THE DIGITALISATION OF AFRICAN AGRICULTURE REPORT 2018–2019 THE DIGITALISATION OF AFRICAN AGRICULTURE REPORT 2018–2019 1ST EDITION, JUNE 2019 2 THE DIGITALISATION OF AFRICAN AGRICULTURE REPORT “Digitalisation is crumbling all sorts of borders and African agriculture will be deeply impacted. Technologies can help stimulate innovation for sustainable agri-food systems and produce better and safer food while preserving natural resources and biodiversity. But we need to be conscious and support solutions that are sustainable and that are tailored to countries’ needs, and embedded into conducive and broader innovation systems. This is in line with the EU’s Digital4Development and SDGs agendas that we are proudly promoting. ” Leonard Mizzi Head of Unit at the European Commission, Directorate-General (DG) for International Cooperation and Development This work has been made possible with the financial assistance of the European Union. However, the contents remain the sole responsibility of its author(s) and can under no circumstances be regarded as reflecting the position of CTA, its co-publisher or the European Union, nor of any country or member State. The user should make his/her own evaluation as to the appropriateness of any statement, argument, experimental technique or method described in the work. This work is the sole intellectual property of CTA and its co-publishers, and cannot be commercially exploited. CTA encourages its dissemination for private study, research, teaching and non-commercial purposes, provided that appropriate acknowledgement is made: – of CTA’s copyright and EU financing, by including the name of the author, the title of the work and the following notice “© CTA 2019 EU financing”, – and that CTA’s or its co-publishers’, and European Union’s endorsement of users’ views, products or services is not implied in any way, by including the standard CTA disclaimer. All requests for commercial use rights should be addressed to publishing@cta.int. THE DIGITALISATION OF AFRICAN AGRICULTURE REPORT 3 CONTENTS GLOSSARY 4 WRITTEN BY: ABBREVIATIONS AND ACRONYMS 8 Michael Tsan, Dalberg Swetha Totapally, Dalberg FOREWORD 10 Michael Hailu, CTA ACKNOWLEDGEMENTS 13 Benjamin K Addom, CTA OVERVIEW 14 EDITORIAL COORDINATION: EXECUTIVE SUMMARY 16 Michael Tsan, Dalberg Key f indings 18 Benjamin K Addom, CTA Challenges 22 Toby Johnson, CTA Recommendations 23 Murielle Vandreck, CTA CHAPTER 1 26 EDITED BY: Why Africa needs a digitally-enabled agricultural transformation Jesse Lichtenstein, Dalberg Martin Schnapf, Dalberg CHAPTER 2 32 Bianca Beks, CTA The D4Ag ecosystem LAYOUT AND CHAPTER 3 96 INFOGRAPHICS: The evolution of D4Ag solutions Mercer Design, United Kingdom CHAPTER 4 122 Where we are headed TEMPLATE DESIGN: Hero, South Africa CHAPTER 5 156 What it will take to accelerate growth and impact PRINTED BY: Proud Press, The Netherlands CHAPTER 6 164 Recommendations for furthering a sustainable, inclusive D4Ag agenda PHOTOGRAPH CREDITS: Cover: © CTA ANNEX 1: COUNTRY CASE STUDIES 174 Ethiopia 175 © CTA 2019 Ghana 178 Nigeria 181 ISBN: 978-92-9081-657-7 Senegal 185 Kenya 188 Rwanda 191 Sahel 195 ANNEX 2: STAKEHOLDER CONSULTATIONS 200 ANNEX 3: METHODOLOGY 202 ENDNOTES 208 BIBLIOGRAPHY 228 4 GLOSSARY GLOSSARY Active use Blockchain Use of a digital solution frequently enough to obtain A digital database containing information such as or even maximise its target benefits. records of individuals, land, and financial transactions that can be simultaneously used and shared within Addressable market a large decentralised, publicly accessible network The potential revenue size of the market that can be (‘distributed ledger’) and memorializes transactions addressed by existing solutions. between parties efficiently and in a verifiable and permanent way. Advisory and information services Digitally delivered information on topics such as Bundling agronomic best practices, pests and diseases, Marketing and distribution strategy that joins multiple weather, and market prices, as well as more products or services together to sell them as a single sophisticated digital services and farm management combined unit in order to deliver more value to software tailored to the specific farmer, farm, or field consumers and/or more economic benefits to the that enable smallholder farmers to make decisions business offering the products; in the context of this that maximise output from their land, improve the report, refers specifically to solutions that cover two or quality of agricultural production, and maximise farm more D4Ag use cases. revenues and profits via lower costs of production, improved ability to identify markets, and/or better Climate resilience price realisation. Climate resilience is the ability to prevent climate- related disasters and crises as well as to anticipate, Agribusiness absorb, accommodate or recover from them in a Businesses collectively associated with the production, timely, efficient and sustainable manner. This includes processing, and distribution of agricultural products, protecting, restoring and improving food and including business entities involved in the production agricultural systems under climate threats that impact and distribution of agricultural inputs and machinery to food and nutrition security, agriculture, and food farmers and those involved in purchasing, aggregating, safety/public health. processing, and distributing farm produce. Climate-smart agriculture Agricultural transformation Climate-smart agriculture is an approach for A state in which agriculture is a vibrant, modern, and transforming and reorienting agricultural production sustainable business that creates value for farmers, systems and food value chains so that they support entrepreneurs, youth, and women, and produces sustainable development and can ensure food security affordable, nutritious, and healthy food for all. under climate change. Artificial intelligence (AI) Crowd-farming Crowd-farming uses digital platforms to link farmers AI is defined as the ability of machines and systems who need capital with sponsors who wish to invest; to acquire and apply knowledge, and to carry out a form of ‘crowd-sourced’ financing in the agriculture intelligent behaviour. context. Big data Large, diverse, complex data sets generated from Data infrastructure instruments, sensors, financial transactions, social Data collection and analytics tools and systems, as media, and other digital means, and typically beyond well as the resulting data assets (e.g., farmer registry, the storage capacity and processing power of land registry, soil, pest and disease databases) that personal computers and basic analytical software.. are relevant to smallholder farmers and/or those who work with them. Big tech Big multi-national hardware, software, and social media companies like Google, Microsoft, Alibaba, IBM, and SAP. GLOSSARY 5 Digitalisation for agriculture (D4Ag) Financial service provider (FSP) Digitalisation for agriculture (D4Ag) is the use Enterprises engaged in the delivery of financial of digital technologies, innovations, and data to services and products including commercial banks, transform business models and practices across the insurers, payments companies, microfinance agricultural value chain and address bottlenecks in, institutions (MFIs) and savings and credit cooperative inter alia, productivity, postharvest handling, market organisations (SACCOs). access, finance, and supply chain management so as to achieve greater income for smallholder farmers, Fintech improve food and nutrition security, build climate Enterprise(s) in the financial sector that either provide resilience and expand inclusion of youth and women. financial services to consumers directly by making use of software and digital communication channels Drone or utilize digital technologies to deliver business-to- Remote-controlled pilotless aircraft that have many business services to financial service providers. applications for agriculture field surveillance and remote diagnostics of agronomic conditions such as Geodata plant and crop diseases, water resources, and soil Information about a geographical location held in quality. a digital format; also called geospatial data and information, georeferenced data and information, as Engaged user well as geoinformation. Farmers who are registered for digital solutions and use them to some extent, but not necessarily to the Geo-referencing level that could be called active or intensive use. Also Adding coordinate information to a digital image such see ‘Active use’. as a scanned map to enable the mapping software to match the map with its real-world location. Enterprise resource planning (ERP) Global positioning system (GPS) Software that digitalises and helps manage and System showing the exact position of an object on integrate core business processes like supply chain earth using satellite signals. operations, logistics, reporting, financial tracking, and human resource activities. Information communication technology for agriculture (ICT4Ag) Extension Use of Information and Communication Technologies An agricultural extension service offers technical (ICTs) in the agricultural sector. In this report we advice on agriculture to farmers, and also supplies distinguish between ICT4Ag approaches that have them with the necessary inputs and services to support characterised earlier efforts to digitalise African their agricultural production. agriculture from the new D4Ag era which involves a broader set of digital tools (i.e., machine learning, big Farmer information services (FIS) data analytics, Internet of Things), wider array of use Services that provide more general advisory cases, and a distinctly more commercial and market- information on agronomic best practices (e.g., based focus for business models. growing, harvesting, post-harvest treatment, storage, inputs, and market prices) without tailoring the Internet of things (IoT) recommendations beyond national, value chain, System in which devices including mobile phones, or district levels. sensors, drones, and satellites, are connected to the internet. Financial access Digital financial services (DFS) relevant for smallholder Machine learning farmers, such as digital payments, savings, Giving computers the ability to learn through analysis smallholder credit, and agricultural insurance, which of big data. increase financial access and equip smallholder farmers to improve yields and incomes and invest in the longer-term growth of their farms. 6 GLOSSARY Macro agricultural intelligence Pastoralists Data analytics solutions and digital decision support Those whose primary occupation is extensive grazing tools that integrate a variety of data sources on on rangelands for livestock production; distinct from smallholder farmers, farms, and markets and convert agro-pastoralists, whose livelihoods depends on both this information into useful country- and value-chain- livestock production and land-based agricultural level insights and decision tools for government cultivation, and who are typically included within the policymakers, extension agencies, agronomists, smallholder farmer definition. agribusinesses, and investors. Pay-as-you-go (PAYG) Market aggregation Digitally-enabled business models in which services Undifferentiated marketing where consumers are are paid for remotely with small, frequent payments treated as a single group. such as daily or weekly installments, and where the product (e.g., off-grid solar water irrigation pump) Market linkages can be remotely deactivated or blocked in the case of Digitally-enabled solutions that link smallholder non-payment. farmers to high-quality farm inputs (e.g., seeds, fertilisers, herbicides/pesticides), to production and Precision agriculture advisory post-harvest machinery and mechanisation services Precision advisory services represent a move from (e.g., irrigation, tractors, cold storage), or to off-take generalised best practices to recommendations markets, including agro-dealers, wholesalers, retailers, tailored to individual agroclimatic conditions (e.g., or even to the end-consumer. weather, soil, etc.), crop varietals, and the economic setting of the farm (e.g., input prices, market prices, Market penetration and market distances). The share of the market that is being reached by a product or a service, typically computed as a share of Registrations a total population or share of total market economic Registrations refer to farmers enrolling in or signing up value (e.g., share of sector revenues or profits). Also for D4Ag solutions. The form of registration depends see ‘Addressable market’. on the type of solution. Mechanisation access services Remote sensing Digital solutions that extend farmer access to Process of gathering information about objects on agricultural machinery or mechanised farm services earth from a distance using aircraft or satellites. (e.g., irrigation, tractors, cold storage). Satellite imaging D4Ag infrastructure/middleware Images of earth collected by satellites. infrastructure D4Ag infrastructure (also sometimes referred to as Smallholder farmers D4Ag ‘middleware’ infrastructure) includes agriculture Individuals who produce crops or livestock on two sector specific data, hardware, and software or fewer hectares of land. Technically speaking this infrastructure that D4Ag solutions rely on to source term only includes farmers and agro-pastoralists who information and deliver their services to farmers and are tied to specific pieces of farmland, but this report other agriculture intermediaries; these are the building uses the term more generally to refer to small farmers, blocks that D4Ag solutions use to do what they do. agro-pastoralists, and nomadic pastoralists. Also see ‘Data infrastructure’. Software-as-a-service (SAAS) Services that can be accessed via the internet rather than through downloading and installing software. GLOSSARY 7 Soil mapping The process of identifying, capturing and depicting soil properties and distribution on a map. Super platform Type of D4Ag solution which bundles together multiple different services for farmers or other smallholder value chain intermediaries and, typically, integrates digital market linkage services, advisory services, and financial services, among others. Supply chain management Digital supply chain management solutions are business-to-business services that help agribusinesses, cooperatives, nucleus farms, input agro-dealers, and other smallholder farmer value chain intermediaries to manage their smallholder relationships. Pest and disease surveillance Monitoring at regional, national, or even farm and field levels to record the prevalence and severity of pests and plant diseases; typically goes beyond simple monitoring to include early warning and advice on pest and disease management. Weather and climate infrastructure Physical (e.g., weather base stations) and digital infrastructure for collecting and recording data on climatic conditions and weather at various levels of geographic granularity, from regional weather patterns down to the agroclimatic conditions (e.g., level of precipitation and temperature) for a farm or specific farm field. Unmanned aerial vehicle (UAV) Aircraft that carry no human pilot or passengers. Also see ‘drone’. Unstructured supplementary service data (USSD) A global system for mobile (GSM) technology in which a user can send messages between a mobile phone and an application programme, including prepaid roaming and mobile chatting, in the network. Fredrick Omondi, CTA Weather index-based insurance Agricultural insurance that uses a weather index such as rainfall to determine pay-outs, thus allowing the system to manage weather and climate risk. 8 ABBREVIATIONS AND ACRONYMS ABBREVIATIONS AND ACRONYMS ACRE Africa Agriculture and Climate Risk Enterprise CGAP Consultative Group to Assist the Poorest AfCFTA African Continental Free Trade Area CGIAR Consortium of International Agricultural AfDB African Development Bank Research Centres AfSIS Africa Soil Information Services CSA climate-smart agriculture API application programming interface CTA Technical Centre for Agricultural and Rural Cooperation ARPU average revenue per user CTIC Conservation Technology Information ATA Agricultural Transformation Agency Centre (Ethiopia) D4Ag digitalisation for agriculture BMGF Bill and Melinda Gates Foundation DRC Democratic Republic of Congo BMZ Bundesministerium für Wirtschaftliche Zusammenarbeit (German Federal DSG digital savings group Ministry for Economic Development ERP enterprise resource planning Cooperation) ESIPPS Environmental Surveys, Information, CAGR compound annual growth rate Planning and Policy WorldBank ABBREVIATIONS AND ACRONYMS 9 eWTP Electronic World Trade Platform LMIC low- and middle-income country FAO Food and Agriculture Organization of MFI microfinance institution the United Nations MNO mobile network operator FSP financial service provider MUIIS market-led user-owned ICT4Ag-enabled FtMA Farm to Market Alliance information service in Uganda G4AW Geodata for Agriculture and Water of MPCI multi-peril crop insurance the Dutch Ministry of Foreign Affairs NAERLS National Agricultural Extension and GDP gross domestic product Research Liaison Service (Nigeria) GIZ Deutsche Gesellschaft für Internationale OECD Organization for Economic Zusammenarbeit Cooperation and Development GODAN Global Open Data For Agriculture and PE private equity Nutrition PFJ Planting for Food and Jobs GPSDD Global Partnership for Sustainable PIP Priority Investment Program Development Data GSMA ROSCA rotating savings and credit association Global System for Mobile Communications Association SAAS software as a service HH household SACCO savings and credit cooperative IBRD organisation The International Bank for Reconstruction and Development SARL société anonyme à responsabilité ICT limitée information and communication technology SDG Sustainable Development Goal (UN) ICT4Ag information and communication SDS security and development strategy technology for agriculture SFSA Syngenta Foundation for Sustainable IFC International Finance Corporation Agriculture IFPRI International Food Policy Research SHF smallholder farmer Institute SMS/IVR short message service/interactive voice ILRI International Livestock Research Institute response IoT Internet of things SNS Smart Nkunganire System (Rwanda) iSDA Innovative Solutions for Decision TAM total addressable market Agriculture UAV Unmanned aerial vehicle (i.e., ‘drones’) ISF Initiative for Smallholder Finance UCFA Uganda Coffee Farmers Alliance ISRIC International Soil Reference and UN United Nations Information Centre USAID United States Agency for International IVR interactive voice response Development KALRO Kenya’s Agriculture and Livestock USSD unstructured supplementary service data Research Organisation VAS value-added service KAOP Kenya Agriculture Observatory Platform VC venture capital KPI key performance indicator VSLA village savings and loan association KPOGT Kalangala Palm Oil Grower’s Trust 10 FOREWORD FOREWORD for Africa’s smallholder farmers and pastoralists. It could drive greater engagement in agriculture from women and youth and create employment opportunities along the value chain. There has been significant growth in digitalisation for agriculture (D4Ag) over the last ten years. In 2019 both the European Union-African Union Task Force Rural Africa Report (TFRA) and the Communiqué from the Global Forum for Food and Agriculture (GFFA) highlighted the power of digitalisation in transforming agriculture. CTA However, despite growth, progress towards D4Ag has been somewhat slow to serve the smallholders Michael Hailu, Director, CTA that produce 80% of Africa’s agricultural output. Nevertheless, the opportunity is there. Agriculture is Agricultural transformation is a priority in the policy expected to be a trillion-dollar market by 2030, ripe for agenda of African governments in their quest to innovation that will drive greater efficiency, sustainable meet the challenges of food and nutrition insecurity, increases in productivity, yield and income. climate change, youth unemployment and overall economic growth. With the right policies, innovation At CTA we staked a claim on this power of digitalisation and investment, the continent’s agriculture could be to more systematically transform agriculture early on. transformed into a powerhouse not only to feed a Digitalisation, focusing on not individual ICTs but the growing population but to create decent employment application of these technologies to entire value chains, for millions of young people. is a theme that cuts across all of our work. In youth entrepreneurship, we are fostering a new breed of Technology, as we have seen in other sectors, is critical young ICT ‘agripreneurs’. In climate-smart agriculture to affecting change and driving development. It is multiple projects provide information that can help bringing countries closer together, reducing barriers towards building resilience for smallholder farmers. to trade and offering a window of opportunity to And in women empowerment we are supporting ‘digital native’ youth entrepreneurs at the vanguard digital platforms to drive greater inclusion for women of innovation applied to different economic sectors. entrepreneurs in agricultural value chains. In agriculture, digitalisation could be a game changer in boosting productivity, profitability and resilience to In other words, at CTA, we know and understand climate change. the power to digitalise African agriculture. But we also understand that the evidence that will attract targeted An inclusive, digitally-enabled agricultural transformation investments to further develop D4Ag on the continent could help achieve meaningful livelihood improvements is lacking. “With the right policies, innovation and investment, the continent’s agriculture could be transformed into a powerhouse not only to feed a growing population but to create decent employment for millions of young people. ” FOREWORD 11 “And as long as we learn from lessons, do it right and manage risks and take into account data sovereignty, inclusivity, sustainability, we will all benefit. ” We realised that it is time to chart the scale of the They say data is the new oil. While I prefer a more opportunity and make some projections that will help in sustainable analogy, for Africa it is certainly the case that guiding policy and investment decisions. It is why we data might be the fuel that drives the transformation of have produced this report together with Dalberg Advisors smallholder farming and keeps the continent on track to and supported by a high-level Advisory Council bringing meet its food and nutrition demands into this century together the key stakeholders that have been engaged in and beyond. All the indicators point to a market that the space. The report is the first attempt to consolidate is ripe for investment now. And as long as we learn evidence and provide proof of impacts and the from lessons, do it right and manage risks and take into knowledge that will allow evidence-based investments. account data sovereignty, inclusivity, sustainability, we will all benefit. While, in the report, we find a young sector, it’s clear that the appetite for D4Ag is burgeoning. However, This report is a valuable first step, we have seen an without the right policy focus and investment there appetite to continually improve our understanding of is a danger that the development will be piecemeal, the D4Ag landscape and chart the opportunity it offers neither sustainable nor inclusive. To capitalise on for entrepreneurs, investors and governments. I hope this opportunity we need to ensure that development our efforts will be valuable in guiding the opportunity is coordinated, that best-practices are shared and a and look forward to the collaborative push that collaborative approach to rolling out and scaling-up I believe will bring D4Ag to life for the benefit of digital innovation, primarily focused on increasing use by Africa’s smallholder farmers and food and nutrition farmers, is adopted. security across the continent. With the baseline that this report provides I believe we are well positioned to start scaling out solutions through partnerships, linking solutions providers, farmers’ organisations, governments, development partners and others. Michael Hailu, Director CTA 12 ACKNOWLEDGEMENTS ACKNOWLEDGEMENTS 13 ACKNOWLEDGEMENTS The design, research and write-up of this report would Dreams TECH Ltd), Christophe Larose and not have been possible without the 120+ agribusiness Milena Pirolli (European Commission DEVCO), leaders, experts and solution providers who shared their Samia Melhem (The World Bank), Christian Merz time, experience and knowledge. We are also grateful (GIZ), Natalia Pshenichnaya (GSMA), Ishmael to the 175 D4Ag enterprises that took the time to Sunga (Southern African Confederation of Agricultural participate in our survey, providing rich insights that Unions), Kentaro Toyama (University of Michigan), strengthened the study. Carola van Rijnsoever, Mariska Lammers and Paul van de Logt (Dutch Ministry of Foreign Affairs) The Technical Centre for Agricultural and Rural and Simon Winter and Robert Berlin (Syngenta Cooperation, (CTA) commissioned Dalberg Advisors Foundation). The members of the Advisory Council – to lead the development of this study. along with several individuals within their respective organisations – provided valuable guidance, input and Dalberg is a strategy and policy advisory firm dedicated support throughout the course of the study. to global development. Dalberg was established in 2001 with the mission of bringing the best of private sector The Digitalisation for Agriculture Steering Committee strategy to address global development challenges by at CTA was central to this effort, guiding the combining rigorous analytical capabilities with deep development of the report and sharing their expertise. knowledge and networks across emerging and frontier The Committee’s members were Benjamin Kwasi markets. Dalberg provides high-level strategic, policy and Addom, Caroline Figueres, Chris Addison, Debbie investment advice to the leadership of key institutions, Kleinbussink, Giacomo Rambaldi, Isolina Boto, corporations and governments, working collaboratively Ken Lohento, Piet Visser, Sabdiyo Bashuna Dido, to address pressing global problems and generate positive and Toby Johnson. social impact. From Dalberg, Swetha Totapally and Michael Tsan We would like to thank the members of the Advisory were the lead authors of the report, with research, Council. The Advisory Council was led by analysis, and writing contributions from Anders Michael Hailu (CTA) and included the following Enghild, Patrick Quigley, Zoe Savellos, and individuals: Vanessa Adams (Alliance for Green Pooja Singhi. Robin Miller and Naoko Koyama Revolution in Africa), Debisi Araba (The International offered advisory support. Jesse Lichtenstein and Centre for Tropical Agriculture), Enock Chikava Marty Schnapf provided editing support and and Stewart Collis (Bill and Melinda Gates Simon Mercer from Mercer Design designed the Foundation), Martin Fregene, Ed Mabaya and report. The photographs from this report are drawn Kemi Afrun-Ogidan (African Development Bank), from CTA’s internal photo bank. Anita Gardeva and Selina Kim (IBM), Clara Colina, Mikael Hook (Rural and Agricultural Finance Learning Lab), Su Kahumbu (CEO Green For more information, e-mail us at press@cta.int. Yieldgap Project Group, 2016 14 OVERVIEW TECHNICAL CENTRE FOR AGRICULTURAL AND RURAL COOPERATION ACP-EU (CTA) Established in 1983 and headquartered in Building on earlier efforts and as part of the research Wageningen, Netherlands, CTA is a joint for this report, CTA is now tracking ~400+ D4Ag international institution of the African, Caribbean organisations across Africa, including NGOs, social and Pacific (ACP) Group of States and the enterprises, government initiatives and purely European Union (EU). CTA is primarily funded commercial ventures that are (1) offering digitally- by the European Development Fund and receives enabled agriculture services directly to smallholder additional funding through a diverse set of farmers or (2) as business-to-business solution providers, international partners. extending digital agriculture products and services to other entities that interface with farmers. CTA promotes food security, resilience and inclusive economic growth in Africa, the Caribbean and the CTA’s current programmes target 900,000 farmers Pacific through innovations in sustainable agriculture and and expect to reach 2 million farmers by 2020. CTA’s actively engaging partner organisations for joint action activities directly contribute toward achieving the UN’s and knowledge sharing. CTA focuses on digitalistion, Sustainable Development Goals with a specific focus on youth entrepreneurship, and climate resilience as its SDG 2 (zero hunger, food and nutrition security and priority intervention areas. sustainable agriculture). CTA’s efforts in D4Ag also map to the European Union’s Digital for Development CTA’s work on digitalisation, in particular, focuses agenda as it supports programmes that advance digital on increasing the profitability and productivity of infrastructure and regulatory reforms, digital literacy and smallholder farmers by leveraging digital solutions and skills, and digital entrepreneurship and employment. strengthening business innovations. It promotes precision agriculture solutions, weather information, soil sensors, CTA aims for this D4Ag report to be a foundational drones for agriculture (where CTA is the key convener of and regularly updated piece of research, which should the African UAV4Ag community) and other data-driven serve as a valuable resource for the entire African D4Ag farming practices, as well as new services for farmers in community, as well as an important tool in advancing the areas of finance and insurance. CTA’s digitalisation the D4Ag knowledge agenda in the years to come. work is closely linked to its other programmatic areas, including a focus on youth entrepreneurship in digital agriculture and the promotion of digitally-enabled, climate-smart agriculture solutions. CTA CHAPTER 1 15 16 EXECUTIVE SUMMARY EXECUTIVE SUMMARY Fredrick Omondi, CTA Context and methodology Against this backdrop, digitalisation Agricultural transformation remains one for agriculture (D4Ag) can be a game of Africa’s most pressing priorities but changer in supporting and accelerating has been difficult to achieve. The statistics agricultural transformation across the are well-known: Africa, especially Sub-Saharan continent. D4Ag addresses a wide scope of Africa, (SSA), needs to double (and perhaps factors and conditions affecting farms, farmers even triple) current levels of agricultural and the agri-food sector as a whole. The productivity to meet continental demand and volume of data – and the supporting layer of stave off food and nutrition insecurity.1 The new digital agricultural solutions – is growing continent must achieve these targets while exponentially at the same time that the quality simultaneously adapting to climate change. of that data is rapidly evolving. For the first Climate change is already impacting the time, it is possible to precisely capture data agricultural sector with increasing climate from individual farms and fields, combine it in volatility and the destructive effects of macro-level data sets, and utilise those sets in droughts, floods, new pests and diseases. With increasingly cost-effective ways. Why are digital so much at stake, it is no surprise that most solutions and agriculture data potentially so African countries have prioritised agricultural transformative? For farmers, they offer access transformation as a key pillar of their national to tailored information and insights that allow strategies. Yet, as the African Union’s 2018 individuals to optimise their production, gain biennial review of the Malabo Declaration access to appropriate products and services, shows, fewer than half of countries (20 out and explore new linkages with markets. D4Ag of 47) are currently on track to meet their provides enterprises deeper understanding commitments by 2025. of their target segments, allowing them EXECUTIVE SUMMARY 17 “D4Ag can be a game changer in supporting and accelerating agricultural transformation across the continent. ” to better tailor their interventions to the costs, is crucial to achieving real agricultural needs of smallholder farmers. Governments, transformation and impact. While it may not likewise, can use improved understanding of be a cure-all, it is clear that D4Ag’s potential farmer segments to improve macro-decision to contribute to Africa’s inclusive growth story policy-making, as well as the design and is significant. implementation of their programmes. The result – if fully implemented at scale – would In this report, we set out to explore the be a highly connected, intelligent, real-time gains D4Ag has made toward reaching agricultural ecosystem that is vastly more its potential. Our ambition, therefore, is productive, efficient, and transparent than for this report to serve as a barometer ever before. The growing quantity and quality for the current state of D4Ag in Africa. of agricultural data and digital agricultural Specifically, we (i) define D4Ag and establish solutions significantly reduce the costs of service, a common language for the sector – the inputs, and information delivery for farmers solutions, their use cases, and their potential; and other value chain intermediaries. This (ii) share how far the sector has advanced as enables them to productively transform their of 2019; (iii) offer our perspective on where CTA traditional business models. the sector will go in the next 3–5 years; and (iv) shed light on what it will take to further D4Ag has the potential not only to unlock the potential of the sector and explore support agricultural transformation but the roles of different stakeholders. to do so sustainably and inclusively. An inclusive, digitally-enabled agricultural Our findings are based on the triangulation transformation could help achieve meaningful of an extensive set of primary and secondary livelihood improvements for Africa’s 250 sources. These include (i) a survey that was million smallholder farmers and pastoralists.2 sent to 430 D4Ag enterprises, with 175 It could drive greater engagement in responses received; (ii) a database that tracks agriculture from women and young people and 390 active D4Ag solutions in Sub-Saharan support employment opportunities along the Africa and more than 70 defunct solutions agricultural value chain – and it could help with detailed information (where available) build resilience to climate change. Still, D4Ag on each, including type of business model, is not a replacement for physical infrastructure, reach, geographic presence, revenue and human networks and human interaction. impact; (iii) interviews with more than Digital tools can improve market efficiency, 120 agribusiness leaders, technology experts, transparency, aggregation, and integration, but D4Ag solution providers, donors, investors, parallel investments in physical infrastructure policymakers and academics; (iv) field visits (e.g., roads and electricity) are still needed to and country case studies in Ethiopia, Nigeria, deliver inputs to farmers and to deliver farm Senegal, Ghana and Rwanda, as well as products to market. Furthermore, human lighter touch reviews of Kenya and the Sahel infrastructure (e.g., extensions, financial agents, region; and (v) secondary research on D4Ag agro-dealers, and agent networks), though market assessments, business models, it entails significant investment and ongoing end-user needs and impact evidence. 18 EXECUTIVE SUMMARY Key findings significantly higher than even a few years ago. Importantly, a small but growing Sector reach and growth number of players are developing strong n A large number of players comprise business models and demonstrating that it is this relatively young sector. As of 2019, possible to generate up to €90 of revenue there are at least 390 distinct, active D4Ag per farmer annually, though the average is solutions across the continent.3 As an much lower (e.g., ~€5 for advisory services, indication of how quickly the sector is ~€25 for market linkages, and €4 for digital growing, nearly 60% of these were launched financial service intermediaries and supply in the last three years, and approximately chain management solutions). While the 20% were launched since 2018. The cost structures for generating these revenues, solutions span five major use cases: advisory of course, vary by solution type, there is services, market linkages, financial access, evidence that some companies are able to supply chain management, and macro achieve 30–40% gross margins. We do not agricultural intelligence. Additional use expect all businesses to achieve this level of cases include D4Ag data intermediaries that revenue or margin, but the data indicate focus on multiple downstream solutions. that strong economics are achievable. Furthermore, the amount of bundling is increasing – over 50% of active solutions n The addressable market is in the low combine more than one use case. billions, though only a fraction of it is being realised today. We estimate that n Reach is growing quickly. D4Ag the total addressable market revenue is solutions have already registered over likely €2.3 billion (mid-range estimate, 33 million smallholder farmers and potentially as high as €5.3 billion in 2019), pastoralists across the continent (13% of of which an estimated €127 million of all Sub-Saharan African smallholders and sector revenues (€107–145 million) are pastoralists and up to 45% of smallholder being realised today (~6% penetration of households, depending on assumptions the total addressable market). The used to calculate penetration). The sector addressable market will continue to grow has been growing at about 44% per rapidly over the next decade with the annum over the last three years in terms growth of the smallholder population, of the number of farmers reached (i.e., improvements in connectivity and rising registered for solutions). A small minority of revenues per farmer as D4Ag business companies (about 15, most of which focus models become more established. These on advisory services as their current primary numbers shed light on business opportunities focus) have begun to reach notable scale to significantly grow revenue, but they also with 1 million plus registered farmers each. suggest that D4Ag companies are still working out their business models and likely n The economics are improving, and a need to create more value for farmers and handful of players are beginning to other customers across the value chain. develop viable businesses with attractive financial models. We estimate n Registrations are concentrated. While that 70% of enterprises generate some there are D4Ag solutions present in at revenue and 80% of those revenue- least 43 out of 49 Sub-Saharan African generating enterprises maintain several countries, over half of the solutions are revenue streams. Of our survey participants, headquartered in East Africa and nearly 26% were breaking even. While robust two-thirds of registered farmers across all baseline data are not available for solutions are based in East Africa, with comparison, we believe that these results are Kenya leading the way. Similarly, the EXECUTIVE SUMMARY 19 largest 20 solutions account for nearly 80% of farmer registrations. Moreover, while products are diversifying to address newer use cases like supply chain management, advisory services continue to dominate the market (two-thirds of total registrations). n Investments remain small, and primarily fuelled by donors, while private investment is lagging. Donors are increasingly making D4Ag an important part of their portfolios. We estimate approximately €175 million in annual donor funding flows for D4Ag. Private sector investment is even more limited – in 2018, there was investment of approximately €47 million into African or Africa-focused D4Ag enterprises, including both start-ups and later stage enterprises. Investment into Africa-based D4Ag start- lower – i.e., likely in the 15–30% range, on CTA ups represented 3–6% of all Africa tech average (based on self-reported data) across start-up investment in 2018. Because these all use case areas. figures are not well documented publicly, we likely have not fully captured all private n Some promising impact metrics are investment. Still, these figures are quite emerging. Though early, limited and small relative to the needs of commercial in some cases, mixed, the overall enterprises on the ground and represent results suggest that D4Ag solutions a tiny fraction of the global investment could achieve transformative results. flows to agricultural technology, which by There are not many verified examples yet, some estimates reached nearly €1.8 billion but the few self-reported examples we do in 2017. Most of the funding has gone to have suggest that some D4Ag enterprises specific enterprises; far fewer investments are seeing highly positive direct and indirect have been made in D4Ag infrastructure impacts on smallholder farmers. The (e.g., farmer registries, soil testing greatest amount of evidence points to a infrastructure, weather stations). link between D4Ag and yield and income metrics. Here, a handful of players are D4Ag use and impact leading the way with noteworthy results. n While D4Ag’s reach figures are Evidence for youth engagement and climate impressive given the relative change is early but promising. The link to nascence of the space, use remains employment is largely hypothetical, though low. Our estimates suggest that 42% of also promising. In terms of gender equity, registered farmers and pastoralists actually however, the data suggest that, barring a used the solutions they registered for with handful of exceptions in which companies any frequency. While there is no standard have made a focused effort to reach female definition for ‘use’ and the nature of farmer farmers, the sector has made little progress. interaction with solutions differs depending on the solution type (e.g., digital financial n Yield and income: A sample of product vs. digital advisory service), the approximately 50 impact data points, number of highly active users is likely even including both self-reported and 20 EXECUTIVE SUMMARY divide that must be overcome in order to engage the significant proportion of farmers from older groups. n Climate resilience: D4Ag has likely already helped reduce some effects of climate change by improving resource use (e.g., soil and water conservation due to advisory services), building resilience (e.g., via digitally-enabled agri-index insurance), and lowering postharvest losses for some farmers. However, the number of data points on climate impact is too limited to make compelling generalisations. Experts suggest that we have just begun to see the effects of D4Ag on climate resilience and that we should expect much more progress Eatradehub independently validated impact studies, in this area in the coming years. with average yield improvements across all data points of roughly 20% for advisory n Employment: While the sector currently services, 70% for market linkages, and lacks precise quantitative data or evidence 40% for digital financial services, with on employment impacts, we believe that corresponding income improvements D4Ag will likely be a net job creator. In typically ranging between 20% to 40%. fact, it could even be a significant job Bundled models seem to have increased creator, opening up hundreds of thousands potential. Based on self-reported data, we of jobs in agricultural technology, D4Ag see yield improvements in the range of support, agricultural processing, and 50–300% and income improvements on the agricultural manufacturing jobs. As digital order of 20–100%. While these numbers solutions justify upscaling, digitally-enabled likely represent the most positive outliers, human agent networks will play a critical they are encouraging and demonstrate that role in linking farmers to inputs, finance some players have been able to achieve not and knowledge. It is also possible that just incremental but actually transformative D4Ag could help increase the share of results through D4Ag. Still, it is important smallholders in tight value chains and the to note that these figures represent the total quality of smallholder jobs. impact on the yield and income of digitally enabled solutions, not just the incremental n Women: The relative uptake among impact of digitalisation. Anecdotally, these women is low – especially considering figures are higher than those of purely the disproportionate burden they bear on analogue solutions and are generated at the farm. In sub-Saharan Africa, where reduced cost and thus higher return on 40–50% of smallholder farmers are investment (ROI). Nonetheless, much more women, only 25% are registered users of research needs to be done to quantify the D4Ag solutions. Companies that explicitly advantages of digital over analogue solutions. target female farmers and make this an important measure of their success tend to n Youth: The high share of youth do better. Overall, the data suggest that engagement – more than 70% of registered companies are not sufficiently prioritising users – is good news. At the same time, this gender as part of their product design, figure likely also indicates an important age marketing and user engagement efforts. EXECUTIVE SUMMARY 21 Forward-looking trends Yara, John Deere and UPL – will n Several of today’s barriers – notably, change the sector’s scale and scope. limited access to technology and Many of these players have already begun connectivity – will begin to be to enter the market via exploratory overcome. In particular, we expect that acquisitions, innovative partnerships, and most farmers will have access to a mobile new product development. Others are more phone by 2030 (~50% penetration for quietly holding exploratory conversations unique mobile subscribers in rural and initiating small-scale pilot programmes. Sub-Saharan Africa, but likely 80+%, based Their presence will bring increased on current trends for share of smallholder financial, human and technological resources households that have access to at least one to the sector, and may be accompanied by mobile phone and reasonable connectivity). major investment in important underlying Many will also have access to smartphones infrastructure. Such improvements could – already more than 25% of smallholder significantly improve sector growth. Still, farmers in countries like Kenya and their entry does not replace the need for Senegal report access to smartphones; these strong local talent. The capabilities of big numbers are projected to grow quickly. tech should complement organisations on The cost of data will continue to fall and the ground that are well positioned to growing, thriving mobile money ecosystems design products that can serve the needs of around the continent will serve as a strong farmers in their region and business models foundation upon which to build platforms that will work given local conditions. The for D4Ag transactions. best models will pair localised knowledge with big tech capabilities. n D4Ag products and services will continue to improve. Over one- n We will enter a platform-led era. third of our D4Ag sector survey Platforms that bring together several use respondents already use at least one form cases, diverse value chains, and the best of advanced technology (e.g., drones, capabilities of multiple players are the most blockchain, machine learning, internet of likely to succeed. Such D4Ag ‘super CIAT things, or big data), and nearly 60% of platforms’ are already emerging, with a respondents expect to integrate new range of private, donor-led, government-led, technologies in the next three years. and public-private partnership models. While D4Ag solutions will leverage cutting-edge we cannot predict who will emerge as the technologies—fuelled by new sources of leader(s), and there are likely to be multiple data and analytical capabilities – to reduce different successful models depending on costs, increase their value proposition and the country, we expect that these platform enhance their precision, customisability players, in partnership with some of today’s and overall capabilities even as they leading specialist D4Ag solution providers, become easier for farmers to access and will bring about in a step change in the use. We will move from a state in which D4Ag sector’s reach and impact. we primarily have observational data to a state in which we can offer users real-time n The reach of digital solutions will insights and predictive capabilities. continue to grow and may include as much as 80% of the smallholder n New entrants in the D4Ag space – farmer population. At 44% per annum, including ‘big tech’ players like the sector’s growth rate is currently very Microsoft, Google, IBM, Bosch and high; access to technology is likely the Alibaba, as well as ‘big agri’ main limiting factor for the spread of incumbents like Bayer, Syngenta, D4Ag solutions. Given that Africa will 22 EXECUTIVE SUMMARY achieve near universal phone access in development remains a major barrier: the coming years, current growth trends 49% of D4Ag enterprises that responded suggest that 100 million smallholder farmers to the survey reported that this was a key could be registered for D4Ag services growth challenge. Similarly, 28% of survey within three years and as many as 200 respondents cited consumer-level barriers million smallholders will sign on by 2030. (e.g., digital literacy) as one of the top three This estimate may be high, however, challenges to adoption and use. and a more conservative scenario of ~60 million registered farmers by 2022 is n Most companies are still working probably more credible, as it will become to develop a viable business model. progressively harder to reach additional While some companies have started to smallholder farmers from remote and reach scale and earn profits, the vast vulnerable populations living in less stable majority of businesses still rely on donor and poorly connected environments. funding and continue to experiment with Nevertheless, the core implication of these business models that are attractive to numbers is that reaching farmers will not funders and customers. In recent years, be the main bottleneck for D4Ag solutions; the sectors have learned a lot about rather, the next phase will require a tight what models do not work; we are still in focus on increasing use among and impact the earliest stages of understanding what for smallholder farmers. models work. For example, experience from several businesses suggests that Challenges farmers are unlikely to pay for D4Ag services (especially advisory services) and n The sophistication of D4Ag solutions has begun to outpace the readiness of that data are challenging to monetise. entrepreneurs, users and government Drawing on these experiences, companies actors to embrace and leverage are beginning to experiment with new them. As discussed above, the underlying approaches, e.g., taking a cut of the value technologies and capabilities of D4Ag created for customer segments. This solutions are advancing quickly. We may have strong promise, but companies now have an opportunity to shift focus will have to continue to deliver greater from technologies and solutions to the value to farmers – and thereby translate underlying enabling environment. For customer reach to customer use – in order Fredrick Omondi, CTA example, insufficient human capital to achieve improved business economics. In the meantime, many companies whose full attention is fixed on developing a viable business model deprioritise or miss important issues like impact and data stewardship, viewing them as secondary in importance or even running counter to their objective of turning a profit. n The lack of D4Ag infrastructure – farmer registries, digital agronomy data, soil mapping, pest and disease surveillance, and weather data infrastructure – in most contexts reduces the effectiveness of D4Ag solutions. Such investments are important building blocks for individual enterprises EXECUTIVE SUMMARY 23 and for the D4Ag ecosystem more broadly because they drastically reduce transaction costs, drive efficiency and increase the effectiveness of solutions. Yet, investment in such public goods and enablers is quite limited and just beginning to emerge at national and local levels. The case for making such investments is not always straightforward; based on some existing approaches, they could produce results at the expense of good data stewardship (e.g., customer privacy, appropriate consent, security, etc.). Good data stewardship and strong middleware can coexist, but we have not yet seen a strong focus on this in the sector. n High degrees of country-level and regional variation in investment expose uneven D4Ag growth across the continent. While market-driven and testing the potential of digital solutions CTA growth in D4Ag solutions in countries in agriculture. In the next decade – the like Kenya, Ghana, Nigeria, Senegal, ‘D4Ag’ age – the aim will be to translate Rwanda and Côte d’Ivoire serves as a this potential into reality – and do so strong inspiration for others, the level of equitably and sustainably. As part of this variation across countries highlights some D4Ag journey, the sector made quick important challenges. For example, it strides toward reaching large numbers of indicates that donors, investors and, to farmers in a challenging environment with a somewhat lesser extent, enterprises are an impressive set of products, services and still risk-averse and likely prioritise the innovative business models. easiest-to-reach markets (e.g., markets where other providers already exist and In the next phase of D4Ag, we have an where the ecosystem is stronger). This also opportunity to improve use and drive greater occurs within individual countries, where inclusivity and impact. But we must do so companies largely target the easiest to reach while actively managing the risks of digital customers. Such uneven growth could tools. This will require sector actors to make further worsen the digital divide between several major investments in the improvement different communities. The experience of of business models and especially the D4Ag other base-of-pyramid markets, such as that ecosystem. As we work to mainstream D4Ag, for energy access, suggests that the transfer we recommend that donors, governments of technological innovation from more and investors: advanced geographies to lagging ones is not an automatic process and can, in many 1 Develop human capital at every cases, be quite slow in the absence of level of the D4Ag ecosystem. well-targeted investments and policies. Developing human capacity will be critical to building D4Ag readiness across the ecosystem, Recommendations from farmers to government officials. The The focus over the last 15 years – the necessary growth in human capital includes ‘ICT4Ag’ age – has been on developing increased awareness of D4Ag, improved digital 24 EXECUTIVE SUMMARY Georgina Smith, CIAT literacy and greater digital skill building among than B2C) offerings and deeper research on smallholder farmers and other actors across D4Ag business models will go a long way in the agricultural value chain. Such growth supporting this objective. will require deeper investment across Africa in those sectors of the developer ecosystem 3 Create greater impact by most capable of boosting human capital, i.e., making D4Ag solutions more start-up ecosystems, incubators, accelerators, inclusive of women, other etc. Efforts must also be made to increase the marginalised groups, and capacity of government workers – particularly smallholders in geographies with in ministries of agriculture, livestock, forestry, relatively less D4Ag investment. fisheries and ICT – to understand how to Today, D4Ag solutions primarily reach the use and deploy D4Ag solutions in various low-hanging fruit – farmers in tight value public initiatives. chains – while many enterprises fail to equitably reach women and other marginalised 2 Drive greater business model segments of the community. To achieve sustainability. equitable growth, D4Ag needs to be more Consistent with other sectors and geographies, inclusive. We recommend that governments Africa needs to prove that D4Ag deployments and donors offer greater support for enterprises can be sustainable in order to drive greater in geographies that have historically attracted investment. Key to driving greater business less investment, and that they incentivise model sustainability will be improving value D4Ag enterprises to target marginalised for farmers, identifying and promoting population segments, especially women, who successful business models and mobilising are systematically left behind. Donors, in funding to support a more diverse set of particular, can play a key role in catalysing companies. A focus on improved product greater targeting of marginalised communities. design, support for consortium/platform-based initiatives, continued push toward B2B (rather EXECUTIVE SUMMARY 25 4 Invest in the missing big technology actors expand their footprint. middleware infrastructure. We have an opportunity to manage these Successful D4Ag solutions require risks before they become realities. To do so, access to a wide range of data governments must design approaches that (from remote sensing data to appropriately balance the need for good data farmer-specific data) in order to stewardship with the desire not to overregulate deliver high-quality services to and stifle D4Ag innovation. farmers. These data need to be accurate, reliable and, 6 Invest in the D4Ag in many cases, available in real time. We knowledge agenda. recommend that governments and donors – We still have a long way to go in learning potentially in partnership with private actors – what works and what does not. As the sector lead the development of important agriculture matures, there is a good opportunity to data infrastructure, including digital agronomy develop a set of best practices and a stronger data (e.g., land, water and crop maps), community of practice with which to share soil testing infrastructure and data maps, lessons learned. Development partners will weather/climate tracking infrastructure, digital likely make these investments, with important pest/disease surveillance systems, farmer data contributions from governments and investors registries and agriculture transaction registries alike. We recommend knowledge investments and commodity exchanges. It is particularly in three major areas: how to design offerings important to get the middleware right – from that meet the needs of farmers, in particular design to policy to implementation – so women and other under-served communities; that everything built on top of it works and research to gather better market and business ultimately helps, rather than hurts, farmers. It model intelligence to drive success in D4Ag; is not enough to make these investments in a and research to gather more robust evidence vacuum. Coordination between governments, on the impact created by different use cases donors, investors, farmers and other interested and business models. parties will likely reduce duplication of efforts and result in higher-quality, efficient 7 Create an alliance of key infrastructure that enterprises can rely on D4Ag stakeholders to promote across geographies. greater investment, knowledge sharing and partnership building. 5 Invest in good data stewardship Investment in D4Ag has been isolated, and design for the risks and scattered and piecemeal. Innovations, limitations of digital systems. deployments, investments, assessments and Specifically, we recommend that governments reports are being unnecessarily duplicated. – with support and input from donors – design There is no ‘go-to-place’ or knowledge CTA and implement appropriate policies and clearinghouse for D4Ag across the continent. regulations to promote good data stewardship. With the results of this report as a baseline, Some of these will be specific to agriculture there is an opportunity for a new alliance for (e.g., policies around farmer registration) digitalisation in African agriculture to lead while others will take the form of good data knowledge sharing, collaboration, and growth governance writ large (e.g., consumer privacy, in the sector. This alliance should be built as informed consent, etc.). Such policies are a partnership between governments, donors, critically missing from the conversation today international bodies, farmer organisations (though they are beginning to emerge) and and the private sector dedicated to advancing will become even more important as the sector inclusive, sustainable D4Ag across Africa begins to invest in a middleware layer and and beyond. 26 CHAPTER 1 WHY AFRICA NEEDS A DIGITALLY-ENABLED AGRICULTURAL TRANSFORMATION Image to go here Photo caption and credit to go here CTA Africa needs an inclusive and environmentally sustainable agricultural transformation to build greater food security, improve nutrition, and expand economic opportunity. D4Ag has significant potential to act as a driving force behind Africa’s agricultural transformation in the coming decades. Agricultural Africa must massively and sustainably malnourished people has declined since 2000, transformation increase its agricultural output – to over a fifth of the population in Sub-Saharan A state in which agriculture more than double current levels of Africa experiences chronic undernourishment, is a vibrant, modern and production – over the next three and around 35% of children under five were sustainable business that creates value for farmers, decades to meet growing demand and stunted in 2016.6 Malnutrition causes stunting, entrepreneurs, youth and achieve food and nutrition security.4 wasting, obesity, and anaemia in reproductive- women, and produces affordable, nutritious and Sub-Saharan Africa, in particular, already aged women, among many other health healthy food for all. (CTA) faces the greatest food security risk of any and non-health consequences.7 Agricultural region. By 2050, its population is expected transformation will help farmers increase to increase 2.5-fold while demand for staple productivity, yield, and income, enabling them cereals will approximately triple over this to consume more nutritious food (that they same time period.5 This growth in demand have grown or purchased). For society at large, will substantially outpace the historical rate of agricultural transformation will likely result in agricultural productivity and yield increases lower prices while improved market linkages in the region. Although the number of will result in greater access to nutritious food.8 CHAPTER 1 27 “Climate change is making farmers even more vulnerable than they already were. ” Africa must realise these gains while sweet potato, banana) and horticultural crops Food and nutrition also adapting to climate change and in Africa. The economic upside of such security mitigating further damage to the improved agricultural productivity would be Condition in which all people, environment. Farmers have always been tremendous given the very large share that at all times, have physical, social and economic access to susceptible to climate variability and extreme agricultural activities contribute to regional sufficient, safe and nutritious weather events. Climate change is making GDPs. The Brookings Institution, for food that meets their dietary needs and food preferences farmers even more vulnerable. They are instance, has estimated that a half-ton for an active and healthy life. already experiencing smaller and more increase in staple yields alone could generate (United Nations) variable harvests, new pests and diseases, a 13–20% higher GDP per capita in many and more severe droughts and floods; all developing countries.11 Smallholder farmer indications are that these conditions will all Individuals who produce crops worsen substantially in the coming decades Agricultural transformation can also or livestock on two or fewer hectares of land (World Bank). as temperatures increase and extreme climate serve as an engine for social inclusion. Technically speaking this term events become far more common.9 To achieve There is an opportunity to better engage and only includes farmers and its objectives, agricultural transformation must empower Africa’s women, who constitute agro-pastoralists who are tied to specific pieces of farmland, improve farmer resilience to these climate at 40-50% of the continent’s smallholder but this report uses the term effects. Agricultural production increases must producers.12 Africa also faces a high level of more loosely to refer to small farmers, agro-pastoralists, and also be achieved in ways that limit further youth unemployment with the projected entry nomadic pastoralists. adverse environmental effects of agricultural of over 100 million young Africans into the intensification and cropland expansion – most job market by 2030 and the demographic Agricultural value chain notably, the overuse of natural resources like reality that, for years to come, more than half Set of actors and activities water, soil degradation and biodiversity loss. of Africa’s youth will continue to live in rural that bring a basic agricultural areas.13 Agricultural sector transformation product from production in the field to final consumption, Agricultural transformation has the could have a major role in generating higher- adding value to the product at potential to drive African economic quality jobs and entrepreneurship opportunities each stage. (FAO) transformation by boosting economic growth for Africa’s youth. Such youth engagement through more formal and efficient smallholder in agricultural employment is increasingly Youth farmer value chains, reducing food imports and important given that the average age for an People between the ages increasing agricultural exports (both within and African farmer is 60 years old.14 of 15 and 35 years. (African Union) outside of Africa), decreasing post-harvest losses and improving efficiency in activities such as For decades, many African governments agricultural processing, storage, transport and have recognised the importance of logistics. Dramatically increased production agricultural transformation and the and resulting increases in economic growth are opportunities it presents, yet several possible. McKinsey & Company has estimated complex and stubborn challenges that Sub-Saharan Africa has the untapped have slowed progress. Given its central agricultural potential to double or triple the importance to their near-term future, a few amount of cereal and grain it produces today;10 dozen African countries have already made the potential for productivity gains is equally agricultural transformation a key pillar of large for many other key staple (e.g., cassava, their national strategies and growth plans. 28 CHAPTER 1 However, as of 2018, only 20 out of the Digitalisation can help 48 countries that completed the survey (39%) are on track to meet their Malabo accelerate agricultural Declaration commitments by 2025, according transformation in Africa to the Africa Agriculture Transformation The strategic use of digital technologies, Scorecard.15 There are many reasons why data, and innovative digitally-enabled agricultural transformation has not been easy business models can (and have already to achieve – not least, the large investments begun to) accelerate sustainable required. An estimated €40 billion annually is agricultural transformation in Africa. needed to harness the power of agriculture to Digitalisation for agriculture (D4Ag) is the transform Africa, whereas only approximately use of digital technologies, data and business €6.25 billion is invested annually today.16 model innovations to transform practices Beyond resource constraints, other major across the agricultural value chain and and often interrelated challenges include address bottlenecks in, inter alia, agricultural poor national institutions and weak enabling productivity, postharvest handling, market environments, underdeveloped transportation access, finance and supply chain management and energy infrastructure, insufficient digital so as to achieve greater incomes for connectivity in rural areas, low availability smallholder farmers, improve agriculture value and uptake of high-quality agricultural inputs chain economics for agribusinesses both large and technologies (such as seeds and fertiliser), and small, expand the economic inclusion insufficient water resources, soil degradation, of youth and women, improve overall food limited financial inclusion for farmers, and the and nutrition security and build climate need for improved human capacity and access resilience – all while mitigating the potential to agricultural knowledge.17 negative environmental effects of agricultural Mwanzo Millinga, IFAD CHAPTER 1 29 intensification. Not only can the integration logistics, and how they make decisions about of D4Ag tools help address these important the future. bottlenecks to agricultural transformation, but we also believe it can do so faster and Third, business models are rapidly diversifying more cheaply than status quo, non-digital as many more commercial actors and approaches because improved cost efficiency, investors enter the space. Despite many accelerated innovation, and rapid product challenges, we argue that this augurs well and service dissemination are the hallmarks for the rise of more commercially viable and of digitalisation. scalable digital agriculture platforms. The idea that digital solutions can be Finally, D4Ag is distinguished by its focus used in agriculture is certainly not on data and data systems as the key new. For the past 15+ years, innovators in input and output – the lifeblood – Africa have been experimenting with various of innovative agricultural business information and communication technology models, which we believe will help drive for agriculture (ICT4Ag) solutions. These systemic change rather than just one-off, efforts – which have largely been one-offs project-level improvements. – have helped farmers, agribusinesses and governments become more comfortable D4Ag can help a range of important actors with using technology in the context of in the agricultural ecosystem. We describe agriculture. We refer to these initial efforts as the potential impacts of D4Ag on these Fredrick Omondi, CTA characterising the ICT4Ag age. Now, fuelled stakeholders in Figure 1. In some cases (though in part by the foundations laid by ICT4Ag, not all), we already see some of this potential we have entered the digitalisation for translating into reality. The level of progress agriculture (D4Ag) age. made, relative to the impact potential of D4Ag, is a major area of exploration in a later This is more than a semantic shift – this section of this report. report argues that we are on the verge of dramatically expanded possibilities Beyond supporting individual actors, for the impact of digital solutions on D4Ag has the ability to promote Africa’s agriculture. The era of D4Ag is intra-regional trade. Aside from positive distinguished from what preceded it in at least impacts on smallholders and other individual four ways. agriculture value chain actors, D4Ag should ultimately make an impact on important First, there is a much broader range of macro-economic conditions and priorities. digital technologies that innovators can As an illustration of this potential, one of the draw on beyond basic information collection Malabo Declaration’s priorities is to triple and communication tools (e.g., satellites, intra-regional trade in agricultural products drones, portable diagnostic technologies and by 2025. D4Ag can help the production of sensors linked to the internet of things). surplus products, improve the connectivity of products to various markets and strengthen Second, there is a move from using digital the efficiency, quality and transparency of technologies for information dissemination to supply chains, ultimately making cross-border the true digitalisation of the agriculture trade across markets more attractive and less ecosystem, including digitalising how farmers risky than it is today. D4Ag could similarly and other agriculture value chain participants encourage greater trade between African pay for goods and services (or access finance), countries and nations outside of Africa. how they connect and transact as buyers and sellers, how they manage operations and 30 CHAPTER 1 Figure 1 Potential D4Ag impacts on African smallholder agriculture ecosystem Category Actor Potential D4Ag impacts (non-exhaustive) Smallholder All smallholder • Greater productivity via the dissemination of agricultural advice and real-time information, better farmers farmers and financial access, and improved linkages to quality agricultural input and reliable off-take markets (SHFs) pastoralists • More sustainable farming practices that help maintain productivity over the long term and reduce costs (e.g., water and input use) in the near term • Increased chances to obtain formal land titles thanks to digital mapping of farm boundaries • Increased farmer incomes as farmers produce greater quantities, face lower crop losses and access fairer input and off-take prices • Improved nutritional outcomes of SHFs as they grow, purchase and consume more nutritious food • Inclusion of SHFs in more commercial value chains due to reduced transaction cost and risks Climate- • Better climate resilience through improved weather forecasts, advice on climate-smart agricultural vulnerable SHFs practices, improved access to weather-adaptation inputs and weather index-based insurance Women SHFs • Better understanding of women farmers’ unique needs and tailored design of solutions due to the capture of large volumes of high-quality gender-disaggregated data • Greater access of women farmers to relevant advice, finance, agri-inputs Rural youth • Greater youth interest in agriculture as digitalisation increases sector attractiveness for the young • More jobs and improvement in the quality of existing jobs in agriculture as digitalisation generates new opportunities in farming and farming-adjacent sectors (e.g., farm agents, processing jobs) • New high tech employment opportunities (e.g., D4Ag software development, data analytics) Business Input providers • Expanded farmer demand for input products (increasing revenue) (e.g., agro- • Improved cost-efficiency of input distribution due to digitally linked value chains and digital tools for dealers, input input supply chain management and logistics optimisation producers) • Greater input value chain transparency, traceability and thus input quality (e.g., widespread use of quality assurance and anti-counterfeiting tools to protect brand owners and farmers) Off-takers • Increased volume of high-quality produce from SHFs due to better practices and input use (e.g., buyers, • Enhanced market efficiency and interconnectedness with more integrated and transparent value processors, traders) chains and less wasteful production and post-harvest stages all contributing to growth and profits • Improved quality and safety of food products coming out of smallholder value chains due to digital traceability and tracking tools and digitalised supply chain logistics Financial service • Lower costs to identify, acquire, and service smallholder farmers due to digital channels and tools providers (FSPs) that directly improve FSP profitability and expand potential universe of economically viable clients (e.g., banks, MFIs, • Improved ability to assess, monitor and manage financial product risks via innovative analytics of insurers, payments digitalised farmer, field (e.g., soil), weather and remote sensing data players) • Lower risks of serving farmers due to digitally-enabled delivery of better advice and market linkages Government Agriculture • Support for national macro-objectives such as sustainable agricultural transformation, food and ministries, nutrition security, job creation and improved climate resilience national • Improved cost-efficiency and more targeted impact of government investment into agriculture (e.g., extension less leakage from agri subsidies, more accountable and cost-efficient agronomy and extension) agencies • Much better macro intelligence on agriculture sector trends, opportunities, and risks at national and sub-national levels allowing for improved planning, resource-allocation and crisis management Agronomy CGIAR, National • Improved linkages between upstream agronomy R&D and on-the-ground agricultural product R&D sector Agriculture development and agronomic advice due to richer and more intensive digital data feedback loops Research Centres • Lower costs of collecting field data (e.g., digital tools for data collection and field trial management) (NARS), private • Improved insights for agronomists into farmers’ wants and needs due to large-scale farmer data agronomy actors • Methodological innovation (geospatial agronomy) due to the availability of much greater volumes of remote sensing (satellite/drone) and ground truth (e.g., digitalised field trials and yield measurement) African • Improved food security due to the much wider availability of lower-cost and more nutritious food population • Improved food quality and safety and faster resolution of food safety issues (i.e., due to traceability) at large • New jobs and entrepreneurship opportunities outside of rural areas but linked to agriculture sector (e.g., D4Ag software development, analytics, derivative financial services and trading jobs) CHAPTER 1 31 We are already starting to see important smallholder farmers. Given the size of this signs of progress, as well as notable segment, its vulnerability, and its importance areas for further improvement. This to agriculture in Sub-Saharan Africa, any report serves, therefore, as a barometer of attempt at inclusive agricultural transformation the progress to date and aims to accelerate must prioritise solutions that deliver value digitally-enabled agricultural transformation to African smallholder and pastoralist by establishing a rich, repeatable baseline for households and other smallholder value sector data and highlighting key emerging chain intermediaries. opportunities. At the same time, the report also acknowledges substantial challenges to progress Of course, digital solutions cannot do and offers recommendations for how these it alone. Major challenges and risks challenges could be addressed. In the sections are associated with digitally-powered that follow, we specifically: agricultural transformation. Digitally- • Describe the D4Ag ecosystem, establish enabled transformation cannot sidestep a common language for D4Ag use cases the need for fundamental infrastructure categories and major solution sub-types, and investments (e.g., roads, energy, irrigation) and explore each use case with on-the-ground important improvements in the underlying examples (Chapter 2). agriculture policy environment. Moreover, digitalisation brings real risks. D4Ag will • Share how much progress has been made in likely accelerate the decline in the number the D4Ag sector as of early 2019 (Chapter 3). of agriculture sector jobs in Africa as • Offer perspectives on forward-looking trends consolidation increases. While some farmers that will define the evolution of the sector may benefit from digital technology, others (Chapter 4). could easily fall behind new types of ‘digital divides’. Women, for example, could be more • Shed light on what it will take to unlock the disenfranchised. Finally, digitalisation creates full potential of the sector (Chapter 5). its own, often poorly understood, risks to • Offer perspectives on the role governments, agriculture sector data privacy and information donors, and private actors will need to play security. Given their information constraints to unlock this potential (Chapter 6). and limited economic resources, smallholders are particularly vulnerable to such risks. We Throughout the report, we focus on the reach explore these challenges and consider how to of D4Ag, its use, and how it impacts overcome them in Chapters 5 and 6. CIAT 32 CHAPTER 2 THE D4AG ECOSYSTEM Image to go here Thompson Reuters Foundation The D4Ag ecosystem in Sub-Saharan Africa presents a complex and fast evolving landscape. At the core of the ecosystem – and this report – are five use cases for D4Ag solutions, which are supported by D4Ag infrastructure (e.g., ag data systems), digital enablers like payments, and a general enabling environment layer. D4Ag Solution Landscape livestock management; support for post- – Defining Key Terms harvest activities such as processing, storage The definitions of D4Ag ‘solutions’ and and transport; linkages to buyers and off-take D4Ag ‘actors’ and ‘enterprises’ in this report markets; and cross-cutting value chain activities are intentionally broad to accommodate the such as input and produce quality assurance complexity and dynamism of the sector.18 and the delivery of financial services. The digital solutions covered in this These solutions encompass a wide report span the full smallholder variety of digital technologies and agriculture value chain, including tools, including everything from agronomic pre-production planning; agricultural advice and information delivered via short input production (e.g., seed production message services (SMS) and interactive voice management), marketing, distribution and response (IVR) to smartphone applications ongoing monitoring (e.g., for farm machinery that link farmers to multimedia advisory and irrigation); support for production-stage content, farm inputs, and buyers. There are activities and decisions for farming and business solutions that rely on sophisticated CHAPTER 2 33 software and data analytics platforms to help linkages; (iii) supply chain management; D4Ag ‘solution’ and agribusinesses to manage their smallholder (iv) financial access and (v) macro ‘enterprise’ definitions supply chains; financial technology solutions agricultural intelligence. Each of these that digitise payments or utilise satellite and five use case categories includes many D4Ag ‘solutions’ weather data to analyse the creditworthiness of underlying sub-types of solutions. There is farmers and deploy new types of agricultural also arguably an additional emerging sixth use Products and services that utilise insurance; and agriculture dashboards and case category of D4Ag ‘super platforms’ digital tools, digital channels, or decision tools for policymakers. – end-to-end solutions that cut across all digitally-enabled data analytics (e.g., machine learning/ other use case categories – which we believe AI) to deliver information, The report defines the ecosystem of are a path to the future of D4Ag and are advice, farming input linkages, market access, logistics D4Ag actors broadly, as well, to include thus covered separately. support, financial services, and NGOs, social enterprises, commercial decision-making tools directly to ventures, government agencies and Figure 2 provides detailed definitions of smallholder farmers or to other intermediaries of smallholder others that offer digitally-enabled these use case along with some illustrations value chains, including agriculture services. They may do so directly of the underlying types of solutions for each. extension agents, agro-dealers, agribusinesses, financial service to smallholder farmers or as business-to- Further detail on each use case follows later providers and policymakers. business solutions for entities (e.g., smallholder- in this chapter. focused extension agents, agribusinesses, financial institutions and policymakers) that While donors, investors, implementers D4Ag ‘actors’ interface with smallholder farmers or make and market intelligence actors continue or ‘enterprises’ decisions about smallholder value chains. This to group D4Ag use cases or categorise D4Ag definition is not limited to purely digital individual solutions in a wide variety Organisations, whether commercial or non-commercial, enterprises. Rather, many of these companies of ways,19 the vast majority of D4Ag that develop D4Ag solutions or meld digital products and digital delivery enterprises still primarily focus on that deliver D4Ag solutions to channels with human agents who support the only one of the five discrete use case farmers and other smallholder value chain actors. delivery of advisory, market facilitation, areas proposed in this report. Given the While many D4Ag enterprises logistical and financial services. early stage of many D4Ag business models have only one D4Ag solution on and the rapid pace of sector innovation, any the market, others hold multiple D4Ag solutions with different This report categorises D4Ag solutions terminology scheme for the D4Ag landscape features and customer bases. into five primary use cases: (i) advisory is necessarily provisional. Furthermore, as Some D4Ag enterprises, such as and information services; (ii) market we will cover in much greater depth later in regional MNOs, deploy multiple solutions under different brands in different countries. Giacomo Rambaldi, CTA 34 CHAPTER 2 Figure 2 D4Ag use case definitions and example solutions D4Ag use cases Definition and link to smallholder farming ecosystem Examples of solutions Advisory & Digitally delivered information on topics such as • Agronomic/livestock management good practices information agronomic best practices, pests and diseases, • Market information systems and services (i.e., agriculture services weather and market prices, as well as more input and crop/livestock price intelligence) sophisticated digital advisory services and farm management software tailored to the specific farmer, • Early warning tools for weather/climate advisory or farm or field that enable smallholder farmers to pest/disease control make decisions that maximise output from their land, • Customised (precision) advisory services at the level of improve the quality of agricultural production and farmer, farm or specific field maximise farm revenues and profits via lower costs • Participatory platforms (e.g., peer-to-peer smallholder of production, improved ability to identify markets communities, curated farmer videos) and/or better price realisation. • Livestock and farm management software Market Digitally-enabled solutions that link smallholder • Linkage to agri-inputs (e.g., digitally-enabled input linkages farmers to high-quality farm inputs (e.g., seeds, distribution, online input marketplaces) fertilisers, herbicides/pesticides), production and • Mechanisation linkage platforms (e.g., shared economy for post-harvest machinery and mechanisation services mechanisation, pay-as-you-go irrigation) (e.g., irrigation, tractors, cold storage), or off-take markets, including agro-dealers, wholesalers, • Linkage to market access (e.g., digitally enabled linkages to retailers, or even to end-consumers. Digital market wholesale buyers) linkage solutions allow smallholder farmers to lower • End-to-end integrated market linkage models (e.g., digital their costs of production via access to lower-cost linkage to both inputs and markets) and/or higher-quality inputs, reduce the costs and • Ag buyer-seller digital marketplaces/exchanges risks of finding and transacting with buyers and ultimately increase their yields and incomes. Supply chain Digital supply chain management solutions are • Traceability solutions (e.g., digital sustainability and organic management business-to-business services that help agribusinesses, product certification tracking) cooperatives, nucleus farms, input agro-dealers and • Enterprise Resource Planning (ERP) platforms for smallholder other smallholder farmer value chain intermediaries farmer cooperatives, nucleus farms, to manage their smallholder relationships in ways agribusiness out-grower schemes that lower costs through greater efficiency, improve value chain quality through better traceability and • Digital quality assurance solutions for farm inputs and accountability and ultimately increase smallholder produce farmer yields and incomes by making it easier for • Logistics management solutions for post-harvest cold chains, more commercial players to formally engage with storage and transport large numbers of smallholder farmers. Financial Digital financial services (DFS) relevant for smallholder • Smallholder farmer payment solutions (e.g., agribiz to access farmers, such as digital payments, savings, farmer, government to farmer, farmer to input supplier) smallholder credit, and agricultural insurance, which • Digital agri-wallets and commitment savings systems increase financial access and equip smallholder farmers to improve yields and incomes and invest in • Smallholder credit (e.g., digital credit assessment/delivery/ the longer-term growth of their farms (e.g., via better collection platforms and products) inputs, mechanisation and expansion to new crops). • Smallholder insurance (e.g., digitally-enabled index weather, Also includes business-to-business digitalisation and precipitation, pest insurance) data analytics services for financial institutions that • Crowdfunding platforms for smallholder farming enable such institutions to serve smallholder farmers at • Business-to-business fintech data analytics intermediaries substantially lower cost and risk. (e.g., digital credit profiles) Macro Data analytics solutions and digital decision support • Government agriculture sector tracking dashboards agricultural tools that integrate a variety of data sources on • Agriculture extension system management tools intelligence smallholder farmers, farms and markets and convert this information into useful country- and value-chain- • Agribusiness and agriculture investor national and regional level insights and decision tools for government intelligence systems policymakers, extension agencies, agronomists, • Agronomy/R&D agenda setting digital tools agribusinesses and investors. • Weather and climate observatories for agriculture CHAPTER 2 35 this report, D4Ag enterprises are increasingly diversifying their business models and bundling services in ways that often blur the boundaries between these use case areas and focus on several use cases at once. Despite these caveats, we believe that the use case categorisation scheme proposed in this report is a useful tool for characterising the current state of the sector and for ongoing tracking of how the D4Ag landscape evolves in terms of the number of solutions, the reach of these solutions into the smallholder farmer population, investment trends, technology and business model innovations and impact evidence. Contextualising D4Ag Solutions in the Broader D4Ag Ecosystem While the five D4Ag use case categories and related solutions in Sub-Saharan Africa are the primary focus of this report, these use cases are only the enabling environment, D4Ag infrastructure, CTA top-most ‘application’ layer of a much and individual D4Ag solution use cases and broader digital agriculture ecosystem. illustrates how D4Ag can simultaneously support macro impacts like agricultural transformation To achieve positive impact on smallholder and smallholder-level impact objectives. farmers at significant scale, D4Ag solutions must be supported by strong underlying D4Ag infrastructure (also sometimes referred D4Ag infrastructure as well as by an to as D4Ag ‘middleware’ or ‘midstream overall enabling environment conducive to technologies’) is the most immediately a well-functioning digital agriculture ecosystem. important element of the D4Ag ecosystem for Additionally, to support and accelerate overall ensuring the scale-up and impact of D4Ag agricultural transformation and to ensure that solutions. As illustrated in Figure 4, this digital solutions produce positive impacts for infrastructural layer includes enabling software individual smallholder farmers, the D4Ag and analytics tools, hardware that captures ecosystem must be supported by parallel data fed into agriculture data systems (e.g., developments in the broader agriculture drones; weather stations; soil, pest, and crop sector. These developments include the diagnostics equipment; and field sensors) and a advent of well-designed agriculture policies, wide variety of data assets and systems relevant increased investment in the formalisation for smallholder farmers and farms. of agricultural input and off-take markets, advances in local and regional agronomy Agriculture data systems cover all the research systems and agricultural trade policies. factors that might inform D4Ag solutions, including farmer data (e.g., The D4Ag ecosystem map in Figure 3 farmer registries that uniquely identify farmers outlines the relationships between the overall and capture details on farmers and their 36 CHAPTER 2 Figure 3 D4Ag ecosystem map Macro D4Ag Impacts Other Ag Transformation Drivers Food & nutrition Ag GDP Social inclusion Jobs Environmental Agricultural policies security growth sustainability Input and off-taker markets Nat’l Non-digital infrastructure D4Ag Adoption Regional integration Smallholder Farmer D4Ag Impacts Yields Income Addressing Gender Youth Formal climate change inclusion employment employment Data supports D4Ag decision- Reach making & Use D4Ag Ecosystem D4Ag Solution Use Cases Advisory Market Supply chain Financial Macro agriculture services linkage management access intelligence See detail in Fig 2 D4Ag Infrastructure Ag data (e.g., farmer registries, transactions, soil, weather, remote sensing) See detail in Fig 4 D4Ag software & analytics D4Ag hardware (e.g., diagnostics, sensors) Enabling Environment Business ecosystem Investment/finance Incubation Doing business Human ecosystem ecosystem environment capital Digital enablers Digital Digital Digital Digital and payments ID literacy data policies Connectivity Connectivity networks, access devices, cloud, etc. farms), agricultural transaction and agronomic data (e.g., field trial and field financing data from commodity exchanges, yield measurement data) and, finally, marketplaces or financial institutions, land agronomic good practices content registry data (e.g., land title registries and adapted to local crops and agroclimatic other data assets and tools that geospatially conditions.20 mark farmer’s fields and their boundaries), localised market data on the prices of Successful D4Ag solutions – particularly essential inputs and commodities, soil data those that are customised to a farmer’s (e.g., granular, national-scale soil property needs – are highly dependent for maps), pest and disease surveillance data, their impact and scalability on the localised weather/climate data, sensor availability, quality and cost of such data from sensors embedded in farmers’ fields agriculture data. Agriculture data systems and agricultural machinery, remote sensing at national and regional levels, however, are data (e.g., satellite and drone field maps), often underdeveloped, fragmented, low quality CHAPTER 2 37 Figure 4 D4Ag infrastructure layer Data Farmer Weather Surveillance Transaction Market registries data data data information (e.g., pest, crop, livestock) (e.g., prices, volumes) Soil Agronomic Agronomy Land Crop data content field data data data Software Hardware Machine Blockchain Drones Diagnostics learning equipment IoT Artificial Other In-situ Other intelligence (e.g., CRM, ERP) sensors (e.g., weather stations) or entirely unavailable in most of Sub-Saharan and logistics sensors embedded in post-harvest Africa today. Without these data layers, D4Ag transport and cold chain equipment). solutions can exist (and, of course, do exist), Critical D4Ag software infrastructure but are unable to realise their full potential includes a wide range of field data collection to respond to the specific needs of each tools, agent field-force management tools, smallholder farmer at sufficiently low cost data analytics tools, and software building and with sufficient quality of data-enabled blocks (e.g., blockchains for agriculture, AI insights.21 We return to this topic in Chapter chatbot tools and machine learning algorithms, 5 when the report explores some of the major background enterprise resource planning (ERP) outstanding challenges and investment gaps to and customer relationship management (CRM) D4Ag solution scale-up. modules). At the intersection of hardware and software sit sophisticated new internet of things The data layer, in turn, relies on and (IoT) solutions for smallholder agriculture that interacts with underlying layers of integrate sensor data with analytics, monitoring hardware and software tools that are and remote management tools. either specific to the agriculture sector or adapted to the needs of smallholder farmer Beyond the availability of essential D4Ag agriculture in the developing world. Hardware infrastructure, D4Ag solutions rely on facilitates data acquisition and storage while the broader enabling environment for software facilitates its processing. digital ecosystems. The overall enabling environment drives access and use of the Essential D4Ag hardware infrastructure D4Ag solutions, ensures the creation and includes agronomic diagnostics equipment growth of strong business models and creates (e.g., new types of portable soil, crop and a safe environment for users. The enabling agriculture input testing tools), remote environment includes connectivity, digital surveillance systems adapted for agriculture enablers and the business ecosystem. (e.g., agriculture-focused satellite networks and drone surveillance providers with specialised First and foremost, D4Ag enterprises soil and crop sensors), low-cost hyper-local rely on the reach, capacity and quality weather stations and ‘in situ’ sensors (e.g., farm of connectivity infrastructure. This field sensors, agricultural machinery sensors includes the penetration and accessibility of 38 CHAPTER 2 communication networks and devices – in build their digital finance products. The order to access smallholder farmers and scale incubation ecosystem, most notably local solutions. While many D4Ag enterprises have technology incubator and accelerator hubs, are designed tools that farmers can use with simple often critical to the growth of early-stage D4Ag feature phones via USSD, SMS and IVR, enterprises and the upskilling of young D4Ag other D4Ag business models depend on greater entrepreneurs in Africa. Finally, the overall reach of connectivity for people and devices ‘Doing Business’ environment includes factors (e.g., models reliant on connected field sensors), such as business registration, taxation and improved bandwidth (e.g., for models that investment regulations, all of which affect the involve video content and other data-intensive work of D4Ag enterprises. applications), lower cost of connectivity and much broader availability of smartphones While we firmly believe that D4Ag (e.g., solutions reliant on smartphone infrastructure and broader enabling environment functionality for field diagnostics of pests elements are critically important for the success and diseases or soils). Another part of this over the overall D4Ag ecosystem, these more connectivity layer are cloud services and upstream D4Ag ecosystem elements are not other back-end systems that allow D4Ag the focus of the analysis in the report and enterprises to better leverage data and process warrant separate treatment in future research information, forming a basis upon which to publications. We do, however, touch on the build more sophisticated solutions. status of these enablers to the extent that they help or hurt the evolution of D4Ag solutions in D4Ag solutions also depend on broader Chapters 4 and 5 of this report. digital ecosystem enablers. National-scale digital payments systems, national digital ID D4Ag Solution Use Cases infrastructure, digital literacy promotion efforts, and conducive digital and data policies, – Overview of the Solution particularly with respect to cybersecurity and Landscape data privacy governance – are important The primary units of analysis for this elements of any well-functioning digital economy report are the D4Ag use cases and and thus critical to supporting the success, underlying solutions. Figure 5 provides an scalability, and sustainability of D4Ag initiatives. overview of major examples of D4Ag solutions; For example, a large share of D4Ag solutions the discussion that follows explores each use in Africa today are at last partly dependent on case in turn. or are building on the success of existing digital payments systems such as M-Pesa. Advisory and Information Finally, the overall business ecosystem Services Use Case is an important determinant of the Digital farmer advisory and information success of D4Ag solutions. This broader service solutions offer on-demand business ecosystem includes human capital (pull) or periodically distributed infrastructure and related educational systems (push) information and guidance to that, ideally, support the promotion of general farmers with the objective of helping literacy and help supply the talent for product smallholders adopt better practices – developers, agronomists, and field agents ranging from the types of inputs they should on which many D4Ag solutions rely. The consider to agronomic techniques, post-harvest investment/finance ecosystems support the handling/processing and marketing advice, availability of investment for D4Ag enterprises and overall farm business management tips. In as well the broader financial systems and addition to distributing information to farmers, institutions upon which D4Ag players can like most other D4Ag use cases, D4Ag advisory CHAPTER 2 39 OVERVIEW OF THE SOLUTION LANDSCAPE Figure 5 D4Ag solution use cases and illustrative sub-use cases Advisory • Participatory and peer-to-peer • Farm management software services • Farmer information services • Precision ag advisory Market • Digitally-enabled value chain integrators linkage • Food e-commerce • E-marketplaces • Mechanisation access services Supply chain • Traceability • Quality assurance • Logistics management • Supply chain ERP • Payments Financial access • Crowd-farming • Savings • Insurance • Credit • Fin analytics • FSP digitalisation Macro agriculture intelligence 40 CHAPTER 2 ADVISORY AND INFORMATION SERVICES Figure 6 Advisory services – sub-use case overview and examples of solutions Precision agriculture advisory • Weather/climate and pest & disease • Remote sensing/satellite imaging • Drone/UAV surveillance • Plant health and soil portable diagnostics • Field sensors • Integrated precision advisory platforms Farmer information services Farm management software Advisory • Dairy services • Poultry • Crop Participatory advisory • Call center/IVR • Interactive chatbots • Peer-to-peer CHAPTER 2 41 and information services solutions often include advisory, and self-service farm intensive data collection from farmers in order management solutions. to improve the quality and relevance of the advice and information they deliver and, at the Advisory Services – same time, to generate a flow of valuable data Farmer Information back to agribusiness, public/NGO extension systems and, in rarer instances, the agronomy Services R&D community. Farmer information services provide relatively general agricultural These types of services, which are delivered information and advice on agronomic either directly to farmers’ phones or with best practices (e.g., planting, harvesting, pest the support of intermediaries like extension and disease management), farming inputs, the agents, financial agents, and agribusiness field weather, and market information (e.g., prices forces can play an important role in helping for key inputs and commodities), typically via smallholders improve their yields and thereby SMS, USSD, and IVR, and occasionally with increase overall productivity, income, and call centre support. Recommendations are not food security. traditionally tailored beyond national levels or general crop types. Farmers access the advice Over the past several years, advisory and information directly, as is the case for most services have become far more advisory service solutions tracked in this report, sophisticated. Historically, digital farmer or via agents such as government extension advisory and information services have focused officers, NGO staff, agribusinesses agents, on packaging and delivering generic best financial service provider agents, and lead practices to farmers. More recently, by better farmers. In such intermediated models, agents tailoring information and advice for individual use digital advisory tools and information farmers, improving the quality of the content repositories to deliver support to individual they deliver, continuously lowering the costs smallholder farmers or farmer groups. CTA of service delivery, bundling advisory solutions with other higher margin services like market linkage, and finding new partnership models for scale (e.g., by partnering with mobile network operators (MNOs), governments, and agribusinesses), a growing number of advisory solution providers have achieved dramatically increased farmer registrations, deeper farmer engagement, and in some instances stronger economics (though, as discussed in Chapter 3 of this report, the economics for many D4Ag advisory services enterprises still remain precarious with limited per farmer revenues and razor thin or negative margins in the absence of subsidies). Advisory services can be sub-segmented into different sets of often overlapping categories. Some major sub-types of advisory solutions worth highlighting are farmer information services, precision agriculture advisory, participatory 42 CHAPTER 2 These farmer information services (i.e., informed by the GPS location of the constituted the majority of the early farm and other specifics of the smallholder wave of ICT4Ag innovators in Africa client), have greater focus on weather and a decade ago. They are exemplified by climate information, and have the tendency enterprises like Esoko in Ghana (in its to bundle other services, such as market earlier stages),22 Grameen Foundation’s linkages, alongside farmer information. Most Community Knowledge Worker (CKW) of the players in this category also now solution in Uganda,23 many early donor-funded have diversified revenue models beyond the ‘e-extension’ agriculture projects from NGOs farmer usage fees and donor subsidies that like Catholic Relief Services (CRS), and were typical of earlier solutions. They now most of the initial MNO-linked agriculture tend also to pursue commission fees and data value-added service (mAgri VAS) solutions24 monetisation revenues from agribusinesses and, like Tigo Kilimo in Tanzania and M-Kilimo in some cases, cost coverage or cost-sharing in Kenya. Many such solutions from that first from MNOs interested in adding value to their wave of innovators are currently defunct. smallholder farmer customers. A large share of existing digital advisory Examples of current farmer information service solutions can still be classed as system solutions include a few different farmer information services today; this models such as large-scale government- category includes many of the largest run farmer information services. D4Ag solutions in Sub-Saharan Africa Examples of such solutions include the 80-28 in terms of the number of smallholder Farmer Hotline in Ethiopia that is managed farmers reached (i.e., registered for the by the country’s Agriculture Transformation solution). Typically, such solutions have Agency (ATA),25 ZIAMIS in Zambia,26 significantly evolved their business models Kenya’s Agriculture and Livestock Research from first-generation farmer information Organisation’s (KARLO’s) suite of farmer services. For example, they have moved applications,27 and the Smart Nkunganire toward delivering more tailored information System (SNS) in Rwanda.28 CTA CHAPTER 2 43 MNO-led or MNO-linked farmer information services represent another major sub-category. There are more than two dozen such solutions in Sub-Saharan Africa, with the most notable examples being Viamo 3-2-1 information services, deployed in partnership with various MNOs across the continent, and Orange’s D4Ag services portfolio. Each of these have farmer information services in more than 10 Sub-Saharan Africa countries.29 Examples of country-specific solutions in this category include Econet’s EcoFarmer in Zambia. Other important examples with significant scale are specialised farmer information system enterprises like iShamba in Kenya;30 iCow in Kenya, Tanzania, and Ethiopia;31 Verdant Agritech in Nigeria;32 Farmerline’s 399 Service patterns), local pest and disease trends, and Espace Géomatique, Burkina Faso in Ghana;33 SMS-based market price highly localised, granular weather data and dissemination services like RATIN,34 and related on-the-ground agroclimatic information several market information services solutions such as field temperature, precipitation, and that are linked to commodity exchange moisture levels. They can consider the specific platforms like the Ethiopia Commodity crop varietals grown on the farm (i.e., advice, Exchange (ECX).35 informed by crop models, is calibrated to the Advisory Services – specific varietals in use on the farm rather than more general crop behaviour models imported Precision Advisory from other contexts). They can also take Precision agriculture advisory services into account the demographic profile of the represent a second major emerging smallholder household (e.g., the household’s cluster of solutions under the advisory budget constraints, risk appetite, level of services use case. Precision agriculture, in farming skill, and level of literacy). Finally, the context of digital advisory services, implies though such solutions are few today, they can a move from offering generalised best practices look at the microeconomic setting of the farm to disseminating recommendations that are (e.g., geographically proximate input prices, highly tailored to individual farmers, farms, market prices, and market distances that affect and, ultimately, farm fields. What this means the farm’s economics). practicably in the African D4Ag smallholder context often remains vague.36 A sufficient quantity and quality of data must be captured in order for precision Precision advisory customises advisory services to function effectively. information selection and This first requires that individual smallholder recommendations based on a large households and farmers be profiled in detail number of factors. Precision advisory and that farm fields and field boundaries be services tend to factor in agronomic features geo-tagged. The resultant data must then be of specific farm fields (e.g., soil properties, integrated with other data derived from such water availability, shade levels, intercropping sources as the remote surveillance of farm 44 CHAPTER 2 FAO fields through drone and satellite imaging, approaches. Since the concept of precision granular weather surveillance, and hyperlocal advisory sits on a spectrum from moderately weather sensors. In addition, new types of to highly customised advice, in some cases, the portable diagnostic equipment and analytics boundary between farmer information services can be applied for in-field pest, disease, soil, and precision advisory services can be blurry. and crop nutrient testing. For the greatest This is all the more true as traditional farmer precision, in situ sensors can be deployed in information solutions, like those provided by smallholder fields and on farm machinery like MNOs, increasingly incorporate localised irrigation and tillage equipment to provide crop, weather, and pest data into advisory ongoing real-time monitoring. algorithms. Nonetheless, a few emerging models can be classified as having elements of The use of such localised data on smallholder precision advisory services. farms theoretically allows for highly tailored advice on planting, irrigation, and harvesting At the somewhat less precise end of times, the selection of the most appropriate the precision advisory spectrum are farm inputs like seeds, fertilisers, and weather/climate and pest and disease pesticides/herbicides, and forward-looking early warning surveillance and advisory farm planning that considers precise (and services. These focus on integrating localised dynamically updated) estimates of crop yields and real time weather and/or pest and disease and market conditions. Taken to the extreme data in combination with basic information of their potential and granularity, precision about the client farmers’ or pastoralists’ advisory models can also allow farmers to locations and agricultural practices. Examples optimise within their fields by informing of weather surveillance advisory solutions variable application of irrigation and other include CTA and aWhere’s CLIMARK inputs like fertilisers and pesticides for specific weather information service for pastoralists portions of a field. in northern Kenya and southern Ethiopia,37 CTA’s project with ECONET in Zimbabwe No current solutions on the ground delivering ICT-enabled weather information in Sub-Saharan Africa incorporate services, Ignitia’s Iska weather forecast all possible elements of precision services in West Africa,38 World Vision’s advisory services, but dozens of players EWEA/FIS early warning platform in Mali,39 are starting to experiment with such and Weather Impact’s weather-based CHAPTER 2 45 smallholder farming advisory products in tailored input advice like highly customised Neil Palmer, CIAT Kenya, Ethiopia, Burundi, and South Africa.40 fertiliser formulations. Examples of such solutions are Sat4Farming in Ghana,46 For pest and disease surveillance, specifically, Earth-I’s ACCORD project in East Africa,47 there are a growing number of pest-specific the Orange Garbal solution in Mali,48 solutions (e.g., Boa Me in Ghana, Rise Geodatics in Kenya,49 MUIIS in Uganda50 Africa in South Africa, and Nuru in Kenya and agribusiness-focused (B2B) players like for the fall armyworm41), as well as large-scale CropIn and SatSure which deliver precision multi-crop solutions like CABI’s Plantwise42 advisory services to smallholder value chains.51 and the Waterwatch Cooperative’s Crop Disease Alert application.43 A number of Another interesting example, though the solutions like WeatherSafe’s Coffee Crop organisation positions itself much more application in East Africa44 and AgriPredict broadly in its ultimate aspirations and in Zambia45 are focused on both weather technology focus, is Precision Agriculture and plant disease surveillance and risk for Development (PAD). PAD is a global management. NGO focused on integrating greater precision into digital smallholder advisory extensions Another category of solutions moving with the support of remote sensing data, other toward greater precision are remote data such as weather patterns and soil types, sensing (satellite) advisory services that behavioural science techniques (for solution provide advice to smallholders based design and testing), and rigorous evaluations primarily on satellite image analysis (i.e., randomised controlled trials (RCTs)) combined with in-depth farmer profiling, of resulting advisory outcomes.52 Satellite weather modelling and, at times, soil data. imagery analytics are the cornerstone of PAD’s precision advisory solutions in Africa. Such solutions frequently also provide climate adaptation and pest and disease advice but A related sub-group of precision have broader mandates than the weather advisory players are drone surveillance and pest surveillance systems covered advisory specialists; CTA and Dalberg earlier, since these solutions primarily focus are tracking over thirty such solutions in on geographically-targeted advice on crop Sub-Saharan Africa. These actors integrate and livestock management practices and/or drone imagery with other data sources to 46 CHAPTER 2 develop and disseminate customised farmer PlantVillage’s Nuru cassava disease advice. Examples of such solutions include diagnostics application,55 the Grainotheque Astral Aerial in Kenya, AgrInfo Jembe in Yiri Drotro fruit and vegetable crop disease Tanzania, Charis in Rwanda, AcquahMeyer diagnostics solution in Côte d’Ivoire,56 and Drone Tech and Ziongate Geospatial’s PEAT’s Plantix application (the most notable Airborne Agric solutions in Ghana, example globally of such solutions in terms of ThirdEye in Mozambique, and WeFly Agri both sophistication and scale).57 in Côte d’Ivoire.53 Like most drone players in Africa, these solutions tend to be of very recent More complex variants of diagnostic advisory vintage; most are in the early stages of testing solutions are models that involve agent- and developing their farmer advisory services intermediated field diagnostic or rely on new into products, as well as developing viable types of portable or farm field sensors. business models. Agent-based diagnostic models include Soil and crop diagnostic advisory services solutions like CropNuts’ Daktari Wa are another emerging cluster of precision Udongo product in Kenya, which features the advisory solutions. These rely on soil or crop collection of soil or crop samples in the field diagnostics as an entry point into the farmer by plant doctors or, alternatively, the training relationship, and typically combine soil and of farmers to self-collect and then test soil crop data with other information about the and crop samples in a professional lab before farmer and farm to generate tailored advice. developing and delivering customised advice to farmers’ phones via SMS and IVR.58 Some of these solutions do not require any specialised equipment but rely on the As alternatives to diagnostic lab processing of images taken via smartphone infrastructures, some solutions rely on applications. The background data analytics novel portable diagnostic tools. Examples enabled by machine learning across large include the Agrocares soil and crop scanner datasets of field images and ground-truth data and advisory application,59 Croptix’s mobile allow such solutions to remotely facilitate yield smartphone-compatible spectrophotometer for measurements, assess nutrient deficiencies, or plant health advisory,60 and Zenvus’s Yield diagnose pests and diseases. Examples include Sky, a portable hyperspectral camera for the Yara International ImageIt application smallholder farmers that feeds into Zenvus’s for diagnosing plant nitrogen deficiency,54 precision advisory solution.61 Some examples Ujizi Kilimo of enterprises that use field/in-situ sensors for ongoing real-time diagnostics include Ujuzi Kilimo,62 Lentera,63 and SunCulture64 in Kenya and Zenvus’ SmartFarm sensor in Nigeria.65 Field sensor-based precision advisory for smallholders is also an area of increasing experimentation by large technology companies. Examples include IBM’s EZ Farm66 and Microsoft’s Farmbeats,67 with several pilots in Sub-Saharan Africa, particularly centred on Kenya. The general trend across many of the precision agriculture advisory solutions in Africa, particularly as the costs of underlying technologies decrease, CHAPTER 2 47 is toward fully integrated precision advisory platforms. Such platforms combine in-depth farmer profiles, transaction data, weather data, satellite data, drone data, and field/machinery sensor data. Such integrated D4Ag products could ingest immense amounts of data about farmers and farm fields in order to generate highly tailored and dynamic advice regarding every element of farm operation. While still in their early stages, such next-generation integrated precision advisory solutions for smallholder farmers already exist and are being deployed by big technology players. Examples are Microsoft’s Farmbeats (and related Digital Agriculture Platform) in Kenya and the Tata Consulting Services (TCS) InteGra precision agriculture advisory platform of content. As in the case of precision Believe Nyakudjara, FAO in South Africa.68 Precision agriculture D4Ag agriculture and farmer information services, start-ups like AgrInfo/Jembe in Tanzania, it is often difficult to draw hard boundaries Zenvus and Kitovu69 in Nigeria, Lentera in between participatory solutions and other types Kenya, and CropIn, are moving in a similar of digital advisory services. Increasingly, direction, combining soil data, farmer data, solutions rely on end-user feedback and field sensors, and remote sensing data from multi-directional data flow rather than satellites and drones. taking more rigid, top-down architectures to information dissemination. Advisory Services – For instance, many digital advisory Participatory and solutions over the years have integrated Peer-to-Peer inbound and outbound call centres and Participatory and peer-to-peer advisory IVR models to source queries from farmers solutions are another important sub- and deliver tailored advice in local languages. type case of digital advisory services. This form of interactivity can be considered Participatory solutions feature tight feedback part of the participatory advisory sub-use case, loops between content providers and end-users, though it also overlaps with other advisory greater levels of farmer interactivity with the models mentioned above. solution (i.e., not just one-way information flows from experts to farmers), and, in many Multiple D4Ag solutions feature call cases, a role – direct or indirect – for farmers centre models that ensure a high degree in creating or customising advisory content. of interactivity. This interactivity manifests Peer-to-peer advisory solutions share some of both in the nature of call centre engagement these features, but also put individual farmers with farmer clients and in the adjustment of and farmer experts into more central roles for content based on rigorous data capture and content creation and dissemination. analyses of incoming queries. Examples include a number of current advisory solutions, such Broadly speaking, digital advisory as iShamba in Kenya, the 80-28 Hotline solutions are moving toward greater service in Ethiopia, and Farm Radio’s interactivity, localisation, and adaptation Mlimi Hotline in Malawi.70 48 CHAPTER 2 dynamically mine farmer queries to improve content relevance and delivery, and the ability to tap into large volumes of farmer-generated content (e.g., logs of prior conversations) to enrich the breadth, depth, and localisation of the advice being delivered. A couple of noteworthy examples include Arifu, a large digital learning and advisory service that works with African farmers via SMS and chatbot applications73 and the chatbot-based advisory platform in Kenya, Farm.ink (and its associated Africa Farmer’s Club Facebook community).74 Another interesting solution in this category is Mahindra & Mahindra’s MyAgriGuru voice chatbot for smallholder farmers. Though this solution is currently limited to India, it CTA Likewise, the use of IVR tools – either in is reaching substantial scale and will likely be combination with call centres and SMS replicated in some way for Africa’s farmers.75 channels or via stand-alone channels – is now mainstream for digital advisory D4Ag advisory solutions do not solutions in Africa.71 There are too many merely exploit new models for models that integrate IVR to mention, but farmer interaction but also, in some it is worth highlighting the work of IVR cases, integrate farmer-generated technology pioneers in the agriculture advisory or intermediated content. The most space like Awaaz.De, VotoMobile (now part established example of such a peer-to-peer of Viamo 3-2-1), and EngageSpark who advisory model is Digital Green, one of the offer B2B IVR-integration services to D4Ag veteran enterprises of the D4Ag sector, which enterprises.72 In addition, a couple of the for over a decade has deployed its farmer video most notable large-scale IVR-based advisory model on a large scale first in India and now solutions are the Ethiopia 80-28 Hotline and also in several countries in Africa. This solution Viamo’s network of IVR-based 3-2-1 Farmer features (i) a participatory process for content information services. production (i.e., topic selection and content adaptation informed by farmer feedback); Newly arrived in the interactive (ii) locally generated digital videos filmed by advisory model space are chatbots for specially trained community film-makers and, D4Ag service delivery. These will become even more critically, featuring local farmers increasingly common over the next few years. who demonstrate and promote improved A growing number of solutions are integrating agricultural practices in local languages; (iii) machine-learning/AI-enabled chatbots, a trend human intermediated instruction of farmer that all experts consulted for this report expect groups for content dissemination and training to accelerate in the next few years. Chatbots (i.e., a private company, NGO, or government are programmes designed to simulate natural extension agent shows videos to farmers and conversations with human users – in this case, facilitates discussions); and (iv) intensive and with farmers – either via text or voice-based systematic data capture and analysis of farmers’ applications. These models offer multiple feedback about the solution content and their theoretical advantages including greater farmer resulting behaviour changes. The Digital Green engagement with the content, the ability to model has been studied closely over the years, CHAPTER 2 49 but the crux of the approach, relevant for this Smallholder farmers are also discussion, is the participatory nature of the increasingly using major social solution both in terms of the content itself and networking platforms to communicate the process of farmer engagement, behaviour agricultural information. It is important to change, and practice adoption.76 Digital highlight, with respect to peer-to-peer D4Ag Green’s model involves farmers in content solutions, that in those geographies where development, but the content is also carefully there is sufficiently strong connectivity, the curated, screened, and triangulated with input top social media platforms in Sub-Saharan from professional agronomists. Africa – i.e., Facebook, Facebook Messenger and, to a greater degree, WhatsApp – are Other peer-to-peer advisory models link becoming increasingly important farmer-to- farmers with each other directly, so that farmer information sharing vehicles.79 This one farmer’s questions are answered by phenomenon is still marginal in many places, another. This approach creates tremendous but as the adoption of mainstream social opportunities for on-the-ground data collection media and communication platforms like and for impacts on farmer behaviour (i.e., WhatsApp increases and as such platforms farmers engaging more with content that is widen their functionality (e.g., WhatsApp’s validated and shared by their peers). But, like widely anticipated move into payments), the any social networking solution with limited potential for such networks to become major curation, it simultaneously presents significant channels for advisory and other D4Ag service risks that low-quality or inaccurate agricultural delivery will grow. In Kenya, for example, a advice and information can be collected and country where the level of WhatsApp adoption distributed based on crowdsourced perspectives is already very high by African and even or direct farmer-to-farmer advice. global standards,80 an expansive, late-2018 smallholder survey showed that WhatsApp A few different solutions exemplify the was already used for farming by half as many peer-to-peer approach. Africa Farmer’s farmers as those who used farming apps.81 Club/Farm.ink, already noted above in the context of the Farm.ink chatbot, for instance, Advisory Services – Farm relies on a Facebook farmer community that Management Software generates farm queries and content that the Farm management software solutions chatbot can mine and pair with professionally for smallholder farmers feature curated agronomic content. interactive tools/applications for farmers CTA Wefarm, the largest-scale peer-to-peer farmer social network in Africa, takes a different approach.77 Wefarm users can ask and answer farming questions and share farming tips, via SMS or online, enabling farmers in rural areas without internet access to participate. N-Frnds78 gives farmers access to professionally curated advisory content on its platform via feature phones (USSD), allows for interaction and communication between business owners, suppliers, and farmers, and includes highly popular features that allow farmers who lack mobile data to engage in group and one-on- one chats to share farming advice. 50 CHAPTER 2 or agents interfacing with the farmers examples include SmartCow82 and that go beyond the delivery of tailored DigiCow83 in Kenya for dairy cows, and recommendations to specific farms. AkokoTakra84 in Ghana and Sen Ngunu85 They empower farmers to make their own in Senegal for poultry. For smallholder decisions with tools to (i) farm budgeting and horticulture and staple crop farming, examples planning (e.g., pro forma upside implications of such self-service management solutions and risks of specific farm investments based include African start-up D4Ag enterprises like on market conditions and/or historical farm Probity Farms86 in Nigeria, AgriGo87 in performance); (ii) farm monitoring (e.g., Rwanda and BudgetMknoni88 in Kenya, as dynamic yield and economic projections); well as international farm management (iii) financial management, accounting, and solutions like Agrivi,89 which can be utilised by record-keeping; (iv) supply chain management African smallholders and are being marketed in the case of slightly bigger or more complex through local partners in several countries, smallholder farm operations; and potentially such as Kenya and Nigeria. even (v) reporting tools that can pave the way to formal financing. The reach of most of these solutions is still very limited given how new they are While there are many sophisticated farm and given the broader challenges noted above management software solutions for for smallholder uptake of more sophisticated large-acreage farms in the developed self-service software. Our interviews suggest, world, the segment of D4Ag services for however, that the uptake and abundance of smallholders is understandably very such solutions will grow quickly in specific nascent. African smallholder farmers face niches such as dairy. Even in the area of staple significant literacy and digital literacy crops and horticulture, while complex farm constraints that curb the potential reach of management tools will likely see low uptake in highly interactive farm management software. near term, novel D4Ag farm budgeting and Furthermore, access to mobile data and/or recordkeeping features could become far more sufficiently sophisticated devices like mainstream as smartphone adoption increases. smartphones, tablets, and laptops is limited. At the same time, particularly in the context of livestock and dairy, interesting solutions are Market Linkage Use Case starting to emerge. Some recently launched Most African smallholder farmers Vystekimages are not adequately linked to markets for a variety of reasons. These include information gaps and asymmetries about market needs, buyers, and prices; remoteness (and related challenges of logistics and transportation costs); overly low and geographically fragmented production volumes to interest bigger buyers; poor quality of produce relative to market requirements or difficulty in meeting the high hurdles of food safety standards and traceability required by agribusiness buyers and processors in more commercial value chains; and, critically, low farmgate prices due to highly intermediated value chains with multiple layers of actors between farmers and end-consumers. On the input market side, beyond challenges of CHAPTER 2 51 MARKET LINKAGE Figure 7 Overview of D4Ag market linkage models Digitally-enabled value chain integrators • Input integrators • Market access (off-take) integrators • End-to-end integrators Mechanisation access services E-commerce services • Pay-as-you-go agriculture machinery • Agri-input e-commerce Market linkage • Food e-commerce • Shared Services for Mechanisation E-marketplaces • Input e-marketplaces • Off-take e-marketplaces 52 CHAPTER 2 financing access, smallholder farmers also agribusinesses of all types improve their have difficulty finding and purchasing reliable margins and grow markets.91 and appropriate farm inputs due to some of the same factors – including information In the past several years there has been asymmetries about which products are significant growth in the number of digital appropriate and have sufficient quality, market linkage solutions available, as well as diseconomies of scale and a related lack of the scale of those solutions. As with many buying power, underdeveloped and fragmented other D4Ag use cases, however, given the agro-dealer networks that increase input nascency of the sector and rapid evolution in costs but still offer very limited availability terms of technologies and business model, the and convenience for input purchases at the definition of ‘digital market linkage’ remains last mile, and other logistics and distribution amorphous. The term is often applied loosely challenges that are common in rural Africa.90 to describe an ever-multiplying array of business models.92 Digitally-enabled market linkage solutions thus have a critically The crux of the concept is the use of digital important role to play in connecting tools to facilitate market connections, which smallholder farmers to high-quality ultimately lead to transactions for goods farm inputs, to production and post- or services between different smallholder harvest machinery and mechanisation value chain actors including farmers; farm services, and – ultimately – to off-take aggregators such as cooperatives, agri-input markets, including agro-dealers, wholesalers, producers and input distribution intermediaries; retailers, or even directly to the urban or farmer services providers (e.g., veterinarians, international end-customer. Digital market agronomists, mechanisation services providers, linkage solutions, by introducing efficiency, financial institutions); produce buyers, traders, transparency, accountability, and trust into and processors; and – moving toward the otherwise inefficient and opaque value chains, ultimate end-consumer – international allow smallholder farmers to lower their costs exporters, domestic wholesalers and retailers of of production via access to lower-cost and finished food products. higher-quality inputs, reduce the costs and risks of finding and transacting with buyers At the most basic level, digital market and ultimately increase their yields and farm linkage solutions can be segmented by both incomes – while at the same time helping Olivier Thuillier, FAO their value chain role and by their level IMAGE REQUIRED CHAPTER 2 53 of human intermediation. The former value chain integrators, mechanisation Fintrac Inc, USAID considers their input market linkage, off- access services, agri-input and food take market linkage or end-to-end market e-commerce services, and virtual buyer- linkage. The latter ranges from purely digital seller e-marketplaces (Figure 7). solutions like virtual agriculture commodity e-marketplaces and trading applications to Market Linkage – digital tools and platforms that function only Digitally-Enabled in combination with last-mile human agents working either for the D4Ag enterprise Value Chain Integrators or for agribusiness organisations that are Digitally-enabled value chain integrators themselves agriculture value chain participants are D4Ag solutions that use digital as aggregators, buyers or sellers. Another tools combined with either in-house important consideration is the breadth of or third-party human agents to link the overall value proposition, i.e., market agricultural markets. At the core of these linkage only versus models that combine models is the ambition to capture value and market linkages with advisory services, supply generate impact for both smallholder farmers chain management, and finance. There are, and agribusinesses by formalising currently of course, a myriad of other nuances that fragmented and informal value chains. Value differentiate digital market linkage business chain aggregation and formalisation can, of models – such as revenue models and course, be accomplished via non-digital means, contracting arrangements – which this report but the key insight of digitally-enabled value does not explore. chain integrator solutions is that digital tools are a powerful means of improving trust, Across these dimensions, four major reducing costs, accelerating time to market clusters of digital market linkage (a critical consideration for seasonal and models stand out in the Sub-Saharan highly time-sensitive agricultural input and African market today: digitally-enabled off-take markets), facilitating transparency 54 CHAPTER 2 in time that the value is generated (e.g., via commissions, revenue shares, or brokerage fees) in comparison to the much more indirect revenue models of most other D4Ag use cases. Major variants of the digitally-enabled value chain integrator model include (i) input market integrators; (ii) off-take market integrators; and (iii) end-to-end value chain integrators. There are already a few dozen solutions in this category today in Africa; the number is growing rapidly as new D4Ag market linkage start-ups enter the market and as traditional smallholder value chain integration actors – FAO and accountability, and ultimately growing such as small/medium-sized agribusinesses, the reach, social impact, and profitability of big regional or international agribusinesses, traditional value chain linkage models. and specialist market linkage NGOs and social enterprises such as One Acre Fund and Largely anecdotal evidence suggests that these Babban Gona – integrate digital tools into market linkage solutions generate tangible their human agent models in order to reduce benefits such as greater ease in identifying and costs, improve profitability, and strengthen attracting farmers (i.e., lower acquisition costs), their competitive positioning vis-à-vis new significantly reduced agent-to-farmer ratios digital disruptors. (i.e., field force efficiencies), lower requirements for agricultural agent skills due to digital While there are a few important exceptions, monitoring and information access (i.e., lower many of these players have relatively limited agent recruiting and training costs), improved reach today. The main reason for this, beyond trust for all parties (i.e., greater stickiness of the newness of these models, is the resource farmers and other value chain actors to the intensity of these solutions and, in the case of solution), reduced value leakage in operations off-take market linkages, the need to develop due to digital tracking (i.e., less agricultural market demand concurrently with quality product spoilage and loss and reduced product supply – something that requires time. input theft), and – critically – benefits from Nonetheless, solutions of this type are growing, economies of aggregation and scale for value can break even quickly and, as suggested in capture, whether in terms of lower input costs our interviews with sector experts, are likely or higher produce prices.93 to see growing attention from investors in the next few years. The unique advantage of digital market linkages in general, and the digital value chain For digitally-enabled input market integrator solution sub-type in particular, integrators, digital technology primarily is that in these models the D4Ag solution serves as a communication and transaction provider is an integral value chain player. By channel by which smallholder farmers and taking on bigger risks and making substantial input providers (e.g., seed, fertiliser, pesticides/ investments in value chain formalisation, herbicides producers, large distributors, the solution provider is theoretically able to and last mile agro-dealers) coordinate on take a much bigger share of the value that the quantity and type of inputs needed, is ultimately being generated at the point aggregate farmer input demand to improve CHAPTER 2 55 the economics of input distribution, and mobile input loan repayments, among other optimise logistics (e.g., input delivery route digitalisation initiatives.99 If it continues to planning). Notable examples of African start-up follow this trajectory, One Acre Fund will, enterprises that fall into this category include in effect, become a digitally-enabled input Farmers Pride in Kenya,94 CowTribe in market linkage platform and may be able Ghana,95 myAgro in Mali and Senegal,96 to convert these digitalisation investments and Agrics97 and iProcure98 in Kenya and into much greater impact and scale. Other Tanzania. models comparable to One Acre Fund, such as Babban Gona in Nigeria, are likewise While most of these digitally-enabled input investing heavily into digitalising elements of market integration solutions are relatively their input supply chain linkages approach. small scale, some have significant reach or significant potential for near-term reach. From the perspective of more commercial For instance, One Acre Fund, the non- models with potential for scale, Safaricom’s profit social enterprise that had more than DigiFarm is currently primarily focused on 800,000 farmer clients in 2018 for its bundled using a combination of digital technologies and input and financing approach, is already its physical network of partner organisation field the largest-scale implementer in Africa agents to link Kenyan farmers to agricultural of agent-intermediated smallholder input inputs, along with input financing, and market linkages for non-commercial (loose) increasingly more tailored advisory services.100 smallholder farmer value chains. One Acre While the organisation and partners like Fund’s model has historically featured little MercyCorp’s AgriFin Accelerate programme technology, but over the past few years the have a broader ultimate vision for DigiFarm, organisation has started to invest aggressively the solution is today a classic example of a in integrating into its market linkage work digitally-enabled input value chain integration a range of digital approaches and tools, model with potential for scale.101 from digital farmer registrations to digital agent field force management tools, digitally- For digitally-enabled off-take market enabled monitoring and evaluation and integrators, digital tools are likewise used FAO 56 CHAPTER 2 horticulture off-take market linkage solution currently being scaled as a for-profit enterprise in India.104 A rising number of solutions use digital technology and human agents to link both sides of the market, from farm input provision through off-take – a model we label as integrated end-to-end market linkages. In the start-up space, one of the most ambitious early-stage solutions of this type is Tulaa in Kenya, an innovative end-to-end market linkage enterprise that is seeing growing commercial investor interest and – based on early independent assessments of its market pilot in 2019 – is already generating significant per-farmer revenue and high levels of impact on smallholder yields and incomes.105 Another Zoona to support the efforts of human field agents. interesting start-up example is Akorion’s The primary role of digitalisation for such EzyAgric solution in Uganda, which solutions is to reduce the transaction costs of combines digitally-supported input and off- aggregating high-quality produce from highly take market linkages with a network of youth fragmented smallholder value chains, thereby service provider village agents equipped with generating cost savings for agribusiness and smartphones, each of whom serves 150–200 incremental value to farmers due to greater farmers as a facilitator of input and off-take certainty (or, via contracts, absolute guarantees) transactions.106 of market access, the reduction in the number of intermediaries between the farmer and At a greater scale, the work of the Farm to the buyer, and farmers’ improved bargaining Market Alliance (FtMA) is also a variant of position vis-à-vis buyers. this end-to-end approach, combining human- intermediated input and off-take linkages (and Leading digital off-take market integration physical market aggregation infrastructure) with solutions in Africa include Twiga Foods the increasingly rich feature set of FtMA’s in Kenya, the best-known enterprise in this in-house digital platform.107 category given its tremendous fund-raising success in the recent years.102 In the D4Ag Another entry point for digitally-enabled start-up space, important examples of digitally end-to-end market linkages in more enabled off-take market integration solutions commercial value chains are digital smallholder include a few Kenya-based players such as financing programmes of the type being Selina Wamucii, Farmshine, and Taimba, pursued by the Kenya Commercial Bank as well as start-ups elsewhere on the continent (KCB) via its Mobigrow product in East like Trade in Ghana.103 Africa and by Opportunity International via its holistic smallholder value chain Finally, another example that has garnered financing model in Ghana and Côte d’Ivoire.108 a great deal of attention globally and is At the core of both products is an approach likely to be replicated soon in Africa is that involves working with an integrated Digital Green’s LOOP, a digitally-enabled ecosystem of farmers, buyers, and agri-input CHAPTER 2 57 providers that are linked not just via digitalised For e-commerce off-take models financing flows (e.g., input payments to agri- like digital grocers, customers of the input providers when the farmer is approved e-commerce businesses should ideally have for a loan, digital payments for produce to good connectivity, digital payments accounts, farmers) but also with improved data insights a smartphone/tablet/PC to access the online and non-financial value-added services such shop, and interest in purchasing (and possibly as advisory services for farmers, agri-input paying a premium for) fresh, high-quality, screening, and market facilitation. locally sourced food. Some workarounds to these constraints exist – such as SMS Market Linkage – ordering instead of an online storefront or Agri-Input and Food cash-on-delivery models in the place of digital payments. In effect, however, this model E-Commerce Services translates into a more niche urban middle- Agriculture e-commerce services are class market for online grocers and thus, online retailers of agricultural produce by extension, has more limited potential as for urban consumers or agricultural an e-commerce market linkage model that inputs for smallholder farmers; they attempts to formalise smallholder value chains rely on online order fulfilment via and link them directly to urban consumers. either shipping or a combination of These constraints, and the challenges of some online and offline (i.e., brick and mortar African e-commerce retailers like Jumia in store) footprints. recent years, have led some sceptics to question the scalability of food e-commerce models and Like digitally-enabled value chain integrator their potential for farmer impact in Africa.109 models, agricultural e-commerce services require a significant amount of value Even if the market is ‘niche’, however, this chain investment and intermediation could still be a highly attractive model for from the D4Ag enterprise in order to source D4Ag enterprises as niche does not necessarily high-quality product, provide additional meet small. The middle class in Africa is value-added activities (e.g., sorting, cold chain already several hundred million strong, and services, and packaging for livestock and fresh this middle class is growing quickly with rising produce, or quality assurance for agri-inputs), GDP and urbanisation.110 By 2030, 47-50% of and then manage the payments and the Africans will live in cities, up significantly from logistics of getting the product to the end-user ~40% today, and for every 1% increase in at the right place and time. urbanisation there is generally a 5% increase in food sales.111 There are thus millions or CTA On the plus side, e-commerce models have a greater potential to bypass intermediaries and can thus theoretically generate more value for both the D4Ag provider and the farmer than other market linkage models, since such models extend the link directly to the product’s end-user as opposed to linking up with intermediary wholesalers or retailers who take their slice of the value. On the negative side, unlike value chain integration models for non-commercial smallholders, digital retail storefronts tend to require wealthier, more sophisticated, and thus smaller customer bases. 58 CHAPTER 2 tens of millions of people in the urban areas wealthier farmers or selling inputs for more of most African countries with discretionary commercial value chains (e.g., livestock or income, but few high-quality retail food stores fisheries products). per capita. At the same time, it is clear that D4Ag With these trends in mind, investment into entrepreneurs are finding ways of mitigating food e-commerce businesses in Africa that some of these challenges by using SMS/ link farmers directly to end-consumers is call centres to handle order-taking from growing. How precisely these D4Ag players connectivity-constrained farmers, offering free interact with the farmer varies by model, so advice on agri-input selection and use to deal the impact on farmers is not always clear, with issues of knowledge and trust, and finding but a number of examples are emerging that local partners capable of facilitating last-mile show that such market linkage models can delivery logistics. The number of such D4Ag be viable and attractive to both farmers and enterprises appears to be smaller than that e-commerce entrepreneurs.112 Examples of such of food e-commerce stores; examples include direct-to-consumer local produce e-commerce Afrimash in Nigeria, FarmIT in Kenya, and enterprises include IzyShop in Mozambique, eMsika in Zambia.114 FarmFresh in Gambia, HMart and Get It Rwanda in Rwanda, Jangolo in Cameroon, Premium Hortus and Jinukun in Benin, Market Linkage – Farmart in Ghana, Village Market and Agriculture E-Marketplaces Foodstock Farmers Market in Nigeria, Agriculture e-marketplaces are D4Ag Khula in South Africa and Herdy Fresh and market linkage solutions that require Kitchen Soko in Kenya.113 little or no human intermediation, and that bring individual buyers and On the input retail side, agriculture sellers together via virtual trading input e-commerce enterprises serving marketplaces.115 smallholder farmers also have substantial constraints on market size, including poor rural Agriculture e-marketplaces provide a connectivity, limited farmer digital literacy, platform for various sellers and buyers of and the high costs of rural transport and agricultural products to transact. For off-take shipping logistics. In effect, these constraints e-marketplaces, sellers can include individual lead to a parallel situation in which digital-only farmers, farmer groups, or cooperatives posting e-commerce sites are often limited to serving their offers. Buyers range from small agri- CTA CHAPTER 2 59 Fintrac Inc, USAID dealer buyers and aggregators to substantial technology-based innovations like the use of agri-processors and wholesalers to last-mile blockchains to build trust via transparent and food retailers. For the input e-marketplace immutable transaction records (e.g., Cellulant’s variant, sellers include various types of input Agrikore). supply chain intermediaries while smallholder farmers typically are the buyers. When e-marketplace platforms succeed, whether on the input or off-take market E-marketplaces can help solve the linkage side, they can theoretically problem of inefficient and fragmented unlock substantial value through agricultural markets when and if they efficiency gains and other positive are able to crack the challenges of knock-on effects. These effects include the identifying and attracting enough buyers ability to use transaction information at scale and sellers. To do so, e-marketplaces need to deliver value-added advisory or market to invest into effective marketing and – more agri-intelligence services across smallholder importantly – must embrace innovations value chains – or to convert those transaction that build trust that is often missing in data into records that value chain participants smallholder farmer value chain relationships. can use as a form of collateral for working The trust-building mechanism can simply be capital or for smallholder farmer input loans. the reputation or brand of the e-marketplace backer (e.g., MasterCard Farmer’s Network), The number of e-marketplace D4Ag a reliable payments platform with which the solutions in Africa is growing – our marketplace is associated (e.g., Cellulant’s database is now tracking more than Agrikore), partnerships with credible 15 such players. The majority tend to government agencies or NGOs (e.g., Farm- be at very small pilot scales today (<25,000 to-Market Alliance), value-added services smallholder farmers registered); a handful, such as free advice, explicit insurance or however, are starting to reach much greater guarantee mechanisms to mitigate the risk of scale and aspire to reach millions of farmers non-performance by counterparties and, lastly, across Africa. 60 CHAPTER 2 Doreen Hove, USAID Of the various examples of e-marketplaces that A few marketplace players are focusing aim to link farmers to agricultural produce on both the input and off-take linkage buyers, MasterCard’s Farmers Network pathways – Lima Links in Zambia and (formerly known as 2Kuze) is likely the most Farmerline serve both produce and input ambitious e-marketplace in Africa today. marketplaces.120 Cellulant’s new Agrikore Incubated by MasterCard’s Lab for Financial solution also focuses on both input and Inclusion in Nairobi, and currently deployed produce e-marketplaces via a blockchain-based in Kenya, Uganda, and Tanzania, this solution smart-contracting, payments and marketplace aims to systemically integrate smallholder system that seeks to ensure that everyone farmers from loose value chains with quality in agriculture (farmers, FMCGs, agriculture buyers via a digital transaction marketplace inputs providers, produce aggregators, for individual sellers and buyers. Participation insurance companies, financial institutions, in the network involves all actors adopting governments, development partners) MasterCard-led payments digitalisation.116 can do business with each other in a trusted environment.121 Smaller start-up examples of e-marketplaces that link farmers to buyers include Usomi’s Across all of these solutions, the interaction Rubi and Mifugotrade in Kenya, Farmster between the buyers and sellers can be simply in Tanzania, Annimart, Zowasel in Nigeria, memorialised as a record in the e-marketplace and eFarm in Cameroon.117 TruTrade in or can incorporate the processing of payments Kenya and AgroCenta’s AgroTrade in for the transaction on those e-marketplaces Ghana also likely fall into this category, though that have third-party payment partners or they do feature village-level entrepreneur proprietary payment solutions such, for agents as part of their models, and so are example, MasterCard’s Farmers Network, not purely digital marketplaces.118 On the Cellulant’s Tingg payments mechanism input marketplace side, examples of active in the case of Cellulant’s Agrikore e-marketplace start-ups include FarmAll in e-marketplace, or the use of Agrocenta’s Kenya and Agro Market Day in Uganda.119 AgriPay for their Agrotrade e-marketplace. CHAPTER 2 61 Market Linkage – While the topic of barriers to mechanisation Mechanisation Access is a complex one with many policy and market failure dimensions, it is becoming D4Ag mechanisation access solutions clear to many sector experts that innovative use digital tools and channels to link D4Ag solutions, in particular, hold the smallholders to farm machinery or potential to address several of the major farm mechanisation services while constraints to mechanisation uptake.126 disrupting or leapfrogging the Some of the key barriers that D4Ag solutions affordability, availability, and logistics can address include high capital costs of constraints of traditional smallholder mechanisation technologies relative to the farmer agriculture mechanisation income levels of most African smallholder business models. farmers, the absence of affordable financing for mechanisation, challenges of supply-demand Farm mechanisation has been the pivot to matching in fragmented value chains with the agricultural revolution in many parts poor information access, the scarcity or of the world and has contributed greatly to absence of distribution infrastructure, and the increased output of food crops. In the issues of equipment quality assurance and African smallholder context, mechanisation ongoing maintenance in remote rural areas.127 – particularly the greater uptake of irrigation and tractors during the crop production and Our review of D4Ag market trends and harvesting cycles, as well as the integration of sector interviews suggest that the two most cold chains and mechanised processing post- immediately promising D4Ag solution areas harvest – has the potential to dramatically in this regard are shared economy for improve yields, generate new, higher-quality mechanisation and pay-as-you-go (PAYG) employment opportunities and income streams, mechanisation solutions. While the number increase resource-use efficiency, and mitigate of start-up enterprises focused on either climate-related hazards.122 opportunity is still relatively small – perhaps a dozen out of the nearly four hundred D4Ag The reality of the mechanisation status quo solutions tracked – it is rising quickly with in Africa is, however, a challenging one. multiple new entrants in just the past two years, Siegfried Modola, IFAD While tractors are used to prepare land on and growing inflows of venture financing. over 60% of cultivated lands in Asia, the corresponding figure for Sub-Saharan Africa The first of these opportunity areas is currently around 5%.123 Likewise, only is the use of ‘Uber-ised’ shared 3.5–5% of the area cultivated in Sub-Saharan economy solutions to link farmers to Africa is currently equipped for irrigation, mechanisation providers and services. by far the lowest of any region globally.124 The most prominent examples of this model Unsurprisingly, there is a growing consensus in Sub-Saharan Africa are the use of digital on the acute need to prioritise smallholder shared service solutions to link farmers to farming mechanisation in order to achieve tractor services,128 though the model is also Africa’s ambitious agricultural transformation readily extendable to other mechanisation goals. The issue has gained significant services that require capital intensive yet momentum in the past year as reflected by the mobile agricultural machinery such as high- African Union Commission’s launch in 2018 cost field diagnostic equipment (e.g., soil of the Sustainable Agricultural Mechanisation and crop testing scanners from enterprises Framework for Africa and the concurrent like AgroCares), land-levelling equipment strong call from the Malabo Montpellier Panel (e.g., precision laser land-levellers from for increased investment in smallholder farmer companies like Trimble that are suited to agricultural mechanisation.125 African smallholder settings),129 and portable 62 CHAPTER 2 mechanised systems for the variable-rate households in the region. A large share application of fertilisers, pesticides, and (60–80%) of the clients of these off-grid solar herbicides (e.g., fertiliser sprayers). PAYG companies are either smallholder farmers or peri-urban and rural Africans The best-established example today is who have at least partial revenue streams Lagos-based Hello Tractor which now has from agriculture. operations across multiple Sub-Saharan African countries and is picking up investors, as well Within this broader PAYG space, SunCulture as technology and distribution partners like in Kenya is the best-established player at the IBM and John Deere International.130 agriculture-energy nexus of PAYG agricultural Other African start-ups with shared economy equipment services. The company currently mechanised equipment rental models include focuses on deploying a PAYG solar irrigation TroTro Tractor in Ghana, E-Tinga and pump, but also delivers value-added advisory FarmAll in Kenya, and Kobiri in Guinea.131 services to its client farmers (i.e., weather Another notable arrival in Africa is Mahindra advisories and tailored advice on when and & Mahindra’s Trringo solution, which how much to irrigate) and has a vision of recently launched operations in Tanzania. ultimately integrating many other types of Trringo already has several years of track agricultural equipment into its platform such as record in five Indian states with 1.5 million post-harvest processing equipment and as cold farmers registered for mechanisation services to storage equipment for dairy and horticulture.135 date – a clear indicator of the potential for the scalability of such solutions in Africa.132 Other examples of PAYG agriculture equipment players in Africa include Azuri’s PAYG agricultural machinery GrowFast and Simusolar for solar irrigation, distribution is another highly promising AgSol for PAYG processing and milling, D4Ag mechanisation model that takes and ColdHubs for PAYG cold chains.136 At advantage of digital payment ecosystems least a half-dozen new Africa PAYG entrants and IoT technology to allow farmers are expected across these different models in to pay for mechanisation equipment in the next 6–18 months, so this segment of the small increments while they use it on market warrants close monitoring for those their farms. D4Ag investors interested in the agriculture- energy nexus.137 As in the case of shared economy enterprises, the potential for PAYG models for mechanisation is far broader than the current Supply Chain implementation of such solutions in Africa, Management Use Case which today tends to focus on deployments Supply chain management solutions of solar powered irrigation equipment. These are primarily designed for and solutions have grown out of a broader solar marketed to agribusiness to make it off-grid energy PAYG sector that historically more convenient, safe, efficient, and focused on household lighting and home profitable for agribusiness to interact entertainment (i.e., TV) products, and features with smallholder farmers. The primary such players as M-KOPA, Zola Electric, focus of solutions in this use case is to help Fenix International, BBOX, and PEG.133 agribusinesses manage their relationships with PAYG solutions reached roughly 2 million those smallholder farmers who are already Sub-Saharan African households across two linked to them via formal off-take or less dozen countries by early 2018,134 and – based formal input purchasing relationships – or to on conservative growth rate estimates – are help them integrate new farmers into their today likely used by more than 3 million value chains. Using supply chain management CHAPTER 2 63 SUPPLY CHAIN MANAGEMENT Figure 8 Supply chain management – overview of sub-use cases and solution examples Digitally-enabled value chain integrators Logistics Quality assurance/ anti-counterfeiting Supply chain management Supply chain ERP solutions • Specialist supply chain ERP enterprises • Big tech agribusiness ERP solutions • Proprietary/in-house agribusiness ERP 64 CHAPTER 2 CTA solutions need not mean becoming a paying horticulture, and cotton.138 For other supply client of a third-party D4Ag provider. It can chain management solutions, the evidence is also include allocating resources to build and still at an early stage but is already sufficiently deploy digital tools in-house. compelling for agribusinesses to invest in integrating these kinds of tools into their We define ’agribusiness’ broadly for the work at significant scale. While this is not purpose of this use case. On the off-take evidence of impact, per se, the growing market side, agribusiness users of supply chain interest and investment in supply chain management solutions can range from large, management on the part of agribusiness global Africa-focused buyers and processors attests to its value.139 – such as ETG, Olam, Mars, Cargill and Barry Callebaut – to national and regional The specific benefits of supply chain African agro-processors – such as the Dangote management solutions depend on the Group in Nigeria and NWK Agri-Services client type. Off-take agribusiness actors are in Zambia – to various types of smaller the primary ‘client’ and ‘user’ of most supply downstream farmer aggregators with outgrower chain management D4Ag solutions. For schemes, such as smallholder cooperatives and such players, the theoretical benefits of these nucleus farms. On the agri-input side of the solutions include lower transaction costs of value chain, business users of supply chain attracting and maintaining smallholder farmer management solutions range from global or relationships, significant cost-efficiencies for regional agri-input players, such as Syngenta, many other types of operations (e.g., agent field Yara and OCP, to small and mid-sized force management, sustainability certification, national agri-input companies to other more transport logistics), improved transparency downstream input value chain intermediaries into and traceability of value chain data, such as input wholesalers and agro-dealers. greater accountability of contracted farmers and agribusiness field agents, better quality The business and impact case for supply of product sourced, reduced post-harvest loss chain management solutions is growing. and waste and, ultimately, greater profitability In many cases, based on self-reported impact and scale. Input agribusinesses also use some data, D4Ag supply chain management forms of supply chain management solutions solutions are already increasing transparency, to establish more direct relationships with efficiency, and operational profitability, their smallholder clients and to better monitor particularly for well-established tight (i.e., and manage the performance and quality of commercial and structured) smallholder value (typically independent and highly fragmented) chains such as tea, cocoa, coffee, high-value agri-input value chain intermediaries. CHAPTER 2 65 For input agribusinesses who utilise supply certification solutions. These solutions, chain management solutions, benefits should also known as ‘tracking and traceability’ or also ultimately translate into improved ‘track and trace’, are digitally-enabled tools profitability due to cost-savings per farmer that link data about specific farms and farmers reached and reduced input counterfeiting, to a view of how food commodities flow as well as stronger and more direct through value chains. relationships with smallholder farmers and other intermediaries that promote input These tools enable agribusinesses to have demand and thus revenue growth. full visibility into the agricultural last mile, maintain a digital record of farmers and other At the individual farmer level, while downstream supplier intermediaries, and smallholders are not the direct clients of facilitate auditing for certification requirements, supply chain management solutions, they are which can become hugely time consuming and often beneficiaries of activities that better expensive in the absence of a strong digital integrate them into formal value chains and data trail. The focus on certification explains should therefore see the eventual benefits of why, historically, most digital traceability higher yields and incomes through value solutions on the African market have focused chain integration. on smallholder products for export markets.141 Supply Chain African domestic agribusinesses have Management – had less of demand for such tools due to fewer standards, low enforcement, Traceability and or low consumer demand for certified Certification Solutions products, but this is now starting to Traceability and certification solutions change142 due to a rising middle class in some help agribusinesses onboard farmers, African countries and, more importantly, document farm compliance with growing recognition by the African agribusiness standards, and trace produce across community that traceability tools can create value chains with higher fidelity and broader value – for example, by helping lower costs. agribusiness better manage instances of food- borne illness and food recalls by making it CTA The demand for traceable and certified agricultural products is on the rise in global markets as international consumers demand more transparency and accountability in supply chains.140 The growing popularity of concepts such as ‘farm-to-fork’ and increased focus on compliance with environmental and social commitment standards and codes of conduct (e.g., regarding labour practices, human rights, and issues such as deforestation and water use) highlights the importance of full visibility into food chains for consumers as well as producers. To comply with an increasing number of both mandatory and voluntary standards and certification schemes, agribusinesses that procure crops from African farmers are increasingly adopting traceability and 66 CHAPTER 2 possible to trace the issue to the source and Sourcing Management platform two years target costly recalls only to impacted supply ago and now works with large global buyers chain actors.143 and processors such as Barry Callebaut, reaching over 225,000 farmers across Africa.145 While agribusiness is the ultimate beneficiary of such tools, smallholders also benefit because Another large new digital traceability and these tools help them access new markets with certification platform is managed directly by a higher prices and, on the input side of the certification standards body, the Rainforest value chain, to protect themselves from inferior Alliance Marketplace 2.0, which builds on agricultural inputs. ChainPoint software’s traceability product, and has broad track-and-trace functionality in Traceability solution providers active support of the Rainforest Alliance’s mission.146 in Africa fall into a few different categories including specialist Finally, in some African countries there are traceability software vendors, big also examples of national, government-run tech firms, certification organisations, track-and-trace solutions. The Namibian and government platforms. Specialised Livestock Traceability System traceability start-ups typically have deep (NamLITS), which has already proven expertise in the technical elements of its worth during recent foot and mouth track-and-trace solution development as disease outbreaks in the country, is one well as the ability to navigate issues of notable example.147 interoperability that are increasingly relevant given the proliferating number of food and Supply Chain environmental certification regimes. Examples Management – Input include solutions such as SourceTrace, SourceMap, EProd, and FarmForce.144 Quality Assurance and Anti-Counterfeiting The growing market for traceability solutions Input quality assurance and has also attracted big technology sector actors anti-counterfeiting D4Ag solutions such as SAP, which launched its Rural help agribusinesses ensure the CTA CHAPTER 2 67 integrity of their brands and help While some of the traceability solutions FAO farmers validate the authenticity described in the last section (e.g., SourceTrace) and quality of received inputs. can be applied fruitfully to input distribution to trace potential sources of fraud, counterfeiting, A major barrier to agricultural technology and mislabeling in input value chains, there adoption in Sub-Saharan Africa is the low are also more specialised D4Ag solutions that quality of many agricultural inputs, coupled with are starting to tackle the issue. a lack of reliable information on input quality.148 One example of such solutions is QualiTrace, Counterfeit products range from benign fake a Ghanaian startup with Africa-wide or adulterated materials to banned substances ambitions which uses track-and-trace that are harmful to crops and human health. technology to authenticate farm inputs and Beyond counterfeit products, the market for fight counterfeiting. QualiTrace not only inputs such as seeds, fertilisers, and pesticides/ authenticates but also provides analytics tools herbicides in Africa is also rife with sub- to trace products as they move from one standard products that do not effectively step to another until the final consumer also perform as they should, have substandard independently verifies the source and quality concentrations, or are simply expired.149 of the product.151 Other interesting examples of enterprises focused on agriculture input The ubiquity of substandard inputs directly authentication are mPedigree and Sproxil, reduces farmer productivity and, together with which have multiple digitally-enabled quality the perception of widespread counterfeiting, assurance solutions for input brand owners, reduces demand for high-quality inputs. This consumers, and governments, including lowers input prices and reduces profits for SMS or IVR unique identifier code producers of genuine products, causing a verification approaches and optical coding form of adverse selection in which counterfeit (e.g, 2D barcodes) that can be scanned by products push high-quality genuine products phone cameras.152 out of the market.150 68 CHAPTER 2 CTA Supply Chain agriculture value chain logistics tracking, Management – Logistics analytics, and optimisation through their apps for agribusinesses and farmers.154 IProcure, for Digital logistics platforms are tools that example, combines digital logistics surveillance, support the surveillance and operational analytics, and supply chain management improvement of physical storage tools with a physical network of agri-input and transport infrastructure and, in agents and warehouses that help agribusiness particular, the transport of agricultural aggregate and optimise smallholder input products across the full span of the supply chains.155 value chain from producers to markets. In the D4Ag context, logistics platforms can Virtual City and WeightCapture make complex, disaggregated value chains combine technologies for temper-proof more efficient and precise, a useful value digital weighing of produce with software proposition given the massive inefficiencies, that monitors the progress of agricultural physical infrastructure gaps (e.g., in terms of products across value chains with digital the quality and availability of roads, vehicles tracking at key hand-off points. Several of and storage warehouses) and corruption, the integrated supply chain ERP solutions theft and red tape that characterise the last- mentioned also have logistics components mile transport of agricultural commodities in their systems – for instance, a product and finished products into and out of rural transfer logistics tracking application that is a areas (and, similarly, the export/import of part of SourceTrace’s solution architecture. agricultural products over longer distances). As in the case of D4Ag Input Quality The use of digital solutions to address Assurance tools covered above, the digital logistics challenges is a much broader logistics solution sub-type serves a relatively topic than D4Ag.153 Most pertinent for small niche, but still has significant promise the purposes of this report are players like for solving the operational challenges of the iProcure, Logistimo, Virtual City, African agriculture sector as part of a broader and WeightCapture, which specialise in portfolio of complimentary digital solutions.156 CHAPTER 2 69 Supply Chain directly (and effectively) with smallholder Management – Supply farmers while also improving intelligence on and control over all aspects of value Chain ERP platforms chain activities. For small and medium-sized Supply chain ERP platforms offer a agribusiness, these types of tools are a means fully integrated package of digital of transforming companies with paper-driven services to agribusiness that duplicates processes into more mature and professional some elements of the solutions data-driven agri-enterprises that have the covered above, but goes well beyond information and management bandwidth to this to include operational analytics, grow in a more intentional fashion. value chain intelligence, and tools for managing smallholder farmers and For smaller and more downstream value chain agent field forces. intermediaries like cooperatives and agro- dealer networks, these tools focus on enhancing The types of data that need to be captured capacity and improving accountability. Finally, for traceability, logistics, and quality assurance for farmers, well-executed supply chain ERP uses are often identical to information needed solutions should make the process of accessing by agribusinesses to monitor key performance formal value chains more painless due to more indicators (KPIs), optimise operational streamlined and less time-consuming data performance, and glean insights into farmer capture; more available, knowledgeable and and agent field force behaviour. While there accountable field force agents; and access to is some resulting overlap between supply value-added tools that can be bundled with chain ERP solutions and those covered in the such platforms – e.g., free, high-quality and sections above, ERP solutions are a largely highly localised advisory services delivered by distinct D4Ag segment both in terms of agents via the supply chain ERP applications. functionality and the kinds of vendors that are involved. The number of supply chain management ERP solutions and Technically speaking, agricultural ERP providers is growing. Examples of platforms are solutions that integrate all core interesting solutions in this category within processes needed to run an agribusiness (e.g., the African D4Ag start-up ecosystem include finance, HR, manufacturing, supply chain, Farmforce, EProd, and Metajua.158 These services, procurement, and others) into a enterprises tend to focus on small to medium- single system.157 We use the term ERP more sized African agribusinesses, typically with a loosely to indicate digital solutions that support range of 1,000 to 20,000 smallholder farmers farmer and field force management tools for being managed per each agribusiness ‘account’ CTA smallholder value chains, typically integrated with traceability, logistics management, quality assurance, and business intelligence elements. The overall value proposition of D4Ag supply chain solutions is to improve the effectiveness and cost- efficiency of smallholder-centred African agribusinesses at every level of operating scale. For the largest agribusinesses (i.e., global buyers/processors or global input providers), these tools are a way to reduce the costs of interfacing 70 CHAPTER 2 or ‘license’. Others like TaroWorks focus on own in-house agriculture value agriculture sector NGOs.159 AgriGo focuses on chain digitalisation tools to support even smaller players like farm cooperatives. smallholder farmer registration, communications, data collection, supply A few supply chain ERP start-ups from chain management/logistics, traceability, other geographies – such as CropIn, and business intelligence needs. The most SourceTrace and Annona – have also widely discussed of these types of platforms for brought their solutions to the African Africa is the Olam Farmer Information market.160 Some of these, including CropIn System (OFIS), which Olam uses to and SourceTrace, aspire to serve large-scale manage more than 250,000 farmers across its agribusinesses and already have extensive countries of operation today (both in Africa experience in working with big national or and Indonesia) – with a target of 500,000 international buyers in India. farmers globally by 2020.163 In addition to serving the immediate internal needs of Olam Big tech enterprises in the supply chain ERP from the perspective of farmer certification segment that focus on smallholder farmers, and traceability, the platform is also a tool such as SAP’s Rural Sourcing Platform for Olam’s country-level intermediaries (e.g., and Accenture’s Connected Crop Solution farmer groups, cooperatives) and field force (ACCS), focus on serving the needs of agents to manage their own organisations, medium-sized and large agribusinesses. SAP’s counterparties, and finances. solution, for example, focuses on global and regional sourcing organisation. ACCS, on While less known in the public domain, several the other hand, focuses on medium-to-large other large buyers and processors active in agri-input organisations and aims to connect Africa have also invested heavily in their own the three key stakeholders in that value chain in-house digital supply chain management and – the field agent, the agri-input company, and track-and-trace solutions that have comparable the farmer.161 Another relevant technology features, but are not always integrated into one initiative is the Connected Farmer solution supply chain management platform. from Vodafone, developed by Vodafone’s Mezzanine team and focused on smallholder value chain SMEs and medium-sized Financial Access Use Case agribusiness to allow such players to effectively D4Ag financial access solutions facilitate and cost-efficiently enrol and manage the farmer access to payments, savings, smallholders they work with.162 credit, and insurance, or – less directly – provide data analytics and digitalisation A few large agribusinesses focused support to financial service providers Marco Salustro, IFAD on Africa have developed their that can then serve smallholder farmers at broader scale and lower cost. By any global measure, African farmers, especially smallholders operating on plot sizes of two hectares or less, face chronic challenges of limited access to financial services – including savings, credit, and insurance.164 From the perspective of smallholder farmers, the overarching objective of financial access D4Ag solutions is to provide a link to high- quality and affordable financial products and services that create an array of new CHAPTER 2 71 FINANCIAL ACCESS Figure 9 Financial access – overview of sub-use cases and solution examples Payments Savings Credit Insurance Financial access Financial analytics FSP digitalisation Crowd-farming 72 CHAPTER 2 opportunities – among them, the ability update to the sector-shaping smallholder to transact at much lower cost with input finance ‘Inflection Point’ reports produced by a providers and purchasers of their products, consortium of leading experts on the subject. purchase the inputs they need to increase their productivity and incomes and significantly Considering the available work of these parallel reduce their risks from weather, pests, plant knowledge initiatives, our primary intent in the diseases, cross-border market disruptions and following sections is to provide an overview a myriad other factors that make smallholder of key financial access solution segments with farming in Sub-Saharan Africa such a some illustrations, rather than diving more financially precarious livelihood. deeply into financial services trends and economics. Issues of smallholder farmer financial access are incredibly complex and the ecosystem D4Ag financial access solutions need around this topic is rapidly evolving given not achieve their positive impact on the rapid transformation in underlying data smallholder farmers directly to qualify analytics and payments technologies, financial for this discussion. services business models, and resulting financial products. Some of the D4Ag solutions covered in this section are, indeed, themselves financial service There are also a number of technical expert providers (FSPs) that use digital channels and organisations, like IFC/CGAP, the MasterCard other types of digital tools to deliver new Foundation’s Rural Agriculture Finance types of digital payments, savings, credit, or Learning Lab (RAFLL), the Initiative for insurance products to the farmers they serve. Smallholder Finance (ISF), and NGOs like This includes both new fintech entrants as well Mercy Corps (via its Mercy Corps AgriFin as some more traditional banks and MFIs that Accelerate programme) and AGRA that are all have integrated digital technology into the way investing in advancing the knowledge frontier they serve farmers and have launched new on the market trends, business models, and digital business units or products.165 impacts of smallholder-farmer-focused digital financial services through regular research Many of the D4Ag solutions that we cover publications – such as the forthcoming 2019 under this use case, however, function Marco Salustro, IFAD CHAPTER 2 73 indirectly. Such solutions are B2B service Financial Access – providers that benefit farmers by working with local financial institutions of varying Payments types and scales. The value proposition of Payments allow smallholder farmers, such D4Ag enterprises to financial institutions input providers, buyers and others to can encompass a few different drivers of exchange money with each other without digitalisation value-addition including (i) cash. Mobile payments significantly lower helping FSPs identify and connect with transaction costs and increase efficiency as smallholder farmers they would not otherwise money can be transferred electronically. Money be able to find (or be able to find profitably); leaves and enters bank accounts with less lag (ii) reducing the operational costs for FSPs time, with little risk of being lost or stolen, of working with smallholder farmers (e.g., and regulatory constraints on the amount of lowering costs of risk assessments, payment cash one can carry become irrelevant. For transactions, credit collection processes, these reasons, the ability to conduct mobile insurance claims processing, etc.), and, payments is a baseline enabler for many other most critically; (iii) de-risking farmers so that types of smallholder farmer financing solutions. they can become ‘bankable’, i.e., so they Sub-Saharan Africa, notably, is the only region can be served at the very least profitably in which more than 20% of adults have a and ideally with sufficiently attractive mobile money account; over the past five economics to justify pursuing smallholder years, the share of adults with such an account farmers clients (and related financial products) has risen roughly twice as fast as that of adults versus other alternatives. 166 with a traditional, formal bank account.167 This report identifies six important sub- The D4Ag payment solutions this types of D4Ag financial access solutions: report is concerned with are derivative (i) payments; (ii) savings; (iii) credit; payment services rather than general (iv) insurance; (v) crowd financing (crowd digital payments solutions like M-Pesa; farming); and a B2B solution area of the services in question are tailored to (vi) financial analytics and process smallholder farmers’ needs and solve digitalisation for financial service for very specific challenges in African providers (see Figure 9). smallholder farmer value chains. Marco Salustro, IFAD 74 CHAPTER 2 The most acute challenge from a payments Such agriculture value chain digitalisation perspective is that cash is still king for most initiatives, driven directly by MNOs or by transactions, agricultural or otherwise.168 traditional FSPs and fintechs leveraging MNO Despite digital payment systems that are digital payments infrastructure, are currently growing quickly – and now becoming an intensive focus for many sector experts and ubiquitous in some African countries, such as intermediaries like the Better Than Cash Kenya – the average African smallholder lives Alliance and the GSMA’s mAgri team.171 in remote areas where mobile network coverage can be weak or non-existent and, most critically, There are a number of D4Ag players that mobile money cannot yet be used to purchase are trying to support farmer payments goods and services from local merchants. digitalisation and the development of broader agriculture digital payments Smallholder farmers are therefore hesitant to ecosystems. accept digital payments from buyers; when such payments do come in a digital form (e.g., One model involves supporting G2P payments from government rural livelihood or agriculture (typically various types of direct transfer sector subsidy schemes), the experience rural livelihoods or agriculture sector subsidy of most African subsidy direct transfer schemes) for farmers via innovative e-wallet programmes – such as Cellulant’s e-wallet models that tie subsidy transfers to agricultural in Nigeria – suggests that farmers prefer to input payments, while at the same time cash out immediately.169 The big near-term trying to add sufficient value to the e-wallet opportunity for smallholder farmer payments account to build farmer familiarity with and in the coming few years is therefore to drive use of digital payments for a wider variety of broader agriculture value chain payment goods and services. Ultimately, the e-wallet digitalisation via business-to-person (B2P) and can serve as a stepping stone to other digital government-to-person (G2P) payment schemes financial products like commitment savings, involving farmers, as well as efforts to create input credit, and agricultural insurance. meaningful agricultural (and non-agricultural) The largest-scale example of this model was product and service choices for farmers where Cellulant’s work earlier this decade with the digital payments are accepted, so that the value Nigerian government’s Growth Enhancement proposition of digital payments increases.170 Support (GES) Scheme.172 Other innovative M-Pesa CHAPTER 2 75 examples include Zoona’s e-voucher model Saving money is, however, very challenging for for agriculture173 and, most recently, the smallholder farmers. The first obvious issue is Smart Nkunganire System in Rwanda, access to appropriate, affordable, and accessible which is helping to drive agriculture payment savings products. According to the most recent digitisation at a national scale.174 regional data, only 19% of Sub-Saharan African adults saved semi-formally via channels In the B2P payments space, innovative models like village savings and loan associations worth highlighting include SmartMoney (VSLAs), and just 9% saved formally through in Tanzania and Uganda and AgroPay in bank, MFI, or savings and credit cooperative Ghana. Both models combine the digitalisation organisation (SACCO) savings accounts.178 of agriculture value chain payments with efforts The second challenge is that when farmers to create broader village-level digital payments do have savings accounts, usage is often low. and digital payments acceptance ecosystems. Saving is hard for everyone; it is especially SmartMoney, for instance, currently serves so for poor smallholder farmers with volatile more than 200,000 rural people and over incomes and urgent expenses. 2,000 merchants, and follows the model of establishing ‘E-Villages’ – village-wide, Digital savings for farmers is an important area of innovation for CTA ledger-based digital money ecosystems that are supported by digitalised payments from solutions that are starting to address agricultural off-takers, on the one hand, and, both access and savings behaviour on the other, by activities to promote digital challenges. payment uptake for agri-input providers and Digital technologies are addressing the a wide range of other small, rural businesses challenge of smallholder farmers’ access to and merchants.175 savings via electronic wallet products that Financial Access – Savings have savings features, either directly when offered by formal financial institutions, or in The use of savings products can make partnerships between payments players who a big difference in the lives of poor already have extensive rural reach and deposit- farmers. Smallholder farmers typically get taking financial institutions with banking much of their income in a few big lump-sum licenses. The primary feature of such digital payments each year during harvest times and savings models is that payments and e-wallets then need to pay down debts, save money are used as an entry point for extending for day-to-day expenses between seasons, and savings account access to large numbers of lay funds aside for next year’s seed, fertiliser, smallholder farmers. and other productivity-enhancing farming inputs.176 Savings are thus needed to ensure We touched on one variant of this model expenditure smoothing across variable seasonal above with national scale e-wallets tied to income patterns, to make farm investments, subsidy schemes, such as Zoona in Zambia and and to build household resilience in the face the IFIKO universal wallet integrated into the of agriculture-related shocks (e.g., pest/disease Smart Nkunganire System (SNS) in Rwanda.179 infestations) or personal financial crises (e.g., Zoona partnered with FINCA Zambia in unanticipated health expenditures). When late 2018 and now allows farmers with Zoona smallholder farmers use savings accounts, this e-wallet accounts to earn a 10% interest on can make a major difference in the amounts their savings.180 Similarly, farmers registered they save and invest in their farms, which with SNS in Rwanda will be able to get access directly translates into increased farming to savings accounts through the Bank of Kigali. profits, improved long-term incomes and Other models in this space involve MNO higher levels of consumption.177 partnerships such as Safaricom’s partnership 76 CHAPTER 2 with CBA on the M-Shwari savings product and credit association (ROSCAs). In the past in Kenya and Econet’s partnership with 2–3 years, these organisations have started Steward Bank on the digital EcoSave product experimenting with digital or digitally-enabled in Zimbabwe. Not all of these products target savings group models in order to reduce smallholder farmers exclusively, the smallholder costs of group formation and support and to farmers tend to be major beneficiaries. allow savings group members to access the broader benefits of payment digitalisation. To address the behavioural challenge – Since 2016, The Aga Khan Foundation, for getting smallholder farmers with access example, has supported the aggressive rollout to savings accounts to actually save – of digital savings groups (DSGs), managed D4Ag players are experimenting with via the Foundation’s DSG Platform, a different types of digital commitment shared software service implemented with savings accounts. One D4Ag solution that both USSD and application interfaces that has extended the e-wallet model in interesting fully digitalises savings group management.183 ways for commitment savings is Agri-Wallet Another interesting example from the D4Ag in Kenya, a recent start-up that has developed startup space is Akobaxi in Uganda, which a free digital wallet for the agricultural sector digitalises village savings groups via a system as a business account for farmers, which they that includes an electronic ‘box’ (a customised, can use to save, buy, and earn. When farmers connected point-of-sale device), Akobaxi’s earn revenue through sales, they can choose to cloud-based software that runs on this device Akaboxi be paid in money through M-Pesa or in tokens for managing and monitoring savings group for their wallet that are earmarked for operations and transactions, electronic ID purchasing input supplies from vetted merchants cards readable by the device for individual and drive beneficial savings behaviour.181 savings group members, and SMS-based communication to savings group members for Another example of the digitalised commitment transaction records.184 savings model is myAgro in Mali, which helps smallholder farmers in West Africa pay on layaway (i.e., via piecemeal instalments) Financial Access – Credit for fertiliser, seed and training packages using In recent years, D4Ag solutions have their mobile phone. Registered farmers can been a major source of experimental save easily by continuously ‘topping up’ their pathways toward confronting the myAgro account in flexible amounts approximately €25–30 billion financing (€0.90–44.90). The myAgro mobile layaway gap facing African smallholder model makes saving for input purchases easy, farmers.185 Given the relatively small size drives input adoption via the commitment of smallholder transactions, the physical and savings model, and, as a result, appears logistical difficulties of serving clients in remote to generate substantial positive impact for rural areas, the complexity of agricultural risks farmers’ yields and incomes.182 (e.g., agroclimatic, commodity prices), and other unusual features of agriculture finance Innovative D4Ag savings products are stemming from its seasonal nature, most also being developed addressing the formal financial institutions perceive lending needs not only of farmers as individual to farmers as too risky or, at the very least, customers, but also targeting informal insufficiently profitable. In Sub-Saharan Africa, farmer savings groups. Organisations like for instance, only 95 of 900 banks surveyed Care International and the Aga Khan provide financing to smallholder farmers.186 Foundation have been working for years on the formalisation and scale-up of informal To address the challenge, multiple savings groups like VSLAs and rotating savings digital lending products specifically CHAPTER 2 77 designed for farmers have been launched One pathway for these approaches involves in recent years and, more broadly, more traditional financial institutions that many lenders are digitalising elements are digitalising their products and interaction of their operations. Digitalisation can models. Examples include the KCB’s come in different flavours in the context of MobiGrow product in Kenya, Advans’s smallholder lending. Some FSPs – including digital cocoa-farmer credit product in Côte both incumbents and new fintech entrants – d’Ivoire, and Opportunity International’s are deploying digitally branded credit products digitally-enabled loans in Ghana.188 that involve little or no in-person farmer engagement, rely on digital communications From fintech innovators, important examples for client acquisition and servicing, and use worth monitoring include the digital Kilimo digital payments for loan disbursement and Booster farmer credit product from Musoni in payment collections. Other FSPs are starting to Kenya,189 the digital agriculture credit model of integrate digital tools, such as digitally-enabled Akellobanker in Uganda,190 Tulaa’s digital automated credit scoring, but continue to use credit offering in Kenya, which is integrated a blend of digital and human channels for into a digital end-to-end market linkage smallholder financing operations. model,191 and digital loans from Apollo Agriculture, also in Kenya, which are For financial institutions, the primary bundled with a digital advisory product. Many motivation for pursuing digitalisation is to of these players rely on digitally-enabled credit reduce customer risk and to lower cost to serve scoring algorithms.192 (e.g., no need for loan officers to travel to the field with paper applications or branch-based As noted in a recent review by the IFC, while loan disbursement and repayment processes), the number of digital lending products is both of which should ultimately translate growing, it is at this stage premature to assess into higher profitability and much broader the extent to which these models are reach that includes otherwise un-bankable commercially viable and at what scale. The clients. The evidence base for the impact of authors of this report are nonetheless optimistic digitalisation on financial service provider about a number of these models based on the economics is still at a very early stage, but the emerging evidence of both smallholder impact indications are positive, and the pace of digital and tangible business model benefits from initiatives and products is picking up.187 digitalisation.193 At the same time, it is also Thomas Mukoya, Reuters 78 CHAPTER 2 IBLI important to recognize that while digital ranging from 10–30% over 3–12-month smallholder farmer credit models address some periods and are divided among participating of the systemic challenges of traditional farmer subscribers and the crowdfunding enterprise finance, they can also introduce their own after the harvest season.196 When bigger risks. These could include the risk of over- investments are in question, many sponsors can indebtedness due to the relative ease of support a farmer together – for example, by accessing credit digitally, consumer protection ‘sharing a cow’.197 concerns about smallholder farmer clients not understanding the products they sign up Facilitated by digital platforms, African for (in the absence of human loan agent smallholder farmer crowdfarming interaction) and data privacy worries given solutions play a bridging role between the large amounts of farmer data (and individual providers and recipients external data focused on the farm, e.g., from of farm financing. Firms on the finance satellites) that are collected and mined by supply side of the model focus on aggressively digital credit solutions.194 marketing farm investment opportunities via digital channels to attract potential farm Financial Access – investors/financiers from international and diaspora communities or African urban middle Crowdfarming class investors. From a demand-generation Another response to the smallholder perspective, these firms recruit smallholder farmer credit challenge, albeit with farmers to join their platform and work with a very distinct business model, them to attractively package the investment ‘crowdfarming’ solutions use digital opportunity to finance suppliers, often with platforms to link farmers who need the addition of a variety of other value-added capital with sponsors who wish to services such as digitally-facilitated off-take or invest.195 Crowdfarming entails sourcing input market linkages, advisory services, and funds from multiple individuals to invest complimentary agricultural insurance. in a smallholder farmer or other small- scale agricultural enterprises. In some cases, The contractual agreement between the investors, often labelled as ‘subscribers’, receive crowdfarming platforms and farm returns in the form of agricultural produce, subscribers provides details on the returns but typically the returns are financial usually on investment per farm enterprise, length of CHAPTER 2 79 IBLI the production/investment cycle, insurance for marketing farm investment opportunities coverage on funds invested, and secure and a starting supply of farmers and investors online payments.198 Farm ‘subscribers’ also – requires minimal investment. Several of typically receive regular information on the the experts consulted for this report have farm’s progress through email alerts and highlighted that this low barrier to entry (and notification of final payments at the end of relative opacity of the actual value-add that the production cycle.199 some of these platforms deliver to their farmer after the initial farm selection) likely means Our research suggests that there are ~30 that some of the copycat models that have crowdfarming enterprises in Africa today, emerged recently do not have much substance with 80% of these businesses appearing in the behind them. They may even present risks past 1-2 years in the wake of Farmcrowdy’s to investors. We do believe that the more success in Nigeria.200 Other prominent established and vetted crowdfarming players examples of crowdfarming businesses include have as much promise as the highly integrated Growsel and Thrive Agric in Nigeria, D4Ag market linkage models that also bundle Livestock Wealth in South Africa, and credit from more conventional sources. Bayseddo in Senegal.201 It is too early to assess the success Financial Access – of crowdfarming models and hard Insurance to generalise about the category in Agricultural insurance offers a valuable terms of farmer value-add given the tool to help smallholder farmers avoid wide diversity of underlying business devastating financial losses and limit models. Many of the D4Ag enterprises are downside risk associated with investing serious, legitimate businesses with thoughtful in their own productive capacity.202 business models, often melding elements of Without insurance, farmers are highly digitally-enabled advisory services and digital vulnerable to external shocks given their market linkages support for farmer clients exposure to environmental hazards (e.g., pests with a crowdsourced financing engine. Other and diseases, weather events), the vagaries of solutions in this segment are much more global and regional agricultural commodities questionable. The minimum viable product markets, and the growing unpredictability version of crowdfunding platforms – a website across all of these factors brought on by 80 CHAPTER 2 climate change. Smallholder farmer surveys products of any time. This skepticism is not consistently show that such risks, particularly entirely unfounded as the costs of insurance climate-related risks, are already impacting products can be high and pay-out mechanisms farmers, often producing disastrous losses. can be slow and cumbersome – or divorced Anywhere between one-fifth and two-thirds of from the reality of the loss-making event as smallholders across a wide range of African perceived by the farmer.206 On the supply side, countries report an instance of major crop loss smallholder agricultural insurance is a over a five-year time period due to catastrophic complex product to design. Most importantly, weather events (e.g., floods, droughts) or due the costs of delivering insurance to often to factors such as pests and disease (which are unwilling and remote smallholder farmer likewise linked to climate change).203 customers can severely constrain the profitability and attractiveness of such Insurance helps mitigate such risks and unlocks products to conventional insurers. opportunity. For example, a recent survey of the literature highlighted that, “with insurance Digital technology is starting to break for agricultural livelihoods, smallholders down the barriers that prevent insurance invest more in their farms, education and providers from serving the agricultural health; whereas, without insurance, farmers sector in general and smallholder adopt lower risk-and-return farming practices, farmers in particular. By aggregating new eschewing investments into more productive sources of data and methods of analysis, D4Ag practices or technologies.”204 Rates of access to insurance solutions allow providers to better agricultural insurance for smallholder farmers predict risk and to execute claims processing are at extremely low levels, however, with only at much lower costs (e.g., automated pay-outs an estimated 20% of smallholders globally and based on remote sensing data). New data only 3-6% in Sub-Saharan Africa using sources in this context primarily constitute such products.205 weather data (weather index insurance) and satellites (satellite insurance), which allow As with other financial access products, the experts to analyse farm plots and weather- reasons for low uptake are multiple. On the related risks and yield implications at scale demand side, smallholder farmers generally and with increasing nuance and detail even as have low levels of understanding of and the costs of such remote sensing data decline trust in complex financial products and, in annually. More complex D4Ag insurance particular, are highly skeptical of insurance models involve a triangulation from more CTA CHAPTER 2 81 granular weather data, remote sensing channel that is gaining in popularity is the satellite data, ground sensors (e.g., field bundling of digitally-enabled agri-insurance precipitation monitors) crowdsourced pest products into MNO farmer advisory and and disease reports that allow for more payment services. For example, Econet in accurate surveillance and projection of Zimbabwe offers its EcoFarmer insurance pest and disease risks, and more nuanced product – weather-indexed insurance for which data about the farm itself (e.g., soil health the pay-out is dependent on abnormal rainfall diagnostics) that enable more refined and all premium and claim payments are predictions of yield losses. processed via digital channels. Examples of key D4Ag insurance solutions Digitally-enabled smallholder farmer include relatively established and large-scale insurance solutions are growing in scale (in terms of farmers covered) specialist firms and have significant promise, but many like Pula and ACRE Africa and more barriers likely still need to be overcome recent digital crop insurance entrants like before the African market will see mass- Oko and World Cover.207 Some of these scale uptake of agri-insurance. Recent players focus primarily on data analytics reviews of the smallholder agri-insurance (e.g., Pula, Oko); others are themselves opportunity broadly, and D4Ag solutions distribution intermediaries – see, for example, for insurance in particular, suggest reason WorldCover in Ghana and SumAfrica in for optimism but while also sounding notes Uganda,208 which identify and acquire clients of caution. IFC’s late-2018 overview of the and service insurance portfolios on behalf D4Ag insurance opportunity has concluded, of or in consortium with more traditional for example, that developing D4Ag “insurance insurers. Still others focus on delivering B2B schemes that balance commercial viability of insurance products to other farmer financing a product linked to a volatile sector where intermediaries, such as the WINnERS risks are not easily mitigated and the need to model in Tanzania of providing weather compensate farmers when they experience (precipitation) insurance coverage to banks that agricultural losses remains a challenge”, and have large smallholder financing portfolios.209 highlights that while there is a good deal of promise and some scale for products like In addition to D4Ag insurance players who index insurance, “most products in this space, partner with insurers and other traditional including those enabled by digital technology, financial institutions, an alternative distribution have yet to exit the pilot stage.”210 EcoFarmer 82 CHAPTER 2 A more in-depth recent study by the space (e.g., banks, MFIs, SACCOs, Initiative for Smallholder Finance (ISF) MNOs) is the FSPs’ limited institutional has likewise highlighted the challenges of capacity for digitalisation; this is an supplier economics (e.g., economically viable opportunity that a number of D4Ag distribution models) and the demand-side solutions are now attacking with B2B challenge of the fact that “the vast majority service delivery models. While scaling of smallholders still don’t understand, trust, up D4Ag financial access products requires or see sufficient value in the products that are overcoming many other demand- and supply- available”. ISF’s review cautiously concluded – side challenges, one common thread is the and this report’s authors concur – that despite constrained ability of traditional, ‘analogue’ many achievements to date and the important FSPs to rapidly design, prototype, and deploy contributions of D4Ag innovators to the sector, digitally-enabled products for farmers.211 Most “agricultural insurance for the smallholder FSPs struggle, for instance, to develop state- farmer market likely requires another five to of-the-art in-house data analytics capacity. ten years of product, process, and technology Many find it hard simply to build up sufficient innovation to break down complexity and management sophistication on data analytics or continue to expand the realm of the possible.” bring their internal data systems to a sufficient state of digitalisation to effectively interface Index insurance, for example, holds strong with third-party analytics vendors who can promise in terms of allowing providers to help. More prosaically, many African FSPs create business and operating models that can struggle with even more basic digitalisation be commercially scaled and sustained in rural initiatives such as digital data capture and geographies provided the pool of policyholders records management, the build-out of digital is large enough and adequately dispersed communication channels with clients, or the geographically to distribute risk. digitalisation of internal credit risk assessment and monitoring functions. The challenges are Financial Access – in part due to resource constraints and the Data Analytics and often very long timelines of internal ‘digital transformation’ initiatives. Another underlying FSP Digitalisation challenge is one of institutional incentives, One cross-cutting challenge for many particularly in the lower levels of incumbent financial service provider types in the organisations, where digital technologies are smallholder farmer financial access more often seen as a threat than an opportunity. Cecilia Schubert, CCAFS CHAPTER 2 83 Cecilia Schubert, CCAFS An emerging cluster of D4Ag solutions FSP digitalisation services. For example, are focused on these FSP challenges. MOBIS, Ensibuuko’s financial management platform, is a, cloud-based microfinance Financial analytics D4Ag enterprises management platform designed uniquely specialise in collecting and analysing data on to help savings and loans cooperatives go the financial habits of farmers and triangulating paperless and become more efficient by such information with alternative data sources digitising how they manage customer data including satellite data, weather data, and and transactions.217 MOBIS serves 50 African soil quality data.212 These approaches use a SACCOs, which collectively reach close to variety of basic and advanced technologies 300,000 farmers in Uganda and now are to analyse this data in value-added ways and expanding in other countries. Similarly, YAPU to deliver risk assessment insights to financial has focused its business model on turnkey institutions such as banks, insurance providers, digitalisation of the lending, data analytics, and MFIs. Key innovators in this space include and customer engagement processes of FSPs players such as FarmDrive, Harvesting, that focus on smallholder farmer finance, YAPU, and SatSure.213 Other initiatives – particularly small and medium-sized MFIs, such as a collaboration between Rabobank allowing such institutions to grow their books Foundation214 and MUIIS project215 and profitability while also boosting smallholder and a CTA-led initiative with IGTF and clients’ yields and incomes through the benefits NUCAFE216 are using this model to increase of sophisticated weather and satellite analytics. access to smallholder farmers and cooperatives in Uganda. These efforts also tackle the issue of inaccurate farmer data acting as a barrier Macro Agricultural to accessing credit. Using detailed farmer Intelligence registries, including GPS coordinates of Solving the complex challenges of farmers’ fields, provides a kind of guarantee African smallholder agriculture to FSPs that they are basing their credit requires timely, accurate, granular, decisions on an accurate representation of the and large-scale data, combined with smallholder farmers they are working with. insightful analyses. Such data and insights are often missing today for key macro Another important variant of B2B D4Ag decision makers including Sub-Saharan Africa solutions in the financial access use funders, government policymakers, case are enterprises that specialise in and agribusinesses.218 84 CHAPTER 2 A number of disruptive technologies such variables of interest like yield projections, as remote sensing via satellites and drones, crop losses, supply-demand mismatches, innovations in low-cost and more compact agriculture jobs trends, climate-impact weather station technologies, low-cost/high- indicators, and granular real-time food and throughput soil testing spectrography, and nutrition security maps. the emerging smallholder-focused internet of things (IoT) are already generating massive This report is not novel in flagging the new datasets about farm fields and agroclimatic ‘Data4Ag’ opportunity; many actors have conditions that have not been previously been on initiatives to develop, support available or have not been available at such and govern this ag data ecosystem for low cost to agriculture sector actors. There is a the past 5+ years. parallel explosion in the volume of geotagged data about farmers and their needs and Major examples of initiatives focused on the behaviours (e.g., data from farmer registries, Data 4 Agriculture ecosystem in recent years open government agriculture data initiatives, include CGIAR’s Big Data 4 Agriculture payments companies involved in agricultural initiative,219 ODI’s Open Agriculture value chain digitalisation, and digital credit and Initiative,220 the Global Partnership for insurance providers). The volume and velocity Sustainable Development Data (GPSDD) of both of these data universes – data about (and its agriculture-focused programming),221 farms and data about farmers – will continue and the rapid growth of the Global Open to accelerate rapidly over the next decade. Data for Agriculture and Nutrition (GODAN) network.222 Despite this dynamism In parallel, innovations in data analytics tools in the ecosystem and the growing volumes of and methodologies (e.g., big data pattern data, there is universal consensus that very few recognition, image processing, machine agriculture actors in Africa actually use learning techniques) mean that there is now macro-scale data analytics and insights tools a growing opportunity to bring very different that can take full advantage of agriculture types of datasets together in unique ways to data’s potential. offer decision makers of all types the ability to monitor real-time agricultural trends at large A small but growing number of D4Ag scale and, more importantly, to forecast key macro agricultural intelligence start-ups L. Sharma, Marchmont Communications CHAPTER 2 85 MACRO AGRICULTURAL INTELLIGENCE Figure 10 Macro agricultural intelligence – overview of sub-use cases and solution examples Macro agricultural intelligence are working to fill the data insights gap – Gro Intelligence – has attracted significant by putting practical and powerful tools commercial investment to date. in the hands of African decisionmakers. D4Ag macro agricultural intelligence We are tracking roughly three dozen D4Ag solutions include a few very different types of actors that have macro agri-intelligence as part organisations. These include government or of their mandate, and under a dozen solutions donor ag data analytics and surveillance that have agri-intelligence as their primary platforms; surveillance and (more rarely) focus. A third of these players appeared in forecasting tools, typically focused on weather the past 1-2 years; 80% of these players data or food security but often now starting to are under five years old. Given their recent integrate other data sources and analytics use vintage, most of the commercial players in this cases for the benefit of government decision segment are still in the pilot or early scale-up makers; the agronomy research community stage; only one of the Africa-based enterprises and its funders; commercial agriculture 86 CHAPTER 2 data analytics platforms that draw on has in recent years significantly broadened and integrate third-party data and then put its use of data sources and its deployment of productised self-service data, data analytics analytics techniques;225 GeoGlam, a donor- and data visualisation tools into the hands funded global agricultural monitoring platform of decision makers; commercial remote that runs tools like the Global Crop Monitor sensing and weather data analytics for early warnings focused on assessing and specialists that have proprietary data forecasting crop conditions in countries at collection assets and specialise in specific risk of food insecurity;226 and more recent data types, but also develop value-added data arrivals like the World Food Program’s intelligence products marketed to agriculture Vulnerability Analysis and Mapping decision makers or other agri-intelligence (VAM) platform227 and Africa country-specific intermediaries; and custom ag data agriculture surveillance platforms currently analytics providers that bundle data and being piloted by CropWatch, China’s leading data analytics with consulting and advisory crop monitoring system, for countries like models (e.g., working with agriculture sector Mozambique.228 investors or specific agribusinesses to deliver value-added market intelligence insights or Of the commercial solutions for Sub-Saharan support specific decisions). Africa macro agri-intelligence analytics and visualisation, Gro Intelligence is For government and donor agri-intelligence the Africa market leader.229 The company platforms, the most prominent example today focuses on aggregating and integrating is likely the World Bank Group (WBG) disparate agriculture datasets – most notably, Agriculture Observatory,223 and country- government agricultural data, weather data, level platforms of a similar type such as soil data, and satellite data imagery – and KALRO’s Kenya Agriculture Observatory then translating that data into trend analysis, Platform (KAOP)224 and a few weather useful visualisations, and (for some variables surveillance observatories, which are likewise like yield) different types of forecasts. The data primarily supported by the World Bank. are marketed to a variety of end-users across Other examples of large donor-funded agri- government, agribusiness, and the private intelligence platforms include FEWS NET, the sector, but the company’s focus is on more leading famine early warning and surveillance commercial (agribusiness and commodity World Bank system that has been in place for decades but investors) decision makers. Other commercial macro agri-intelligence players tend to focus on both self-service data decision tools and bespoke agri-intelligence analytics for private sector and public sector clients. Examples of such solutions include Tata Consultancy Services (TCS) AgEye,230 SatSure’s 6th Grain,231 McKinsey’s ACRE,232 and Dalberg’s CubicA.233 Finally, a number of players specialise in satellite or weather data analytics for agriculture with a strong focus on macro agri-intelligence applications. AWhere is the most established example of such solutions for agriculture-related weather analytics. In the satellite data space, interesting examples include SatSure234 and Satelligence.235 CHAPTER 2 87 Georgina Smith, CIAT Global big tech players like Microsoft (via programmes (e.g., World Bank’s Kenya their Microsoft AI for Earth team)236 and agriculture transformation programme and it’s Google (Google Earth Engine)237 are also KAOP component). exploring macro agri-intelligence applications that have relevance for Africa, but have not yet In the case of more commercial macro agri- developed their tools into products targeted at intelligence solutions, as noted above, most the agricultural space in the region. of the actors in this category are at an early stage of proving their value and business The macro agri-intelligence opportunity models. Furthermore, despite growing interest, is still in its very early days and data and data analytics monetisation in the commercial prospects for many of the context of developing Africa is still a very models are uncertain, but we are likely difficult business with sceptical and resource- to see many new solutions in the next constrained institutional clients and fairly few years. risk-averse agribusinesses (when it comes to paying for third-party data and data analytics From the perspective of government and technologies). This makes the economics of donor-funded macro agri-intelligence platforms, stand-alone macro agri-intelligence businesses our expert interviews suggest that we are on challenging in the near term; however, the cusp of significantly increased investment since macro agri-intelligence is often a into national agri-intelligence system supplementary or ancillary data stream for development, either as stand-alone projects or many players in the sector, experimentation as knowledge and monitoring and evaluation and market entry will continue to grow quickly (M&E) investments bundled into much even if it outpaces commercial viability for larger national agricultural transformation many actors. 88 CHAPTER 2 Farmerline An Emerging D4Ag a fully distinct and mature use case – many Use Case – D4Ag D4Ag enterprises are just beginning to build out their service bundles and to refine their ‘Super Platforms’? value proposition – D4Ag super platforms There is an emerging D4Ag use case of were repeatedly highlighted in our expert ‘super platforms’, solutions that bundle interviews as a fast growing and highly multiple D4Ag services and deliver a promising path forward for the sector. The fully integrated digital value proposition report’s authors strongly endorse this view. to smallholder farmers and other agricultural value chain intermediaries. D4Ag super platforms link farmers to buyers and to the broader ecosystem D4Ag ‘super platforms’ are solutions that of finance, advice, and other services, straddle many – and, at times, all – other thereby eliminating layers of D4Ag use cases. At the very minimum, super intermediaries and creating immediate platforms combine digitally-enabled market economic value. linkages, digital finance, and digital advisory services into an integrated service bundle While there are many variations of these for farmers. When they operate at scale, models, all super platforms follow the logic of these platforms can deliver immense value value chain supply and demand aggregation to smallholder farmers, greatly reduce risks and formalisation. Typically starting with and transaction costs for all agriculture value digital payments, often bundled with digitally- chain actors and, at the same time, generate enabled off-take linkages, these solutions attractive economics for D4Ag enterprises. result in more reliable access to markets, which, in turn, encourages farmers to invest We have adopted the term D4Ag ‘super in productivity enhancements – most notably, platforms’ – a helpful encapsulation of the purchase of farm inputs. Farmers buy the the scope and ambition of such business necessary inputs through the super platform models – from MercyCorp’s AgriFin due to convenience, more attractive prices Accelerate team.238 Other names for these (i.e., improved bargaining power vis-à-vis models or analogous concepts in the literature input sellers), and strongly aligned incentives include holistic service delivery models on input quality, since the super platform (SDM) and ‘integrated digital agriculture also partakes in the upside of higher farmer marketplaces.’239 Although they are not yet productivity and incomes. CHAPTER 2 89 SUPER PLATFORM SOLUTIONS Figure 11 D4Ag super platform solution examples Government Payments MNO Bank Smart Nkunganire System Start-up market Agribusiness Donor-led/PPP linkage specialists Global digital e-commerce Smallholder farmers also have an incentive to ensure cost-efficiency, support traceability, and access credit (and bundled agri-insurance) from improve time to market. the platform. These financial services are likely to be far more affordable than alternatives The core insight of emerging D4Ag super due to the super platform’s privileged access platforms is that product and service to the farmer’s data and, most importantly, its bundling is essential to unlocking ability to monitor input purchases or off-take maximal smallholder farmer impact transactions. Digitalised advice and information and maximally attractive economics for supports and de-risks every step of this journey D4Ag intermediaries. by helping smallholder farmers minimise risks of crop loss, improve their financial literacy Service integration, in the highly fragmented and agronomic practices, and understand off- and inefficient market environments that take market needs and quality requirements. characterise smallholder farmer agriculture in Finally, super platforms can also include Africa, can create surprising levels of synergy in digital supply chain management services to terms of doubled or even tripled farmer yields 90 CHAPTER 2 While D4Ag super platforms share value chain aggregation features, emerging models are very diverse. Solutions in this category vary across several different dimensions including player type, the scope of services offered (i.e., number of use cases covered by solution), the depth and sophistication of each service (e.g., light touch farmer information services vs. in-depth precision advisory), the level of human intermediation involved, and the approach to service bundling (i.e., multi- player partnership/consortia vs. integrated super platform solutions that build and deploy all services in-house). The first important dimension to consider CTA and incomes, operational efficiencies, improved is the type of player that is promoting the farmer trust and loyalty, quality control over D4Ag super platform products, as approaches, value chain inputs and outputs, and valuable constraints, and incentives differ substantially data and insights.240 Service bundling can be by actor type. very costly for D4Ag enterprises, but successful D4Ag super platforms ultimately generate the As illustrated in Figure 11, the range of arbitrage opportunities and overall increases players who have built, are building, or may in economic value that single-use-case D4Ag aspire to build D4Ag super platforms in Africa solutions are never able to achieve. Over time, is very wide. this compensates for the incremental costs and complexity of bundled service delivery. On the government side, the two most prominent government-linked platforms are A related insight for super platforms is that the SNS in Rwanda241 and, at an earlier in rural smallholder farmer markets that lack stage, ongoing efforts by ATA in Ethiopia vital infrastructure, particularly agricultural to consolidate national-level digitalisation finance and logistics infrastructure, the initiatives and assets into a more integrated combination of human agents and digital national advisory, market linkages, payments, technologies can meaningfully plug many of and financing platform.242 Globally, another these gaps. D4Ag super platforms do not just example of government-led D4Ag super leapfrog infrastructure gaps; rather, they often platform’s is India’s eNAM platform, which fill them with new and essential physical and several African governments have been human last-mile infrastructure (e.g., market studying with an eye to replication.243 Another and knowledge facilitation agents, input/ non-commercial example of note, this time off-take aggregation points, storage facilities, from a public-private consortium, is the knowledge hubs, and payments hubs). Super Farm to Market Alliance (FtMA), which platforms can deploy and maintain such is building out an ambitious digital platform infrastructure at a reasonable cost due to scale that integrates sophisticated digital (precision) and network effects (i.e., many uses for physical advisory, digitally-enabled input and infrastructure and field agents to ensure off-take market linkages, supply chain high utilisation) and through efficiency gains management, and digital finance (payments, delivered by digital technologies. credit, and insurance).244 CHAPTER 2 91 The second clear cluster of D4Ag super already covers advisory services, payments platform designs are models driven (EcoCash), and agri-insurance. Digitally- by different types of financial service enabled value chain market linkage services are providers. These include several of the very in the product pipeline.248 largest D4Ag solutions in terms of reach that we are tracking across the entire region. Several leaders from the payment space are also pursuing the super platform opportunity. KCB’s large and rapidly expanding Most notably, MasterCard, as part of MobiGrow platform, which already MasterCard’s Lab for Financial Inclusion combines elements of advisory services, in Nairobi, launched an ambitious agriculture market linkages, and payments and credit, value chain digitalisation solution in 2017, is the most prominent example of a initially called 2Kuze, and now operating bank-led super platform, though other in East Africa as MasterCard Farmer banks such as Opportunity International Network (MFN) and in India as e-Rythu249 and Advans are also experimenting Cellulant’s new Agrikore product, a with elements of this model.245 blockchain-based agriculture payment and market linkage digitalisation solution, also For the MNOs, super platforms are also an has great aspirations for scale and super attractive opportunity to tap into agricultural platform features. payment digitalisation revenues and other ancillary revenue streams.246 The best known Another major group of D4Ag solutions example is Safaricom’s Digifarm solution, pursuing the super platform vision are which already features advisory services, credit smaller start-ups in Africa that focus extension, and input-side market linkages, and on digitally-enabled market linkages. is planning to both deepen (e.g., moving to These solutions typically already integrate more precise advisory service) and broaden advisory services, payments, and other value- the range of digital services on the platform.247 added financial services; occasionally they also Econet, via its EcoFarmer D4Ag platform, include logistics and supply chain management Stephanie Malyon, CIAT 92 CHAPTER 2 AFRICA’S SUPER PLATFORM FUTURE? Figure 12 Overview of Alibaba’s Rural Taobao in China Enable rural residents greater Rural life access to a broader variety of digitalisation goods and services and big data infrastructure Rural Taobao In 2019, Rural Taobao service centres are Urban-rural in 1000 counties and 30,000 villages, integration with 60,000 last mile Taobao assistants. Help farmers earn more by €400–500 annual investment by AliBaba. selling agricultural products directly to urban consumers 3-year plan announced in 2018 to establish Entrepreneurial service centres in 150,000 rural villages and employment in 1000 counties, supported by 300,000 opportunities Taobao assistants. This would cover 33% and 25% of the villages in the country. Mechanisation inputs Farm inputs (e.g., tractors, tilling) (e.g., seed, fertiliser, pesticides) Input linkage Payments Alibaba AI & big data Agri-insurance engine for agriculture advice Advisory & information Financial Credit services access products Taobao Urban-rural rural store crowdfunding advisory & offline SHF services group lending Credit Supply chain scoring management Market linkages Supply chain management/ last mile logistics Farm product Farm produce sales (C2C, sales (B2C) livestreams) CHAPTER 2 93 CTA services as part of their interaction with input Another variation of super platforms worth or off-take markets. Several relevant examples noting are ‘in-house’ D4Ag platforms, such include iProcure, Twiga, and Tulaa. as the OFSIS platform that sits at the heart Tulaa, in particular, has already prototyped of the organisation’s digitalisation strategy. an end-to-end D4Ag super platform model in OFIS is now being supplemented by the miniature as part of its market pilots in Kenya. newly launched Olam Digital Origination The Tulaa model incorporates agents but platform, which supports direct digital also digitalisation throughout the value chain, transactions between Olam and its farmers and including digital payment e-wallet, digitally- includes additional features such as traceability, enabled input and off-take market linkages, advice to farmers on yield and quality the provision of digital credit and, finally, optimisation, and payment facilitation.251 digitalised farmer advisory and supply chain and logistics management features. The final potentially paradigm- shifting models worth considering are While digital market linkages are a typical D4Ag super platforms led by global entry point for such models, several players e-commerce leaders. Such platforms are not are exploring a move to a super platform currently in the Sub-Saharan Africa market model from the digital advisory angle. but, given the growing interest of players like MUIIS in Uganda, a solution funded by the Alibaba in Africa, the entry of such models Dutch government and launched by CTA, into the region in the medium to long term started with precision advisory and agri- is well within the realm of possibility and insurance services, but is now moving to could revolutionise the way that African last- integrate more payments, credit, and market mile value chains operate. Alibaba’s Rural linkage elements.250 Similarly, WeFarm, Taobao initiative and business model, which the large Kenya-based peer-to-peer digital is continuing to grow and evolve rapidly in advisory enterprise, is considering pivoting China, shows one logical evolution pathway its model to include digital input and off-take for the D4Ag super platform concept and – marketplace components, as well as linkages independently of whether a player like Alibaba to digital finance. decides to replicate this in Africa – holds 94 CHAPTER 2 CTA many lessons for African D4Ag entrepreneurs, and, on the other, to help Chinese farmers funders, and investors (see Figure 12).252 earn more by selling their products to urban consumers – while also dramatically improving At the core of the Rural Taobao concept farmers’ productivity and encouraging the lies the idea of using a combination of growth of value-add rural enterprises through digital technologies and human networks better linkages to farm inputs, mechanisation, to more closely link China’s farmers and and a full suite of relevant financial products. rural hinterlands to the economic growth engine of urban China and, ultimately, The central market linkage engine of this to global trade networks. The primary model gives farmers opportunities to market entry point for this vision is Alibaba’s rural- and sell their produce directly to urban focused e-commerce strategy, which combines buyers on Taobao, the country’s biggest C2C a rapidly growing network of on-the-ground digital marketplace, either directly or through Rural Taobao Service Centres and agents intermediary food and agriculture enterprises (assistants) with B2C (TMALL.com) and C2C that have Taobao ‘storefronts’. Farmers can (Taobao) e-commerce platforms and other also get linked to markets via agribusiness enabling digitalised logistics (i.e., Cainiao), intermediaries that market their goods on payments (Alipay), and financial services (Ant Tmall, the country’s leading B2C e-commerce Financial) infrastructure, all fully owned by platform. On the input side of the equation, or affiliated with the parent Alibaba Group, farmers can purchase high-quality and lower- China’s biggest company and one of the cost agricultural inputs from a dedicated world’s most valuable brands. Taobao inputs and mechanisation marketplace, with delivery to rural areas facilitated through The vision of Rural Taobao is to use this web Taobao’s rural service centres. of enterprises and digital solutions, on the one hand, to enable rural Chinese access to a From an advisory and farmer information broader variety of modern and low-cost goods services perspective, farmers can receive and services (i.e., agriculture inputs, health, some advice and support from the trained insurance, and modern consumer goods) service centre staff, but also potentially have a CHAPTER 2 95 pathway to accessing digitally-enabled precision advisory services powered by Alibaba’s “Alibaba has invested heavily into Taobao, on Agriculture ET Brain artificial intelligence business. ET Brain currently is only piloting the order of €400–500 million annually. such precision agriculture advisory solutions ” for larger farms, but may extend this to last-mile agents staffing Taobao’s service smallholders in future phases. centres have covered 30,000 villages – a strong foundation for future growth and impact.253 From a logistics and supply chain management perspective, logistics The Taobao super platform model management, traceability, and other related deserves close monitoring by anyone functions are digitised and managed through thinking about the future of the D4Ag proprietary cloud-based software solutions by space in Africa. Despite vast differences in Cainiao, Alibaba’s partly-owned rural logistics cultural and economic context, there are many partner for the Rural Taobao venture. parallels between the Taobao Rural context and the Africa agricultural transformation Finally, for financial access, the entire vision and, more broadly, Africa’s rural network is supported by Alibaba’s payments infrastructure and jobs challenges. One (Alipay) and financial services (Ant Financial) important lesson is likely to be the scale businesses, with targeted third-party of investment required – Alibaba alone is partnerships (e.g., agri-insurance from China investing 10x annually in Taobao what the Insurance), all integrated via a common entire private sector investment community is payments network and data collection and investing in all of Africa’s D4Ag enterprises analytics infrastructure. each year. Another obvious point is the value of fully-integrated and digitised super Alibaba has invested heavily into Taobao, platform models for the African context given on the order of €400–500 million annually the growing (though anecdotal) evidence of since the launch of the venture in 2014; the Taobao Rural’s successes. Finally, the Taobao company is projected to continue a similar case is an important example of the value of pace of investment over the next few years – melding of digital tools, physical infrastructure a good indication of the level of investment and human last-mile networks. Purely digital needed to seriously move rural infrastructure models have their place, but optimal impacts forward. This appears to be yielding strong and economics are unlikely to be achieved results, both in terms of financial viability without using human agents – supported with (e.g., financial service and rural e-commerce digital tools – to facilitate markets, provide revenues) and in terms of scale: Taobao has advice, deliver financial services, and support reached likely over 100 million farmers with last-mile logistics in places where rural new goods, services and finance while 60,000 infrastructure is weak or entirely absent. 96 CHAPTER 3 THE EVOLUTION OF D4AG SOLUTIONS Image to go here Farmerline Led by a handful of strong players, the sector is growing rapidly. D4Ag solutions already reach up to 13% of Africa’s smallholder farmers and generate up to ~€144 million in earned revenue annually, with growing evidence of the sector’s positive impact on smallholder farmers. Fig 13 Sector timeline -- from ICT4Ag to D4Ag World Summit on the Dutch Ministry Release of Information Society World Bank CTA hosts ICT4Ag of Foreign Affairs CTA-Dalberg Forums 1 and 2 releases report International establishes the Geodata ICTforAg Digitalisation recognised the on ICTs in Conference for Agriculture and annual event in of African e-agriculture concept agriculture in Kigali Water Facility Washington DC Agriculture Report 2003- 2005- 2011 2013- 2014- 2016- present present 2019 2007 2011- 2014- 2014- present present present 2019 Web2forDev promotes GSMA starts mAgri Global Open Data USAID forms WB Africa adoption and dissemination partnerships for Agriculture and New Alliance Disruptive Agricultural of low-cost applications between MNOs Nutrition launches ICT Agriculture Technology for development and agriculture Extension Challenge and organisations Challenge Fund Conference Industry growth is dynamic: numbCeHrA PoTEfR 3soluti9o7ns has increased at a 45% CAGR over the last 6 years D4Ag solutions have multiplied in number Figure 14 D4Ag solutions by year of launch Prior to 2010, conversations about number of active solutions, EOY 2018 digitally-enabled agriculture had already begun – primarily among donors and 390 multilateral agencies – but there were very few D4Ag solutions in Africa or 314 globally.254 The few enterprises that did exist were just starting to offer basic solutions like market prices, weather information, and 230 generic agronomic advice using SMS/USSD messages over common feature phones. These 163 early discussions, partnerships, and sector convenings (shown in Figure 13) and the 112 intensive experimentation during the ICT4Ag age helped set the stage for the transition to 66 the D4Ag era over the past 5-10 years. 42 Digital solutions have skyrocketed in Before 2013 2014 2015 2016 2017 2018 number (see Figure 14).255 CTA is tracking 2012 more than 460 solutions; of these, as of February 2019, 390 were active and providing useful services.256 This number is high given that nearly 60% (227 out of 390 active solutions) launched in the last three years, and nearly 20% of the total have launched since Figure 15 D4Ag solutions by primary use case early 2018.257 Moreover, 90% of these solutions number of active solutions, EOY 2018 are being offered by unique enterprises.258 These totals are also conservative: our research 137 likely did not uncover all active solutions in Africa, and we exclude data on hundreds of time-delimited, donor-funded ‘deployments’ 105 and ‘projects’ that have utilised digitally- enabled agriculture services in Africa in recent years but are not stand-alone enterprises or organisations with ongoing operations.259 56 50 Almost two-thirds of the solutions we have tracked report either advisory 33 or market linkage solutions as their primary use case (Figure 15). Advisory 9 services[137] have been popular among donors and private enterprises because of their ease Advisory & Market Financial Supply Data Macro of delivery; unlike other use cases, farmers’ information linkage access chain intermediary agri- receipt of information does not necessarily services management intelligence require coordination with other market actors or institutions – or as deep an understanding of specific local value chains. Market linkage 98 CHAPTER 3 solutions (105), though more difficult to and, most notably, access to large and develop and implement given the higher level relatively low-cost datasets and advanced of investment typically required per farmer data analytics (e.g., machine learning, AI), reached, have also begun to grow in number. which were either not available or too costly until just a few years ago.261 As underlying The financial access, supply chain technologies have matured and spread, supply management, and macro agri- chain management and macro agri-intelligence intelligence use cases are at an earlier solutions have grown in number and scale. stage but are developing rapidly. We expect this trend to continue. Financial access solutions (56) are typically complex, requiring the collaboration and partnership of multiple actors (e.g., banks and Digital farmer registration mobile network operators) and often building figures are growing rapidly on the existence of key enablers, particularly We estimate that the number of farmers mobile money.260 Supply chain management registered for D4Ag solutions in Sub-Saharan solutions (50) typically require relatively large Africa has grown at roughly 44% per year enterprise-quality software investments in order over the past three years, and likely in to be considered by agribusiness users, are the range of 50–60% CAGR over the subject to network effects, and require large past eight years, to reach a total of numbers of clients to ensure viability – all of 33 million smallholder farmers as of which limit the number of deployments. Macro the end of 2018.262, 263 The definition of agri-intelligence solutions (9), meanwhile, registered users requires clarification require a more advanced D4Ag infrastructure (Figure 16). Figure 16 Sizing the number of registered D4Ag users – methodology considerations Individuals vs households This estimate counts users that may be households or individuals using the same device. D uplication This estimate includes duplicated users (e.g., one farmer registered for multiple D4Ag solutions). We later apply a 20% haircut to account for this. A pples and oranges This estimate includes users of passive solutions (savings accounts) and active solutions (market linkage apps) though use has different implications in these cases. CHAPTER 3 99 Depending on numbers used to size the overall The definition of ‘registration’ depends on the smallholder farmer population (i.e., individual type of solution – a farmer may be registered farmers and pastoralists vs farm households), for an MNO simply by providing a name this figure represents 13% of all smallholder and phone number or texting a short code to farmers in Sub-Saharan Africa and up to 45% register for the service, while registering for a of all smallholder farmer households in the government-provided solution might involve region (the highest end of the range assumes sharing census-level details. The overall figure only one user per household). The number of 33 million does not, however, account includes farmers that have either registered for the possibility that some proportion of themselves for D4Ag solutions, have been registered farmers have been registered registered by agents, or have been registered with more than one D4Ag solution, and are through an enterprise (typically an agribusiness, therefore doubly counted.266 cooperative, or financial institution) that uses a D4Ag solution provider to reach and D4Ag enterprises (including both manage relationships with smallholder farmers commercial and non-profit) account for in its value chain.264 These numbers do not the majority of registered smallholder include registrations of non-farmer end-users farmers. Commercial enterprises and like exTtheonusigonh asgmenatlsl, igno nveurmnmbeenrt, egnodv-eursenrms ent and NMGNOOs collectively reach approximately 60% of decdiseiopnlo tyomolse, natnsd d eenmteorpnrsitsrea ctleie nsitgs.n Sifuiccha nt reach(a maximum of 20 million) of registered actors are also users of D4Ag solutions buSt mwae llholdesrms arlelhgoilsdteerr efadr mbeyr sD (4seAe gF isgoulruet i1o7n).s T, bhiys solution type do not include them here as we were not able number includes financial service providers to capture them reliably (and they were not (FSPs), which currently reach ~5.5 million our focus segment).265 farmers through digitally-enabled insurance, Figure 17 D4Ag solutions and registered users, by type of actor number of solutions and millions of smallholders, EOY 2018 1.5% 390 33M 2% 1.5% ~1% 7% 20% Government deployment 15% Agribusiness 20% MNO deployment NGO ~5% Commercial enterprise 75% 54% # of solutions Registered users 100 CHAPTER 3 MNO D4Ag solution savings, and credit solutions. As mobile part, this relatively low number is due to the example payments become mainstream in many fact that most agribusiness solutions target Safaricom is the largest MNO countries, recent growth in FSP activity is smallholder farmers in tight value chains – and in Kenya and has been a clear likely to continue.267 such farmers represent a maximum of 7% pioneer in the D4Ag space. Safaricom’s M-Pesa, one of the of smallholder households.271 Input dealers first mobile money platforms, Mobile network operators (MNOs) and like Yara and Syngenta and mechanisation has been a critical enabler of D4Ag in Kenya – numerous governments each account for roughly players like John Deere are also active in the enterprises that rely on M-Pesa 20% of registered smallholder farmers. D4Ag sector; for the most part, these players, to operate might very well MNOs have at least 6.5 million smallholder as well as big buyers and processors like not exist without Safaricom’s leadership. Safaricom has also farmers registered to their D4Ag solutions, Barry Callebaut, have partnered with other rolled out a suite of financial typically advisory and information services organisations to digitise their farmers or use and information services for smallholder farmers delivered with other partner organisations. third-party supply chain management solutions, through DigiFarm, which Six major MNOs across the continent so we capture their potential reach within offers “discounted products, currently offer a total of approximately the D4Ag enterprises category (e.g., Barry customised information on farming best practices and 15 D4Ag solutions.268 Governments similarly Callebaut’s reach counted as part of the SAP access to credit and other reach about 6.5 million farmers through their Rural Sourcing Management Platform). financial facilities.” Since its launch in 2017, DigiFarm has own D4Ag solutions. rapidly expanded throughout ‘Engaged’ and ‘active’ the country, reaching a Large agribusinesses such as Olam,269 reported 950,000 users, users make up a minority 20% of whom are active. Cargill, Mars, and ETG likely reach no more than 500,000 African of registered users smallholder farmers with proprietary We estimate that 42% of registered (in-house) digitally-enabled supply smallholder farmers have engaged with chain management solutions.270 In large D4Ag solutions to some extent – in the Figure 18 MNO D4Ag solutions in Africa (EOY 2018) Niger Mali Orange Labaroun Kassoua Orange Sandji Orange Senekela Orange Garbal Somaliland Senegal Telesom M-Dalag Orange Mlouma Burkina Faso Orange 3-2-1 Uganda Airtel 3-2-1 Côte d'Ivoire Rwanda Orange mAgri MTN 3-2-1 Ghana Vodafone Farmer's Club Nigeria Kenya MTN Iska Weather Airtel 3-2-1 Safaricom Digifarm Tanzania DRC Vodacom 3-2-1 Madagascar Vodacom 3-2-1 (42502) Orange M-Kajy Orange M'Vola Orange HayVokra Zambia Orange Bazar.mada MTN 3-2-1 (6-6-7) Airtel 3-2-1 Malawi Botswana Orange mAgri Mozambique Airtel M’chikumbe Vodacom 3-2-1 South Africa Vodafone Connected Farmer Zimbabwe Econet EcoFarmer CHAPTER 3 101 Figure 19 Definitions of D4Ag user types Agribusiness examples Igara Tea Growers Registered users Engaged users Active users Factory (IGTF) and CTA, in partnership with the consulting firm Environmental Surveys, Information, Planning and Policy (ESIPPS), built a spatial data management system. The digital profiling of tea farmers involved compiling 33m 42% 15–30% geo-referenced information about them and their land The number of accounts The proportion of The percentage of accounts using GPS-enabled tablets. The in the database; users that know how used regularly enough for data are stored and spatially the most frequently to use solutions and users to feel the full benefit cited number have done so of the solution analysed by an online system. The profile database is linked to a financial and accounting system, allowing smallholder- owned IGTF to build track case of most solutions, at least once from one solution to another, lack of standard records of transactions with member farmers. The system monthly in the past year.272 We propose active use definitions even within specific use can thus serve as a basis engaged users as a new term to filter out cases, and the varying levels of use necessary to for fertiliser distribution and tracking. IGTF has benefitted farmers who are registered for digital solutions realise the benefits of a solution (for example, tremendously from this digital but do not use them, while also acknowledging market prices might need to be accessed daily solution, which is currently that truly active users are currently difficult or while planting guidance might only be helpful being scaled up at the national level. impossible to measure in a report of this type annually). The lack of a meaningful definition due to the lack of consistency in definitions of ‘active use’ or sufficiently precise data to In 2014, Olam International, one of the and the absence of comparable data. The give it parameters suggests a need for further world’s largest suppliers of estimate of engaged users is based on CTA- data collection, analysis, and study on the cocoa beans and products Dalberg survey data and augmented with desk part of enterprises, donors, and others. In like palm oil, coffee beans, cotton and rice, developed research and interviews with implementers of any case, based on interviews and data from an in-house Olam Farmer large D4Ag solutions that did not respond to those studies that have attempted to measure Information System (OFIS) – a digital supply chain the survey – but it should not be interpreted different levels of farmer activity for specific management and advisory as suggesting that the farmer is necessarily the D4Ag solutions (e.g., GSMA’s MNO mAgri solution. Since then, it has direct user. ‘Engaged’ in this case might mean case studies), it is clear that the level of truly been refining the solution and scaling it across its ecosystem that each farmer included in this estimate active use is in many cases far below engaged of smallholder farmers. has used a D4Ag solution, but it could also use – e.g., active ‘power users’ accounted for Sub-Saharan African farmers constitute a large share of mean that someone – such as an agent – has just a third to half of engaged use levels in the the 160,000 smallholder helped the farmer use the application, or used case of many MNO solutions, leading to registered for OFIS and Olam it on the farmer’s behalf. Our database tracks our provisional estimate of active users being expects to digitise all 500,000 of its farmers globally by 9.5 million ‘engaged’ users of D4Ag services, 15–30% of registered users, cumulatively, 2020. Beyond OFIS, in 2019, which constitutes 42% of the registered users across all use cases.273 Olam announced the launch of their Digital Origination (23 million) for whom an estimate of engaged platform, which supports direct users is available. Extrapolating to the broader Registrations are digital transactions between D4Ag population suggests that there were Olam and its farmers and ~14 million ‘engaged’ farmers utilising D4Ag highly concentrated includes additional features such as traceability, advice to solutions in 2018. D4Ag registrations of smallholder farmers on yield and quality farmers are highly concentrated by use optimisation and payment facilitation. Estimating ‘active use’ – use of a digital case, actors, and geography. Advisory solution frequently enough to obtain or services account for over two-thirds of registered even maximise its target benefits – is farmers today, the top 20 players reach more at this stage impossible, as noted above, than 80% of registered farmers, and nearly 70% due to the varying definitions of active use of all registered farmers are in East Africa. 102 CHAPTER 3 Figure 20 Smallholder registrations, by primary use case millions of registered farmers, EOY 2018 2.4M 33.1M 2.5M 7% 5.6M 8% Supply chain management Market 17% linkage 22.6M Financial access Note: This count excludes solutions 68% with indirect reach – such as FarmRadio and Agribusiness TV, which reach tens of millions Registrations are concentrated in of farmers. Also excluded are advisory and information; other farmers registered by business to use cases are still nascent. business solutions if those farmers are already counted as part of the user base of a farmer-facing D4Ag solution provider (e.g., Arifu registered users who are Advisory Total also counted as part of Safaricom services Digifarm’s solution). Note: This count excludes solutions with indirect reach, such as FarmRadio and Agribusiness TV which reach tens of millions of farmers as well as government-to-farmer or business-to-farmer digital payment solutions Source: Dalberg analysis. [1] Dalberg analysis, Lowder, et al, 2016. D espite significant straddle 4 or more D4Ag use cases (Figure bundling in the sector 21). In the earliest stages of D4Ag, advisory services were easiest to deliver using common today, registrations are technologies. In particular, the evolution heavily concentrated of SMS/USSD-enabled enterprises to offer among advisory services generalised information on feature phones without a need for supporting systems. Advisory services account for over two- However, the value generated by such advisory thirds of registered farmers today.274 services remained low. In recent years, D4Ag This is consistent with the distribution of use enterprises have looked to combine other use cases by number of solutions. (see Figure 20) cases, including market linkages, with their As discussed above, this concentration is in advisory offerings. More broadly, as noted in large part because advisory solutions tend to Chapter 2, D4Ag enterprises are increasingly be easier to scale. Other areas remain more moving toward ‘super platform’ business nascent as they require greater feedback models that combine market linkage, advisory from and tailoring to farmers, more complex services, and financial services, and often also operational logistics and, in many cases (e.g., have supply chain management and macro for market linkages), the integration of human agri-intelligence features. Farmers tend to see agent networks, which hampers reach. more immediate returns from these services, which increases farmer uptake and willingness D4Ag enterprises are increasingly to pay. Moreover, bundling use cases offers bundling services across mutiple use farmers services they need more holistically, cases into their solutions. Today, more enabling greater choice, and drives operational than half of surveyed enterprises offer services synergies across different solutions. across multiple use cases and ~9% of solutions CHAPTER 3 103 Figure 21 D4Ag solutions by number of use cases offered number of solutions, EOY 2018 183 207 (53%) solutions offer bundled use cases 108 65 26 Note: Number of use cases offered 8 represents the number of different kinds of services an enterprise could provide users. Bundling presents 1 use case 2 use cases 3 use cases 4 use cases 5 use cases a unique business model and may result in higher/lower levels of reach, revenues, use, impact, etc. Registrations are heavily in reach, several have now achieved concentrated among a meaningful scale. Around 75% of enterprises reach fewer than 100,000 farmers relatively small number and nearly 30% of enterprises reach fewer of players than 1,000 farmers. Yet, 16 commercial and The 20 largest D4Ag solutions account non-profit enterprises now have more than for nearly 80% of all registrations (see half a million users (see Figure 22). In the Figure 21). The scalability of a solution financial access use category, for instance, depends on a number of factors, including the the top players are highly concentrated: enabling environment, market size, revenue ACRE’s Agricultural Loan Cover (1.7 million model, and value-add, as well as the type of smallholder farmers), Bank of Kigali (1.5 solution sponsor (financial service enterprise, million) FarmDrive (1 million farmers), and MNO, government, agribusiness). MNOs, for Pula (600,000 farmers). example, may already have direct access to farmers through large agent networks, which Three out of the six MNO players with likely accounts for the disproportionate number D4Ag solutions that we are tracking have of users registered with this solution type. collectively registered 5 million farmers Other types of players may face the more and account for nearly 80% of farmers expensive and time-consuming prospect of reached by MNOs. Viamo has millions of having to build out their own agent networks registered users through its 3-2-1 product, in order to reach individual farmers. Orange reaches at least 1.2 million farmers across ten different D4Ag solutions, and Although most commercial and non- EcoNet in Zimbabwe has 1 million registered profit D4Ag enterprises are quite small users through EcoFarmer. By the time this 104 CHAPTER 3 Figure 22 Top 20 solutions, by number of registered users Rank Solutions Registered users Primary use case 1 Ethiopia 80-28 hotline 4,000,000 Advisory services 2 Viamo 3-2-1 (multiple solutions) >3,000,000 1 Advisory services 3 TCS InteGra 2,000,000 Advisory services 4 n-Frnds >2,000,000 1 Advisory services Esoko Digital Farmer Service 5 ACRE Africa 1,700,000 Financial inclusion 6 Bank of Kigali/TecHouse 1,500,000 Financial inclusion 7 WeFarm 1,400,000 Advisory services 8 Orange (multiple solutions) >1,300,000 Advisory services 9 ZIAMIS 1,150,000 Advisory services 10 Esoko Digital Farmer Service 1,000,000 Advisory services iCow 11 Econet EcoFarmer 1,000,000 Advisory services 12 Safaricom DigiFarm 950,000 Market linkage 13 Arifu 900,000 2 Advisory services 14 iCow 821,800 Advisory services 15 Pula 611,000 Financial inclusion 16 Digital Green 500,000 Advisory services 17 Agroforce/Virtual City 500,000 Supply chain management 1 Estimated number of registered users that access agriculture content; many more registered 18 Waterwatch Cooperative 500,000 Advisory services users for these solutions overall in Africa (~15 million for n-Frnds and 19 RATIN 400,000 Advisory services ~10 million for Viamo). 2 Large share double-counted with 20 KCB MobiGrow 380,000 Market linkage Safaricom Digifarm. report is published, Safaricom’s DigiFarm rolling out a suite of financial services for solution is likely to reach more than 1 million farmers in addition to the advisory services farmers, as well. it already provides. There are likely other government deployments in the works, but our Government reach in Sub-Saharan research did not come across them. Africa is almost entirely through three solutions: Ethiopia’s 80-28 advisory As discussed above, our agribusiness service (4 million farmers), ZIAMIS reach estimates are derived primarily in Zambia (1.15 million), and Bank of from large agribusinesses (e.g., Olam, Kigali’s SNS solutions, deployed in Cargill, Twiga, SAT4Farming). Olam likely partnership with Rwanda’s Agriculture has reached the largest number of farmers to Board (1.5 million). The latter is currently date – the company claims that it has already CHAPTER 3 105 Figure 23 D4Ag solutions by number of registered farmers The top 20 solutions, each with more than ~400k registered users, account for 78% of total reach millions 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 Solutions (ranked ordered by size from 1 to 390) Registrations are concentrated by geography; the majority are in East Africa. compiled data on 160,000 cocoa farmers. It is many emerging solutions but the number likely that the concentration of D4Ag solutions of registered users remains low (3 million among major agribusinesses may simply mirror registered by solutions active in West Africa). the concentration of major agribusinesses in In contrast, Southern Africa’s user base is high the agribusinesses sector more generally. (5.8 million) but the number of solutions is limited. Central Africa falls far behind on both Registrations are fronts (see Figure 25). We will discuss regional variation further in Chapter 5. concentrated by geography; the majority Most companies are now are in East Africa generating some revenues While D4Ag enterprises are present in Increasingly, enterprises have been able to nearly every country in Africa, D4Ag has generate at least some revenue (Figure 26). not developed evenly across Sub-Saharan Based on the CTA-Dalberg survey data, of the Africa. D4Ag enterprises operate in 43 out 175 respondents, an estimated 70% of African of 49 countries in Africa,275 but while at least D4Ag solutions generated some earned revenue 17 countries have more than 20 enterprises, – a number lower than the likely 80%+ of 25 countries have fewer than five. East Africa D4Ag enterprises in Africa that are revenue- dominates both in terms of registered users seeking.276 The remaining organisations were (over 20 million) and a high number of active either entirely donor- or government-funded D4Ag solutions, whereas West Africa has entities or were very-early-stage start-ups that 106 CHAPTER 3 Figure 24 Regional breakdown of D4Ag solutions solutions and registered users (millions) by sub-region of HQ and sub-region of primary focus, Registered users EOY 2018) Solutions by primary region Western Africa Solutions Users Eastern Africa HQ: 145 3.1M Focus: 162 4.3M Solutions Users HQ: 124 21.0M Focus: 146 21.8M 27 solutions headquartered in the G5 Sahel, account for 573k users. Another 33 solutions have users in the region. Central Africa Solutions Users HQ: 18 0.60M Focus: 20 0.85M Southern Africa Solutions Users HQ: 43 3.9M Focus: 46 5.8M Figure 25 Most D4Ag enterprises are now generating some earned revenues number of survey respondents by use case 82 Never generate revenue 7 75 4 3 Will generate revenue 4 Generating revenue 71 39 68 2 31 4 1 29 1 2 2 33 29 25 Financial Advisory Market Supply Data access services linkages chain systems CHAPTER 3 107 did not yet report revenue streams. In a few reported D4Ag enterprise revenues, expressed Revenue generation cases, non-revenue-earning solutions were in annual revenues per registered farmer, tend example – N-Frnds in-house (i.e. non-monetised) digital platforms to be highest for market linkage enterprises. In Rwanda, N-Frnds from agribusinesses or MNO solutions that Aggregating across survey, desk research, and understands the tremendous value of data for both banks derived value indirectly without revenues (e.g., interview data, and rounding for convenience, and farmers and has built a ‘free’ farmer information services that generate we see ~€25 average revenues for market viable business model around it. The company leverages the value through improved customer retention linkage solutions per registered farmer annually data it records on transactions and stickiness but do not generate direct D4Ag (€3–45 range), in comparison to ~€5 for between farmers and off-takers revenues). Around 80% of the revenue-generating advisory and information services (€1–9 range), to link farmers to banks to facilitate lending opportunities. enterprises had several revenue streams.277 ~€4 for digital financial services (€0.5–7 Smallholder farmers pay range, and ~€4 for supply chain management nothing for the service; instead, N-Frnds charges Of the revenue-generating firms in the survey solutions (€1–7 range).282 banks a small acquisition fee sample, 26% reported running profitable for every loan extended to and sustainable businesses that could survive As the sector pivots to a greater focus on N-Frnds’ network of farmers. In this way, N-Frnds’ business without donor long-term subsidies, a figure market linkage (or rather market linkages model targets businesses that that is within range of earlier D4Ag sector bundled with other services) from solutions are able and willing to pay for these data, as opposed to overviews.278 Most D4Ag enterprises are thus focused more on advisory services – something farmers who would be unlikely largely supported by grants and still have a that we heard universally in our interviews but to use the service if they had way to go before they are sustainable and are unable to demonstrate empirically in the to pay for it. scalable. This profitability number may absence of comparable historical data – one seem disappointing but is not unexpected. would expect that average sector revenues Only 40% of the commercial enterprises in would rise quickly. the CTA-Dalberg databases have been in operation longer than three years, which is A small but growing number of players have often seen as a reasonable benchmark for already started developing business models time to profitability for tech start-ups and, that can generate up to €90 in annual per more broadly, new small and medium-sized farmer revenue. Achieving these types of businesses.279 This share of profitability among revenues requires multiple revenue streams and start-up enterprises is also in line with early- extensive product bundling, i.e., characteristics stage start-up investor expectations in Africa.280 of emerging D4Ag ‘super platform’ models. To generate such economics D4Ag actors Sector economics are improving and the must essentially become active agriculture share (and number) of profitable enterprises value chain participants, taking a share of both is growing. While there are no baseline data agricultural input costs and off-take value as with which we can make a comparison, compensation for their digital intermediation. anecdotal evidence from interviews suggests This approach can work well when D4Ag that these results are significantly higher than solutions are able to successfully consolidate what was common even a few years ago in fragmented value chains by removing other terms of the share of D4Ag solutions that are intermediaries (e.g., digitally linking farmers to profitable. Extrapolating to the overall sector, retailers for the post-harvest sale in ways that even assuming very high levels of new business bypass last-mile village agents and traders), failure (e.g., 50–75% failure rate over three reducing value chain ‘leakage’ (e.g., using years), these numbers suggest that the number digitised logistics and just-in-time market of profitable and thus potentially investable linkage to significantly reduce post-harvest D4Ag actors could double from ~75 D4Ag losses) or, in an ideal state, capturing both solutions today to over 150 in 2021.281 of these effects. The substantial surplus value created can then be shared in ways that There is also a clear trend of rising annual leave both the farmer client and the D4Ag D4Ag enterprise revenue per farmer. Self- intermediary with dramatically improved 108 CHAPTER 3 economics. The D4Ag enterprise can benefits (and even if they do realise benefits, then further supplement such revenues they may not attribute the benefits to the with ancillary revenue streams such as advisory service). There are signs of emerging inancial services fees/interest or even data willingness on the part of farmers to pay for monetisation revenues. market linkage solutions where results are more immediate. Overall, while 70% of revenue- While the cost structure for generating these generating enterprises have user payment revenues varies dramatically depending on revenue streams, user payments do not appear solution type, there is evidence that some to constitute the majority of their revenue. companies are able to achieve 30–40% gross margins. We certainly do not expect all Because of the challenges of generating businesses to achieve this level of revenue revenue from farmers, organisations or margin, but the data indicate that with have oriented themselves to generate extensive revenue bundling, strong economics their revenues from other businesses, are achievable. This is already a major leap even if the final service is to the farmer. forward from a time when D4Ag solutions Such B2B payment models allow for a range of centred on advisory services, as that model payment streams from players with greater is marked by low per-farmer revenues and ability and willingness to pay than the typically razor thin margins. smallholder farmer. These models include monetising data and fee for service. FSPs often Important business model shifts account partner with banks and other FSPs rather than for the high share of revenue-generating work directly with farmers, while supply chain enterprises in the D4Ag space. By and management enterprises partner with larger large, digital service providers have learned agribusinesses. For example, Tulaa relies on that farmers will rarely pay for digital products commissions from farmer market linkages and and services – and especially advisory services, related transactions. Farmforce, meanwhile, where it can take time for farmers to realise enables off-takers (processors or agribusinesses) Figure 26 Estimated total addressable market calculations Solutions Addressable Annual revenue per user Total addressable market (million) farmers (million) (min) (max) (min) (max) Advisory services 250 1 €1.00 €9.00 €250 €2,250 Financial access 73 2 €3.00 €14.00 €219 €1,022 Market linkage 73 2 €3.00 €50.00 €219 €3,650 Supply chain management 73 2 €0.50 €9.00 €37 €657 Total (assuming no digital constraints) €725 €7,579 Total factoring in connectivity constraints Conservative scenario: (39% of smallholder farmers have mobile subcriptions)3 €283 €2,956 Less conservative scenario: (70% of smallholder farmers have access to phone in household)4 €507 €5,305 Notes: 1 Assumes that every smallholder farmer is part of addressable market for advisory services subscriptions (i.e., possible to have multiple subscribers from family for one farm) 2 Assumes that farms or households are a relevant unit for market sizing as multiple subscrption for the same product unlikely or impossible 3 Sub-Saharan Africa farmers with unique subscriptions in 2018 (~39%, estimated based on 44% unique subscriber rate in the region and 1.3 ratio of urban to rural connections based on GSMA data) 4 Sub-Saharan Africa farmer households owning at least one phone (~70%, estimated based on smallholder farmer survey data from sources such as CGAP smallholder diaries) CHAPTER 3 109 Figure 27 Known and estimated earned revenue by primary use case revenues by use case ~€127M average €, EOY 2018 (€110–145M range) 3 3 9 ~€107M 3M 21 3M Advisory & information services 6M Market linkage 18M 39 Financial access Supply chain management 26M Data intermediary Macro agri-intelligence 54 51M Tracked revenue Estimated sector revenue to access, monitor, and manage a large number conservative estimates (smallholder farmers of farmers for a fee paid by the off-taker. with unique mobile phone subscriptions) to less constrained estimates (households owning at Overall, we estimate that the total least one mobile phone) in order to arrive at a addressable market (TAM) is a directional estimate. maximum of €5.3 billion, depending on key assumptions around the number of At the lowest end, the TAM is somewhere addressable farmers, average revenue per between ~€0.3–0.5 billion. These figures user (ARPU) by use case (see Figure 26 and apply the lowest end of ARPU for each use additional information on these calculations case. They are likely to underestimate the in Annex 3: Methodology) and constraints TAM because the ARPUs underlying this around smallholder farmer connectivity. 283, calculation are likely more representative of 284, 285 These ranges are wide primarily for the lowest performers in the market, rather two reasons. First, the ARPU by use case than an average. At the highest end, the varies significantly: individual enterprises TAM is approximately €5.3 billion, assuming within a use case have widely varying business the highest ARPUs for individual use cases models and few reliable examples with data as well as limited constraints around phone points exist today. In our estimates we have ownership (i.e., if a smallholder family owns therefore applied the highest and lowest ranges at least one phone, family members are able based on available estimates from enterprises to use D4Ag services and are therefore part themselves. Second, there are no reliable of the addressable market). These figures are estimates of smallholder farmer ownership of likely to be overestimates; only a handful of mobile phones, and there are multiple ways companies are achieving the highest end of to arrive at such a figure. As with ARPU, ARPUs (though in a few cases like market we similarly applied a range of the most linkages, there are examples of companies 110 CHAPTER 3 Figure 28 Quantifying impact – a directional view based on limited data Smallholder farmers Income Productivity Digital Bundled D4Ag models advisory services 30% 23% Productivity (10-70%) (0-75%) Income Digital market linkages 37% 73% 57% (15-100%) (5-300%) (20-100%) 168% (50-300%) Digital financial Examples of bundled models with services 18% 38% self-reported data (e.g., Zenvus, (16-20%) (25-50%) MyAgro, Kituvo, Tulaa, SunCulture) *Note: Yield and impact data across ~50 data points cited in literature or captured in USAID ICT4Ag impact database *Note: Yield and impact data across ~50 data points cited in literature or captured in USAID ICT4Ag impact database Source: USAID Impact Database, BMGF Impact Analysis, Dalberg analysis Source: USAID Impact Database, BMGF Impact Analysis, Dalberg analysis outperforming our current range) and success in terms of revenue – and while D4Ag household ownership of a phone is likely solutions do reduce agribusinesses’ costs and/ not fully indicative of ability to access D4Ag or increase their revenues, these benefits do not services. Still, these figures provide useful come from user payments – and agribusiness bounds and suggest that the likely TAM is data are difficult to access publicly. Taking the somewhere between the midpoints of the midpoint of the revenue range (€140 million) conservative and less constrained estimates (i.e., and the midpoint of total addressable market €1.6 billion and €2.9 billion). As the sector (€2.3 midpoint estimate, €1.6–2.9 billion range evolves and more data points emerge, the depending on which constraints to connectivity range of estimated values for the addressable one assumes), we estimate that market market will likely become narrower and more penetration today is 6% (between 4–8%). precise. Evidence of results is The D4Ag sector likely generated emerging though much about €110–145 million in revenue in 2018,286 a small fraction (6%) of the total more is needed addressable market.287 This figure includes Evidence of D4Ag impact is currently commercial enterprises (including financial limited. Only a few market leaders currently service providers), NGOs, and MNOs, but systematically track the impact of their work. excludes governments and agribusinesses. Among those that do, there is little agreement We do not include governments because on metrics or methodologies, so comparisons they typically do not charge users or frame are difficult to make across solutions and CHAPTER 3 111 Figure 29 Impact across the smallholder value chain – USAID’s view on the evidence Digitizing the agricultural value chain I WHY PLANNING INPUTS ON-FARM POST-HARVEST ACCESS TO PRODUCTION Storage I Processing I Transport MARKETS • Help farmers plan • Reduce • Help extension • Improve links • Increase • Reduce costs • Increase ability of what, when to plant counterfeits services reach between farmer of transport smallholder farmers more farmers farmers, negotiating to sell to larger • Tighten relationship • Reduce costs and processors power by • Increase markets by allowing with buyers, risks for buyers • Provide timely providing choice of buyers to track crops processors reminders/alerts • Reduce post • Increase access to market prices different types to source harvest loss of transport • Adapt to climate quality inputs • Use behavior (certification and with • Track for farmers change change media to provenance) • Enable sellers to digitally-enabl provenance • Increase promote best • Provide data for know demand in ed harvest for supply access to • Increase market practices among farmers to make advance loans and chain timely information available farmers business decisions on digitally optimization information so to farmers so that cash flow and • Provide • Increase precision warehouse and grading that farmers they have more maximizing profit convenient and and/or receipts know if and choices secure ways for adaptability of when farmers to • Inform harvest farming transport is purchase, save, practices to interventions and arriving and receive credit reduce post crop choices inputs harvest losses. through applied data • Monitor storage conditions <--------------------- USING CONNECTED DIGITAL TOOLS TO BETTER INTEGRATE THE ENTIRE MARKET SYSTEM ---------------→ Source: USAID. 2018. ‘How digital tools impact the value chain.’ aggregate impacts are difficult to arrive at. labourers needed. We define impact on income Finally, efforts at tracking impact data not as D4Ag increasing farmers’ incomes. Impact only vary in design and focus, but also vary on income depends on ease of access to well- considerably in robustness. For at least half of priced inputs, fair prices from off-takers, and the 40-50 impact data points collected for this other factors. Increased income improves report across different African D4Ag solutions, quality of life for farmers and their families and the evidence we are left to draw on is based helps establish food and nutrition security. on fairly small samples and/or is self-reported by enterprises. Only a handful of players have Although conclusive evidence has yet to applied randomised controlled trials (RCTs) or emerge, some providers have shown that quasi-experimental evaluation methodologies to D4Ag can impact the productivity and measure impact. income of smallholder farmers. According to self-reported data as well as randomised Productivity and income are the most control trials and other impact studies universally understood aspects of impact conducted by D4Ag enterprises, the degree and – but are just two among a wider range of range of impact differs significantly depending impact types discussed later in this chapter. on use case. Advisory services (10–70% income We define impact on productivity as D4Ag increase, 0–75% yield increase) and financial increasing farmers’ yields, or crop produced access (16–20% income increase, 25–50% per hectare of land. Higher productivity can yield increase) tend to have lower impact on drive increased revenues, commercialisation, incomes and yields than do market linkages and reduction in the number of agricultural (15–100% income increase, 5–300% yield 112 CHAPTER 3 increase) (see Figure 29). It is important to note services and FSPs.288 We believe that financial that the sample sizes used to determine these access solutions with different business models ranges and the average indicated are small; and structures would vary in terms of impact they should be construed as an indication of as well, but data on impact metrics are too what is possible, not a definitive representation limited to reach more specific conclusions. of the space or use case. Bundling services appears to create These providers are likely the outliers; the more impact. A handful of enterprises that companies that are tracking impact may bring together use cases report very high represent best-case scenarios – and even in impact numbers (20–100% income impact, these cases it is still difficult to attribute that 50–300% yield impact). This suggests that, impact to D4Ag solutions alone (as opposed to, when structured well, combining offerings for example, a particularly strong business case, across use cases could have an additive impact or other aspects of the solution). Anecdotally, on users. these figures are higher than those of purely analogue solutions and are generated at The impact story is far from complete, reduced cost. but the information we do have is encouraging. Robust evaluations and The span of these ranges indicates that, trustworthy impact metrics are hard to find depending on their business model, across the D4Ag space. The sector requires even solutions within a single use case significant investment in capturing impact data can vary significantly in the value they if we are to better understand successes and offer farmers. Within advisory services, failures to date and in the future. Of critical higher-end solutions are more precise and importance will be user-centric research and participatory, but there are insufficient data design; in-depth case studies of both successful points to parse out what balance of precision and less successful actors; better evidence of and cost is optimal for farmers. Market linkage the on-the-ground impact of different use solutions that integrate farmers with input cases and business models, using standardised providers and off-takers, often using agents and and rigorous impact metrics; and a better intermediaries, appear to have even greater understanding of the specific contributions of impacts on yields and income than do advisory digital vs other business model enablers. CTA Robust evidence is particularly critical as the number of players in the sector explodes and enterprises begin to move from pilot phase to scale – a point at which it is notoriously challenging to maintain strong impact. For whatever impact measurement data do exist, far too little gets captured and published. The CGIAR Big Data in Agriculture initiative has recently launched a process to start to collect this data from the sector, something we believe is overdue and essential for moving the knowledge agenda forward. We will discuss these impact-related challenges and subsequent recommendations further in Chapters 5 and 6. CHAPTER 3 113 D4Ag’s impacts matter not only for individual smallholder farmers, but also “ for other agricultural actors – and many Bundling services appears to create more impact. A agribusinesses are already realising handful of enterprises that bring together use cases value from D4Ag solutions. The benefits report very high impact numbers (20–100% income to the broader ecosystem may have a number impact, 50–300% yield impact). This suggests that, of indirect positive impacts on smallholder farmers. Digitalisation allows companies to when structured well, combining offerings across use better understand farmers in their value chains cases could have an additive impact on users. (e.g., profiling, monitoring farmer activities) ” and thus offer them better, more tailored products and services. D4Ag improves internal process efficiencies, as well, by enabling better market aggregation and coordination – thus Digital solutions are also helping cutting costs. When information across multiple governments make more informed farmers and catchment areas is combined, decisions and are supporting agricultural companies can know the quantity, quality, and planning, albeit more slowly than location of the produce available and better for agribusinesses. As an example of manage volume fluctuations for their own use how governments are beginning to make and/or as they take the product to market. use of these data, Ethiopia’s Agricultural Digitally-enabled coordination and supply Transformation Agency (ATA) has used chain management also reduce the number of its highly popular 80-28 system, as well as agents needed on the ground, which cuts costs. e-vouchers, to support smallholder farmers For example, in Southern Africa, large fertiliser while building robust datasets of them, their companies have begun to use predictive needs, and government priorities to address weather data to project farmers’ likely yields, those needs. In Rwanda, the government informing decisions about how much fertiliser has leveraged digital solutions to consolidate to provide on credit. In Rwanda, government- farming activities, facilitating big-picture led consolidation of localised farming activities, decisions around commodity pricing, storing, driven by advisory service solutions, has and crop input supply. While these examples underpinned improved efficiency for off-takers are promising, the potential for digital solutions and price leverage for producers. to support macro-level decision making is still largely under-tapped (and completely Recognising the strong potential gains in un-tapped in many countries); we discuss this D4Ag, some agribusinesses have started further in Chapter 4. In the meantime, it to invest in building out capabilities is also important to consider the impacts of in-house, which would allow them to reduce D4Ag for youth, climate change vulnerability, costs and own their valuable proprietary data. employment, and women as part of the overall For example, Twiga Foods has embedded impact story. The next section takes a closer its entire value chain with digital solutions to look at the impacts of D4Ag through these create an entirely cashless network. lenses. 114 CHAPTER 3 A DEEPER DIVE INTO HUMAN IMPACTS OF D4AG For a number of reasons, D4Ag will not necessarily have the same impact on all segments of the agricultural labour force or the population more broadly. This report looks particularly at how D4Ag affects young people, climate resilience, employment and women. The next several pages examine how D4Ag could benefit these groups and efforts, the progress and emerging signs of impact so far, and potential risks and challenges. Youth D4Ag is seen as a way to attract more youth into agriculture. Over 60% of Africans are under 25 years old. Every year, 10–12 million youth across the continent enter the job market in search of work.289 Vast numbers of young people continue to work in farming in rural areas – agriculture remains the continent’s largest employer – but urban migration among young people is booming, driven by the promise of higher wages and an escape from the drudgery with which farming is often associated. In this context, experts wonder whether D4Ag has the potential to slow or even reverse this trend. As Michael Oluwaghemi, co-founder at LoftyInc Allied Partners and operator of WeHub, explains, D4Ag “puts the ‘sexy’ back in agriculture for our youths. Our farms could become the offices of the future.”290 Inoussa Maïga, Mediaprod Youth are more likely than their parents to use D4Ag solutions, but it is hard to prove that this affects their choice of career. Based on our survey data, on average, two-thirds of D4Ag users are under age 35, likely due to the simple fact that younger people tend to be more digitally savvy. As yet we have no conclusive evidence that this means young people are actually more likely to consider working in agriculture. However, the attention governments and donors have paid to youth employment in Africa has increased sharply in recent years. As a result, we expect that new research will help us better understand the continent’s employment challenges and will yield more data on the ways in which the digital transformation of agriculture impacts the sector’s ability to create jobs for young people. Even without conclusive data, the chances seem good that D4Ag is pulling more young people into agriculture. D4Ag solutions bring clear benefits, some of which are particularly relevant to youth. First, D4Ag makes jobs in the sector more lucrative by increasing yields and profitability. Many digital solutions also make farming work more convenient and less gruelling, and open up opportunities for youth across the value chain, further increasing Springboard Nigeria its appeal. At the same time, funding from all over the world is going to support entrepreneurship in Africa today – much of it with an agricultural tie-in. For example, in Nigeria, Wennovation Hub (WeHub) “empowers [young] African entrepreneurs to solve their immediate socio-economic challenges by leveraging CHAPTER 3 115 A DEEPER DIVE INTO HUMAN IMPACTS OF D4AG technology and local resources and build[s] their community and collective networks through collaboration.” WeHub has supported over 6,000 young entrepreneurs and invested in around 30 start-ups.291 Much of this space has a strong tech focus, and many of these tech start-ups are run by youth and focused on agriculture. These outfits are also more likely to hire other young people and design products/services that appeal to youth users. Recognising this potential, donors and incubators are working through D4Ag to bring more youth into agriculture. For example, USAID, Syngenta, IREN and the Toyota Kenya Academy created a forum for youth to present their products to possible investors called the Young Innovators Agribusiness Competition.292 Kosmos Innovation Centre and Reach for Change’s Senegal Start-up Accelerator have provided a half-year of incubation support and €1,800 in seed funding to five youth-led D4Ag start-ups.293 Charis UAS Climate resilience Climate change will hit Africa harder than most other continents. Temperatures are rising fast, extreme weather events are expected with increasing frequency, and nearly 70% of Africans work in agriculture – among the most vulnerable sectors to climate change. Farmers will have to cope with changing water cycles and rainfall, more frequent natural disasters, more expensive fuel, and a host of other challenges that have yet to emerge.294 Smallholder farmers bear more risk than others because they depend more on weather-reliant crops and have limited resources to mitigate the stresses climate change will increasingly place on agriculture. The impacts are already being felt. For example, multiple weather shocks in Malawi over the last 20 years have resulted in multiple instances of severe flooding and droughts, including a particularly severe cycle of drought and flooding in 2015. The 2015 weather events resulted in 90,000 hectares of cropped land becoming unusable and the declaration of a national emergency.295 Digital solutions can help farmers become more ‘climate resilient’. First, D4Ag can help improve the quality of short-term and long-term weather information that farmers receive by increasing the accuracy and the location- specificity of weather forecasts. Specific use cases promise additional benefits – advisory services, for example, can provide farmers with additional guidance that can help them adjust to changing weather patterns. We have also seen digitally- enabled weather insurance help farmers protect themselves financially against more volatile weather. In addition, market linkage solutions could provide farmers access to new, more customised inputs as their land and water resources change. For example, farmers may need fertilisers with more or less nitrogen as soil contents change.296 More broadly, by increasing their productivity, D4Ag 116 CHAPTER 3 A DEEPER DIVE INTO HUMAN IMPACTS OF D4AG can help farmers earn additional income needed to invest in adapting to climate change. Policymakers and others operating at the macro level could also harness D4Ag to help systems become more climate resilient and even mitigate the effects of climate change. The vast volumes of data that D4Ag solutions can produce will help policy to become more evidence-based. For example, services that track smallholder farmers’ use of inputs more precisely and in real time could help policymakers understand how climate change is altering the environment. This would also allow top-down decision makers to better tailor policy and programmes to hyper-local environments. Among the most promising digital technologies for climate change mitigation are satellite imagery and remote sensing to evaluate land use and land cover; there could be opportunities for such solutions to help smallholders in the near future. iDrones Zambia Hard evidence of the impact of D4Ag on climate resilience has yet to emerge. As the effects of climate change become more apparent, however, it will likely become easier to observe how digital solutions are enabling farmers to navigate these unprecedented challenges. Already, however, a number of early- mover providers have developed and launched D4Ag solutions that promote climate resilience effectively. Several players providing farmers with data and coaching on adapting to climate change have experienced success. They either offer farmers more accurate/long-term weather forecast data to help them plan better or offer coaching on a broader set of climate resilience techniques. In many cases, these players combine data from a wide range of sources (satellite data, weather stations, GPS, etc.) in order to improve the quality of forecasts. Standout examples include the Grameen Foundation’s Community Knowledge Workers (CKWs), who help Ugandan farmers by providing information on weather- specific agronomic techniques, pests, functioning markets and storage facilities. Digital technologies support CKWs in the form of an online monitoring system FAO and smartphones with relevant applications.297 Esoko also sees information dissemination as an important path for climate change adaptation. It sends climate forecasts, agronomic advice and market prices to farmers in Ghana via mobile phones. This pilot programme increased users’ productivity by a stunning 90%. Interestingly, this model places more emphasis on human intermediation, as employees train farmers on how to use the solutions – which may in part explain its success.298 Weather insurance can provide a safety net for climate-vulnerable smallholders, although it remains unaffordable for those most at risk. CHAPTER 3 117 A DEEPER DIVE INTO HUMAN IMPACTS OF D4AG ACRE Africa, an insurer with partners in Kenya, Rwanda and Tanzania, has developed a suite of products that enable farmers to handle climate risk using a state- and satellite-based weather index, area yield index, hybrid weather index, multi-peril crop insurance (MPCI) and dairy livestock insurance.299 Its success has been attributed to the fact that it bundles insurance with other solutions (e.g., input credit) and sends pay-outs to farmers using mobile money.300 There are still crucial climate data gaps in Sub-Saharan Africa, but the private sector is becoming aware of the opportunities these gaps represent. The quality of data remains far below the standards of most industrialised countries. For example, one major gap in climate-resilience-focused D4Ag is hyper-local weather information. Weather forecast technology is not yet advanced enough to provide the kind of reliable, five-to-seven-day outlook that smallholder farmers need. And even where raw data are available (e.g., from satellites, ground stations), the gap between data and prediction is significant. Yet gaps like these CTA that go unfilled by government present an opportunity to the private sector. Cutting-edge enterprises like aWhere and Ignitia disseminate more accurate local weather information than ever existed before on the continent and continue to invest in R&D to advance this technology. CTA has launched a project in partnership with the International Livestock Research Institute (ILRI) and private insurance companies to promote a market-driven approach to promoting climate resilience in Southern Ethiopia and Northern Kenya.301 Meanwhile, Ignitia raised €988,000 in Series A funding in late 2018.302 Similarly, GSMA has highlighted the opportunity for MNOs to improve their AgriVAS offerings by incorporating weather index insurance products and to invest in location-based services to collect weather monitoring data and offer highly localised services to farmers.303 As the effects of climate change become more apparent and piloted solutions start to demonstrate impact, we expect climate-related digital solutions to expand rapidly in number. Employment It is too early to say for certain, but it looks likely that D4Ag will create more jobs than it will destroy. Evidence for how D4Ag will affect employment is perhaps the least available of any aspect of impact discussed here, likely because of the breadth of the issue and the number of indirect effects that need to be considered. Some commentators argue that D4Ag will create new jobs that will require new roles and the development of new digital skills. Others point out that automation will likely eliminate or reduce a host of familiar roles and occupations. The reality is that both are likely to happen. Without clear evidence to rely on, our hypothesis nonetheless is that D4Ag will likely be a net job creator, perhaps significantly so. 118 CHAPTER 3 A DEEPER DIVE INTO HUMAN IMPACTS OF D4AG Emerging D4Ag solution providers in Africa have employed tens of thousands and this number appears to be growing. Based on current trends, the number of D4Ag solution providers in Sub-Saharan Africa will continue to rise rapidly. The jobs created by this will more often than not be relatively highly skilled – for example, tech developers and business managers. If a few hundred of these providers are active, and each hires 10–100 employees, tens of thousands of new jobs will have been created. Many more jobs will be created among the networks of field agents working with these providers. Today, extension worker density in Africa is about 1 to 1,500 farmers. Successful D4Ag solutions, however, often work with a higher ratio of extension workers to farmers, on the order of one field agent for every 200–500 farmers across use cases like advisory services, input/off-take market linkages and financial service intermediation on the ground (e.g., support Sonita Tossou for informal digital smallholder farmer village savings and lending group). D4Ag solutions are able to substantially reduce farmer-to-field-agent ratios because digital technologies allow for the upskilling and more efficient monitoring and management of young and inexperienced field officers who require less training and are far less expensive than professional agronomists. Beyond reducing the costs of field agents, digital solutions also improve agent profitability or cost- coverage. With the help of digital solutions, such agents generate incremental value for farmers and other value chain intermediaries like input providers, off-takers and FSPs, thereby making it much easier for D4Ag enterprises to retain such agents or for other players to hire them in large numbers. It is also critical to note that such agents are not a replacement for existing African professional agronomists, but more a complementary last-mile human network that supports value chain formalisation on the ground. If D4Ag solutions were to become ubiquitous in farming across the continent, this would imply between a threefold and sevenfold increase in the number of field agents. In absolute terms, this would mean the creation of hundreds of thousands of jobs. CTA D4Ag will increase not only the number of jobs but also their quality. Today, just ~7% of smallholders in Africa work in tight value chains. D4Ag can help them enter well-organised value chains that will increase productivity and, by extension, the level and stability of their income. Digital solutions can help achieve this by improving communication and reducing transaction costs. We also see the opportunity for D4Ag to create formal jobs further up the value chain in agriculture processing and manufacturing. As these sectors tend to be higher value-add, this would translate into higher paying jobs for today’s farmers. CHAPTER 3 119 A DEEPER DIVE INTO HUMAN IMPACTS OF D4AG While these prospects are encouraging, policymakers need to think about groups that will inevitably lose out from this transformation. There is no doubt that D4Ag will automate significant numbers of people out of jobs. It is important to look not just at the aggregate impact of D4Ag on jobs; there will be winners and losers. If and when D4Ag becomes truly pervasive, we will see a divide between the ‘haves’ and the ‘have-nots’, or those who were left out of the agricultural transformation journey. The ‘have-nots’ may be driven out of farming altogether by consolidation, stricter quality assurance and price competition. We do not expect this to transpire in the short to medium term – most of the D4Ag industry is still trying to develop viable business models that do not rely on grant support. But it is important for policymakers – especially those investing in D4Ag solutions – to keep this in mind as they ramp up their support for D4Ag and form their visions for the future of agriculture in Africa. Women Filippo Brasesco, FAO D4Ag solutions, in theory, have the potential to be transformative for women. Most women (60%) working in Sub-Saharan Africa are employed by the agriculture sector.304 They play leading roles across the agricultural sector as buyers (e.g., in the pineapple value chain in Ghana) and local processors (e.g., as members of Sooretul, an e-commerce platform in Senegal). As with men, digital solutions can increase incomes and yields for women farmers by improving USAID 120 CHAPTER 3 A DEEPER DIVE INTO HUMAN IMPACTS OF D4AG agronomic practices, connecting them to markets, and providing credit. But D4Ag offers an additional value that is particularly relevant to women. Due to social norms, many women across the continent are largely confined to their homes. Digital tools like advisory services and market linkage can allow them to access products and services despite this restriction. In doing so, these tools have the potential to increase women’s ability to organise and work collectively – one of the most significant drivers of women’s empowerment. However, this potential has yet to be realised. Few D4Ag users are women. Enterprises surveyed report that women comprise 25% of their user base, which is consistent with data from large solutions, and indicates lower reach to women. Moreover, a large share of respondents (57%) did not feature reaching women in their top priorities. A number of factors contribute to the gender disparity in D4Ag engagement – among them, the underlying gender gap in digital access. Women in Sub-Saharan Africa are 15% less likely to own a mobile phone and 41% less likely to use mobile internet than are men.305 Given that the vast majority of solutions require one or both of these, it is much harder for enterprises to reach women. Reports suggest that the main barriers to female mobile engagement in developing countries are affordability, literacy and skills, safety and security and relevance.306 Yet, providers in Ethiopia, for example, have shown how to work around low digital literacy levels or internet access (see Ethiopia case study in Annex); similar principles could be applied elsewhere to Antonello Proto, FAO reach more women. On the supply side, businesses, donors and governments appear to view a specific focus on engaging women as too great a challenge given the barriers to engaging any farmer in D4Ag solutions. Today, D4Ag solutions primarily reach what providers consider the lowest-hanging fruit – (male) farmers in tight value chains. Most enterprises and initiatives fail to prioritise outreach to women and other marginalised segments – and, unsurprisingly, fail to reach them in significant numbers. To address the gender gap in D4Ag, the entire sector needs to make women a priority. This will require mainstreaming gender in D4Ag initiatives by building gender concerns into donor programming and enterprise solution design. It will also require advocacy to ensure that gender becomes a funding priority. Industry players can take steps to make it easier to work with women – starting with more inclusive data and solution design. Gender-disaggregated data remain sparse, which hinders problem identification CHAPTER 3 121 A DEEPER DIVE INTO HUMAN IMPACTS OF D4AG and trends analysis with respect to women’s empowerment in agriculture. For example, Technoserve’s Coffee Initiative began collecting data at the individual – instead of the household – level in order to more accurately track training attendance and coffee tree ownership by gender. This was one of multiple measures that may have contributed to increasing female participation in the programme from 6% to 42%.307 As providers then move into solution design, more effort is also needed to involve women users in this process. Rapid prototyping and testing should help ensure that D4Ag solutions are responsive to women’s needs. Many of the most active players in D4Ag have applied this to various elements of their businesses targeting women. For example, MyAgro recognised that women farmers typically have smaller land plots and less liquidity than men, and began selling inputs in smaller batches for crops that women typically grow.308 Implementation decisions are also crucial. Other agriculture operators CTA in Africa have demonstrated the imperative of disseminating information and products in safe, convenient and inclusive locations. For example, the Wakulima Tea Company in Tanzania developed 30-minute trainings about application of inputs including fertiliser, held while farmers wait for tea collection trucks; this increased attendance, particularly for women, who perform 70% of tea harvesting. Having gender-diverse programme representatives also matters. A World Bank and International Food Policy Research Institute (IFPRI) study found that female extension agents are more likely to serve female farmers than are male agents (the ratio of women to men was 1.30 for female agents and 0.53 for male agents).309 MyAgro’s wider work stands out as an exemplar of how to build a strong base of women users. MyAgro is a mobile layaway programme in Mali and Senegal that equips farmers to buy seeds, fertiliser and training packages. In a short period, it has demonstrated impressive impact by spurring 50–100% increases in harvest yields and €108–334 additional income per farmer. It has also managed to build a user base that is 60% women. MyAgro attributes this achievement to a number of factors: (1) it involves women in its design phases, particularly for products used in the types of farming dominated by women (e.g., peanut farming, or farming on plots smaller than three hectares); (2) it offers smaller seed and fertiliser packets and mobile layaway options, which benefit women, who are more likely to be cash poor; (3) it disseminates information and products through women-dominated village savings and loan associations (VSLAs); (4) it develops village-level distribution centres to work around women’s mobility constraints; (5) it focuses explicitly on recruiting female field agents; and (6) perhaps most importantly, it also tracks the impact of these efforts by collecting and analysing gender-disaggregated data.310 122 CHAPTER 4 WHERE WE ARE HEADED Image to go here CTA We are entering a new phase of more powerful and more capable D4Ag solutions, fuelled by the power of data and ongoing business model innovation. We will see better products, underlying improvements in D4Ag infrastructure, greater investments and many new players. Within three years, the sector could approach 60–100 million registered smallholder farmers and generate annual revenues of €260–380 million.311 Over the next 3–5 years, we expect to remote sensing and farmer data) and see five major trends in the African the corresponding growth in sector data D4Ag space: analytics capacity to deliver more precise, real-time and impactful D4Ag solutions to 1 Accelerated business model the market innovation with an increased focus on solutions that formalise smallholder 3 Increased adoption and use of value chains including D4Ag market innovative technologies for D4Ag linkage services and bundled services, (e.g., remote sensing, diagnostic, D4Ag ‘super platforms’, and agriculture IoT sensors), several of which will payment digitalisation initiatives, which will move beyond experimental pilots to deliver more value to smallholder farmers, scale, contributing to the data revolution agribusiness, and FSPs, and lead to more highlighted above, and also unlocking new attractive D4Ag sector economics business models and impact opportunities 2 Growth in the availability, 4 Increased Africa D4Ag investment affordability and use of valuable by tech VC investors and large agriculture data at scale (e.g., commercial players including big CHAPTER 4 123 technology companies, MNOs and agribusinesses and, in parallel, growing “Data capture continues to get better, faster investment from philanthropic funders into supporting D4Ag infrastructure public and cheaper, which has led to a growing goods (e.g., national-level agronomic data wealth of available information. collection, weather and pest surveillance, ” farmer registries) 5 Continued improvement in D4Ag at the last mile for smallholder farmer market enablers, setting the stage for linkages, mechanisation, logistics and financial much more dramatic agriculture service delivery.312 digitalisation progress in the longer (5–10 year) time frame, including This pivot, which is already underway, is the growth in connectivity and phone access, result of several interrelated business model expansion of digital payments and digital insights – highlighted in Chapters 2 and 3 ID systems and the continued growth and of this report and recapped here – reached maturation of Africa’s D4Ag incubation by leading D4Ag sector actors and experts. and investment ecosystems. These lessons are informing where and how entrepreneurs, commercial investors and Taken cumulatively these trends should donors are allocating their resources for the translate into more impact at both the next phase of the D4Ag sector’s growth. smallholder farmer and macro-economic levels and, critically, a stronger D4Ag The first of these insights is that D4Ag business and impact case for the next solutions that focus primarily on decade of agriculture sector digitalisation in data collection and the delivery of Sub-Saharan Africa. At the same time (as information and advisory services are noted in the discussion below and in Chapter important but insufficient. On their own, 5 of this report), ensuring and sustaining information and advisory services are unable the positive evolution of the D4Ag sector to maximise farmer impact in the absence of will require a concerted focus on addressing parallel and closely linked systems that ensure systemic challenges to D4Ag scale-up and farmers’ access to inputs, markets and finance. managing emerging risks. In addition, solutions narrowly focused on Accelerated D4Ag information and advice delivery are highly business model innovation constrained in their economics due to the limited willingness of farmers and other will transform the D4Ag smallholder farmer value chain actors to pay landscape for advice and information. The willingness to All D4Ag use cases will see rapid growth pay is not zero and is growing over time, but in the next few years, but the relative the economic value that can be generated per emphasis of the sector will continue farmer (e.g., via farmer fees, data monetisation shifting toward digital solutions that or B2B payments by agribusiness) is still aggregate and formalise smallholder insufficient – and will remain so for the value chains. We project a clear pivot foreseeable future – to sustain high margins in of business model innovation and sector most contexts. Such economics are, therefore, investments to digitally-enabled market typically inadequate to provide for national formalisation and aggregation solutions, or region-wide scale-up of digital advisory particularly those that utilise digital tools to and information solutions without substantial support and supplement human agent networks ongoing donor and government subsidies.313 124 CHAPTER 4 Looking forward, this does not mean that advisory services being a standard component, digital advisory and information solutions will but typically not one that is monetised or that no longer be in favour – rather, the number of is essential to the business model’s viability. solutions with an advisory services component There will still be a continued niche for and the reach of such solutions will continue to specialised digital advisory enterprises (e.g., grow quickly. Large-scale public (e.g., Ethiopia for weather data, pest and disease data) who 80-28) and donor-funded (e.g., Digital Green, provide B2B information and/or capacity- PAD) digital advisory and information services building services to other D4Ag enterprises, will grow and remain important as generators but these will be relatively few in number of essential public goods. However, we predict compared to market linkage solutions. that ‘pure play’ advisory solution models among commercially-minded D4Ag enterprises The second related observation will become far less common over the next few concerning D4Ag business models years. recognises the value – in terms of both D4Ag impact and economics – of Commercial D4Ag advisory solutions bundling solutions.314 Incipient evidence will broaden their mandate by suggests that breakthrough impacts on farmers combining the advisory service value (>50% increases in incomes, >100% growth proposition with digital market linkages in yields) are possible with the help of D4Ag (input, mechanisation and off-take solutions. However, results like these typically linkage services). They will either do require a holistic approach to serving the needs this directly by incubating market linkage of smallholder farmers by providing digitally- solutions in-house to augment or sit alongside enabled market linkages, advisory services and the advisory product (e.g., the path taken by financial services.315 Esoko, Farmerline and Digital Green) or via third-party partners with whom they will share From a business economics perspective, aside value. In line with this trend, we expect that from the increased upfront complexity and cost the majority of D4Ag solutions in 3–5 years of setting up such solutions, bundled solutions will primarily focus on market linkages, with are also uniquely attractive. The key drivers Fiondella, IRI/CCAFS CHAPTER 4 125 for improved profitability and scalability of bundled solutions include costs savings due to operational synergies and, more importantly, increasing willingness on the part of farmers to pay for those bundled products that can generate instant economic value – which can take the form of lower input costs or higher, more guaranteed off-take prices alongside the harder to quantify long-term effects of improved farmer productivity and resilience through better practices (which farmers are often unwilling to pay for in the near term). The most immediate implication over the next few years will be the rise of bundled D4Ag ‘super platform’ solutions as the most common architecture for D4Ag service delivery. The idea of bundling to enhance D4Ag solution impacts and economics is not new. It has informed, for instance, several phases of Mercy Corps’ AgriFin Accelerate programme for the past seven years, starting with bundles of Africa’s myriad policy regimes, value chains Georgina Smith, CIAT finance and advisory services in a handful and cross-border trade and logistics challenges, of country pilots and broadening to much a winner-take-all approach for D4Ag platforms broader commercial concepts exemplified by is unlikely for the foreseeable future. Safaricom’s DigiFarm. What is new today are the improved and still evolving ideas about how The most likely scenario in the next to make such models work and, as discussed in few years is a complex ecosystem of depth in Chapter 2, the resulting emergence of competing and sometimes collaborating D4Ag ‘super platforms’ as a distinct category super platforms: commercial providers with of D4Ag solutions. proprietary, custom-built digital platforms that formalise loose value chains via direct We foresee a proliferation of D4Ag super agent market integration models (e.g., Tulaa, platform solutions – many at national or Twiga, One Acre Farm), micro-entrepreneur value-chain levels – competing with each platform models (e.g., Kuza), farmer hubs other, likely with multiple successful (e.g., Multiservices Agricole in Senegal), bank players and models emerging in the platforms (e.g., KCB MobiGrow), value chain interim. We predict that the D4Ag super management solutions designed for agribusiness platform model will become the dominant (e.g., SAP Rural Sourcing Platform, Olam’s approach in the sector in just a few years, in-house digital stack), government-affiliated or but this does not necessarily mean that the -led platforms (e.g., Smart Nkunganire System sector will be dominated by a few big unitary in Rwanda), solutions from different specialised commercial digital agriculture platform D4Ag vendors bundled under common super providers. That is one possible outcome, but platform commercial brands and farmer an improbable one given the diversity of interfaces (e.g., Safaricom’s DigiFarm), families sector needs. In the longer term (5–10 years), of inter-linked digital solutions or enterprises a progressive winnowing and consolidation (e.g., Farmerlink, Esoko and – a the very large of solutions is likely, but with Sub-Saharan end of that scale – Alibaba’s Rural Taobao 126 CHAPTER 4 system in China) and, finally, looser consortia – which we do believe have a role to play and models, such as the Digital Green-led digital will continue to be important – are unable to agriculture consortium and related initiatives in match the impact of hybrid models due to the Ethiopia, which embrace a more open digital familiar barriers of connectivity in the field, agriculture ecosystem but link independent digital literacy, farmer trust in digital content players together via a common mission, and the difficulty of localising content – all common distribution channels and common issues where human intermediation can help. application programming interfaces (APIs) to ensure the delivery of holistic solutions to For these reasons and others, many sector farmers. experts have concluded in recent years that, while direct-to-farmer D4Ag solutions are an Another important insight for the important supplemental or ancillary channel future of D4Ag business models is for smallholder farmer engagement, for that transformational impact on maximal impact and commercial sustainability smallholders requires digitally-enabled – in the words of a recent D4Ag business human networks, not just purely digital model review by the Syngenta Foundation, solutions. Human networks consisting of a funder of several such models – “field last-mile agents or ‘field forces’ of various forces [will and must] remain an essential types (e.g., agriculture extension officers, digital actor in disseminating and embedding digital finance agents, market linkage agents, advisory agriculture solutions” on the ground.317 micro-entrepreneurs, ‘lead farmers’) have been a feature of D4Ag solutions for years (roughly We believe that D4Ag hybrid ‘digital + 25–35% of solutions in our database feature human’ business models will become agents in some way),316 but much of the energy much more common for less formal in the African D4Ag sector in the past decade agriculture value chains in Africa.318 has been focused on the 65–75% of solutions The logic of sector impact and sector that are direct-to-farmer via SMS, USSD, economics will push D4Ag super platform IVR channels or, more recently, smartphone players inexorably in this direction given the applications. This focus has been unsurprising lack of existing last-mile agent forces needed as virtual, i.e., ‘pure digital’ models are cheaper to support digitally-enabled market linkage to deploy. and logistics operations. Our interviews with sector experts repeatedly One well-trodden pathway to greater highlighted that purely digital D4Ag solutions integration of human and digital tools will Georgina Smith, CIAT CHAPTER 4 127 continue to be D4Ag enterprise partnerships In terms of scalability, such models do require CTA with existing third-party agent field force more upfront investment and present greater organisations to digitalise how such risks, but these are risks that should be organisations interact with their smallholder quantifiable and manageable for commercial farmers (e.g., One Acre Fund extending digital investors as the evidence for hybrid business tools to its agents or Digital Green providing models accrues over time. Large corporations a digital overlay for existing national extension may be willing to take on such bets for the agents). The costs of agents in such models are same long-term, profit-driven reasons that not born by the D4Ag solution, but by third Alibaba in China is investing into Rural parties. In most cases, however, such third- Taobao’s last-mile infrastructure of stores and party organisations simply do not exist for agents (60,000 agents today, with plans to informal agricultural value chains, and other expand to more than 300,000 agents over the alternatives are needed. next few years).320 Donors and governments, for their part, should have a strong interest in More novel and promising from an supporting and de-risking such models, given impact standpoint are approaches that they function as direct rural job creation that involve D4Ag players building engines. their own agent field forces, salaried or commission-based, alongside their digital The final D4Ag business model trend platforms (e.g., myAgro, Tulaa, Twiga, that we believe will be notable in the DigiFarm) or using a digital platform as a next few years is an increased focus on tool for recruiting, training, capacitating and agriculture payment digitalisation as managing agricultural micro-entrepreneurs in an entry point for D4Ag solutions. There the field (e.g., Kuza). Such models have rightly is growing recognition today that expanding been seen as more costly and operationally digital payments and building responsible complex than purely digital solutions. When digital payments ecosystems are fundamental considered in light of the impact potential and to creating a more productive and sustainable sustainability of hybrid models, however, the agricultural sector.321 barriers to integrating human agents (often fairly low-wage-earning youth who can be By enabling farmers to receive compensation, upskilled and managed via digital tools) are transparently and securely for their crops, likely more easily surmounted than what is digital payments allow them to save money commonly believed, leading to a high return and reinvest it in their agricultural activities. on investment.319 For agribusinesses, digital payments generate 128 CHAPTER 4 CTA the benefits of security and speed, as well substantial energy on supporting agriculture serving as an entry point into broader payment digitisation, with a primary focus digitalised supply chain relationships with on formal agribusiness procurement from smallholders that can generate marketing highly commercial value chains like cocoa upside, improve product quality/traceability or in West Africa. generate other operational efficiencies. Our interviews and desk research The GSMA mAgri team has estimated that suggest that agriculture payment the potential market for agricultural payment digitisation initiatives will continue digitalisation is already substantial and likely to increase in scale and ambition in to grow quickly. By 2020, the potential value the next few years. Building off existing of formal procurement payments to farmers pilots with GSMA and others, MNOs have in Africa will be ~€300 million annually, of announced an increased number of agriculture which only 5–10% is captured via payments payment digitisation projects and partnerships digitalisation today322 – a major opportunity in 2018–2019. The launch of the new GSMA for MNOs on the continent. In addition to Innovation Fund for Digitisation of the revenue potential, GSMA has assessed Agricultural Value Chains as this report that “digitising agricultural payments could was going to press will likely add further generate measurable indirect benefits for momentum to such initiatives. Development mobile operators related to the acquisition of banks like the African Development Bank new mobile money users, increasing loyalty, (AfDB) and the World Bank are embracing the increasing volume of transactions and overall payment digitisation opportunity for priority activity on mobile money accounts to support geographies (e.g., AfDB’s Togo smallholder a sustainable agent network.”323 payment digitisation project). In the past few years, players like GSMA and Even in less formal value chains, payment the Better Than Cash Alliance, as well as digitisation is increasingly becoming a standard corporations like MasterCard, have focused feature of D4Ag super platform projects, such CHAPTER 4 129 as MasterCard Farmer Network, DigiFarm which are IoT, blockchains and machine and KCB MobiGrow’s work in East Africa, as learning.326 well as smaller-scale D4Ag platforms like Tulaa and Twiga. Organisations are already seeing As discussed in depth in our overview of results and that will likely add further impetus emerging D4Ag solutions in Chapter 2, to the digitisation movement. One Acre Farm, we already saw many examples of how for example, has moved aggressively to digitise sector actors are making use of these data loan payments with its 800,000 farmers and, to enable more tailored, precise, real-time based on early results in a few geographies, recommendations for farmers; give financial has reported reductions in payment losses service providers the ability to better assess and collection costs (of 80%), increases in and control risks; and provide valuable insights operational efficiency (approximately ~50% into smallholder supply chain needs and less time spent by agents on payments opportunities for agribusinesses. collection) and higher farmer satisfaction relative to cash-based loan payments.324 While we are excited about the promise of advanced technologies and the growth in data, While we cannot predict what share of farmer many technologies (e.g., drones, field sensors) payments will be digitised and by when will likely remain in the experimentation phase based on the data available, it is clear that in the African smallholder farming context payment digitisation is on its way to becoming for years to come and do not yet have fully a standard feature of D4Ag solutions and settled business models, or at least not yet at interventions. scale. It is therefore important, as D4Ag actors experiment with these technologies, that they Vastly larger data volumes continue to capture the evidence needed to build the business and impact cases such as and growing data technology investments. analytics capabilities will result in more impactful We are already seeing an explosion D4Ag solutions in raw data capture from a range of sources, yet the agriculture data D4Ag solutions will increasingly use ecosystem remains fragmented. The cutting-edge technologies – fuelled by sheer amount of data collected has increased new sources of data and improved exponentially.327 This includes farmer data, analytical capabilities – to increase their soil/land/crop data, and water and climate Aurora Photos, Alamy value proposition. This will enhance the precision and relevance of D4Ag solutions, even as they become easier for farmers to access and use. We have seen signs of this trend in our research; over one-third of the respondents to the CTA-Dalberg survey already use at least one form of advanced technology – defined here as drones, augmented/virtual reality, blockchains, machine learning, the internet of things (IoT), big data, artificial intelligence/machine learning, and voice activated technology.325 Nearly 60% of respondents expect to integrate new technologies over the next three years, the most popular of 130 CHAPTER 4 data. The trend is explained in part by the them to generate powerful insights. On a more ubiquity of mobile phones (e.g., mobile institutional and policy level, it is becoming surveys), but a number of other technologies increasingly clear that data aggregation is facilitate agriculture-specific data capture at only possible with better defined data even greater scale and lower cost – namely, regulations and innovative data-sharing drones, sensors, and satellites. business models; progress on both of these fronts is at an early stage. The data capture from these sources continues to get better, faster, and cheaper, which Strong data analytics capacity – essential has led to a growing wealth of available in deriving insightful recommendations information for both D4Ag intermediaries for farmers from increased data volumes and farmer end-users. However, despite the – is developing rapidly but currently lags growing volume and promise of data, we are behind the pace of data generation and still seeing a very fragmented data ecosystem, capture. Data analytics and machine learning with many valuable datasets – including much – two methods by which to leverage these raw of the data from the public agronomy research data – are in more experimental stages but are community at national and regional levels – quickly improving. There are many forms of locked in organisational silos, not fully digitised, data analytics, each of which serves a distinct or embedded in proprietary systems owned by purpose: descriptive, diagnostic, predictive, financial institutions and agribusinesses. prescriptive or cognitive.329 A handful of agriculture sector actors have begun to Sector actors have started to recognise experiment with integrating those capabilities the importance of aggregating data. into their businesses. The most common These is a growing focus in the sector, led models to date have involved specialist by open agriculture data initiatives from agriculture data analytics vendors who collect, organisations like GODAN and the Open analyse, and sell data to interested parties, Data Institute (ODI), on ways to ensure that or in-house teams that accumulate data from whatever data are captured are stored in an other places.330 accessible, usable format, and are employed by a broad range of players to improve farmers The focus for many players over the lives.328 On a technical level, cloud storage and next three years will be on continuing to big data analytics tools facilitate the low-cost improve the quality of data capture and storage and aggregation of data in ways that then developing meaningful, actionable allow others to easily access them and use insights from these data sets. Big data has CTA CHAPTER 4 131 an important role to play here; we expect the should lead to markedly better products for ‘winners’ to be those who are able to combine B2B and B2C users, as they will be specifically the various datasets in the most meaningful and precisely designed to meet these users’ way and package the insights so that they needs. resonate with farmers.331 Machine learning will be an important tool for accelerating this This data-driven approach will push process. As algorithms learn and improve, they past some of the limitations of today’s can have increased relevance and power for solutions in order to target what people specific enterprises and farmers. want. Data-informed solutions will be designed around a deep understanding of their However, not every organisation will users’ behaviours and needs; as such, they have the financial and human resources should encourage higher uptake and create to follow this path. The use of data – and real impact for farmers. Eighty per cent of especially the more advanced technologies survey respondents indicated that they have around data – requires specific skill sets and tailored or plan to further tailor their products sufficient resources to invest. Many players for smallholder farmers. Moreover, the today lack one or both of these. We expect ongoing collection of data and use of pattern- that many D4Ag organisations will try to recognition and machine learning tools should embrace the potential of data, but only a small enable D4Ag solution providers to recalibrate percentage (though impossible to quantify) will their solutions based on user results and the be able to take advantage of it. Thus, in the ability to diagnose what is and is not working. coming years, we may also see some greater consolidation within the sector as data analytics This ‘data revolution’ will not only leaders outcompete their slower-moving rivals. allow for improved user information and feedback loops but will also Successful solutions will be those that extend the offerings that solutions can ‘crack the code’ on how best to can provide smallholders. For example, use data.332 These solutions will be able to chatbots that share photos with farmers and integrate the many sources and types of data voice-based solutions that allow farmers to in a compelling way to best deliver value to hear advice rather than read it have begun the farmer. The data-informed output must be to overcome the challenges of illiteracy and insightful, precise, simple to use, and – most low connectivity. Additionally, data-driven importantly – truly address the pain points that solutions can provide smallholders with critical farmers care about most. This ‘data revolution’ farm guidance with an unprecedented level CTA 132 CHAPTER 4 of precision, localisation, and customisation. into these systems would be solutions skewed Similarly, drone technology is being used to towards men, their outputs would reflect the create highly accurate maps that can be used same biases. These technologies also come with for mapping land boundaries with a range of other important risks and concerns around possible uses, such as land titling and clarifying data governance and consumer protection land ownership.333 These and other methods (including privacy and informed consent). should further bridge the gap between reach In Chapter 5, we discuss how governments, and impact. donors and investors can ensure that these technologies are adopted in an effective and The increased use of data in agriculture appropriate manner. is not, however, without risks. To begin with, many of the technologies in question User design, experience, and (e.g., machine learning, data analytics) understanding must also go hand in leverage similarities. In other words, they hand with such data-based insights. One rigorously use data from one case to predict commonly cited benefit of data analytics is another. This reliance on commonality could that it “can reduce the amount of direct input present a challenge in a sector as massive needed from the farmer”.336 But by distancing and varied as agriculture.334 The agricultural themselves from farmers, solutions may more sector in Africa comprises nearly 70% of the easily misrepresent their desires and needs. workforce and differs widely from place to The balance between data and ground-level place in crop, climate, human context, farmer knowledge is an important one to strike and characteristics, etc.335 will be discussed more later. Moreover, when it comes to data analytics, Longer time horizons are the key to and artificial intelligence especially, there is managing these and other risks. It is a danger of reinforcing existing biases. As critical that players take time to think through one illustration, today’s solutions currently the consequences of the models and methods reach very few women or other marginalised they design before implementation and follow groups. The algorithms in question are based up with rigorous evaluation and adjustment on inputs of historical data. Since all inputs – even if doing so slows down the pace of FAO CHAPTER 4 133 transformation. Moreover, by its very nature the agriculture sector moves more slowly than the technology sector; tech players will need to practice patience and re-orient themselves toward a more long-term approach. Failing to do so will risk entrenching existing issues in the design of new solutions, creating new and unanticipated consequences, and veering away from an inclusive agricultural transformation. Innovative technologies for D4Ag will support the agriculture data revolution and also enable new business models and impact possibilities Solutions built on emerging technologies – several of which are beginning to move from experimental pilot to scale – will contribute to this new age of data- In the case of the internet of things, for C. Schubert, CCAFS driven agriculture by providing new and example, we estimate based on the CTA- better sources of data, improved data Dalberg database that in 2019 likely fewer storage and aggregation, and stronger than 50,000 smallholders in Africa had capabilities of analysing and using this a field sensor on their farms and perhaps data. IoT helps generate massive amounts of several hundred thousand were starting to data. Big data makes it possible for the storing, experience the benefits of machinery sensors in processing, and analysis of this data to arrive at tractors via Hello Tractor and irrigation pumps potentially powerful insights. Machine learning via SunCulture. allows us to improve solutions on an ongoing basis, building algorithms that understand users Likewise, we estimate that across the 30+ even better than we may. Each technology is smallholder-focused drone start-ups in Africa, individually powerful; in combination, they only a few hundred thousand hectares of land create a virtuous cycle that can generate even have been scanned and, likely, only tens of more precise and tailored products, pushing thousands of African farmers have had the boundaries of what D4Ag can do. their field analysed via drone flyovers in the past few years. There is a long way to go It is important to note that the use of for these solutions to become mainstream in these technologies in Africa is still early the sector, but in every single case there are and experimental in nature. This is due (in encouraging signs of major investments on the some cases) to the nascency of the technologies way or new commercial entrants focused on themselves, regulatory and policy constraints technology integration. (e.g., policy constraints on drone operations), the relatively high levels of capital investment Here we provide a snapshot of each of these required, and the lack of additional skills technologies and their relevance to agriculture needed among people designing and using in Sub-Saharan Africa, as well as early these technologies or trying to adapt them to examples of their application and a glimpse at the African context. their future potential. 134 CHAPTER 4 NEW TECHNOLOGIES POWERING THE PATH FORWARD The internet of things (IoT) Collecting and transferring vast amounts of data with mobile phones, sensors, drones, and satellites IoT – a term used to describe the connection of devices to the internet – enables the generation and transfer of massive amounts of data. IoT enables devices that gather data (e.g., sensors, mobile phones, drones, satellites, etc.) to transmit the data they capture over the internet. Importantly, IoT allows one to capture data from a source without being there in person; this ability is the basis of the surge in available data today. At 30% year-over-year growth in connections since 2015, IoT is growing quickly in Africa.337 This growth in IoT connections has the potential to help transform agriculture through the use of a range of devices to bring precision farming – historically, a luxury only Western countries could afford – to Africa. Much of this growth is being fuelled by the falling prices of IoT technology. For example, the FarmBeats project has developed a cheap alternative to a drone that can capture farm data from the air. “Tethered Eye” Charlie Pye-Smith, CTA helium balloons act as aerial sensors, collecting images of farm conditions and then refining the data collected by sensors on the ground.338 IoT devices use a vast array of sensors to capture localised and valuable data to support agriculture in Africa: (i) location sensors that use GPS signals that capture precise latitude and longitude details of individual farms; (ii) soil sensors, which help determine soil properties, pH conditions, nutrient levels, air permeability and moisture levels; (iii) weather stations that use a combination of sensors to capture climatic data including air temperature, soil temperature, wind direction and speed, rainfall and atmospheric pressure; (iv) storage sensors that check gas levels, moisture, and other conditions that could contribute to post-harvest loss;339 and (v) livestock sensors that measure location, activity, and health metrics like temperature for animals.340 Combined, the insights from these IoT devices can provide farmers in Sub-Saharan Africa with a number of benefits, such as boundary mapping, weather prediction, yield monitoring, disease detection, fertiliser calculations, and harvest predictions. The insights emerging from sensor data are meant to help farmers make better decisions (e.g., concerning input use) – based on localised, customised, and real-time information – that ultimately improve crop quality and result in greater yields. For agribusinesses and FSPs, these insights can be used to tailor marketing activities (e.g., offer more customised fertilisers) or even extend services to farmers that otherwise would not be available – for example, CHAPTER 4 135 yield prediction data can give FSPs the comfort they need to offer farmers loans; similarly, weather data can help insurers extend insurance to farmers. For governments, maps with such detailed information can help improve macro-level decision-making and resource allocation. It can also help increase the value of their extension agents on the ground, who can make recommendations based on individual farmers’ needs rather than, for instance, relying on outdated, generic soil cards. We are starting to see some promising signs emerge for each of these use cases. Ujuzi Kilimo, a Kenya-based D4Ag firm that uses soil sensors and data analytics to send highly localised advice to farmers via text message, draws on data from satellites, sensors, institutions, and local weather to “generate insights using machine learning and data analytics.”341 Zenvus, based in Nigeria, uses soil data to optimise inputs and drive access to finance; Zenvus is currently Charlie Pye-Smith, CTA making use of the IoT technology in its Smartfarm products to collect vast amounts of soil data from the smallholder farmers it works with. These data both inform the use of fertilisers and pesticides at the farm level and are being sold on a subscription basis to banks to increase lending, insurance, and investments. There is some emerging evidence that these technologies are creating positive impact on the ground, but they are still too new to make definitive claims. For example, players such as Microsoft FarmBeats, Zenvus, Ujuzi Kilimo and Lentera, which use on-farm sensors, report that farmers receiving advice are able to substantially improve their yields due to improved advice precision. While these advancements are encouraging, they are typically not yet rigorously measured with external validation and robust impact measurement techniques. IoT for agriculture is still experimental in nature in Sub-Saharan Africa; even with rapidly declining field sensor costs, it will likely take 5–10 years or more before IoT solutions are mainstreamed at any scale. The underlying technologies are still expensive (though rapidly falling in price), devices do not always work (i.e., sensors have often been built for Western markets and have not sufficiently been tailored for local markets), and farmers and actors do not always know how or choose not to implement the insights and recommendations. Furthermore, the growth of IoT, as with much of D4Ag, is uneven and often limited to the usual suspects: Kenya is leading the way on IoT uptake for smallholder farming, and experiments are underway in Ghana, Nigeria, Rwanda, and, to a lesser extent, Senegal.342 136 CHAPTER 4 NEW TECHNOLOGIES POWERING THE PATH FORWARD Big data Bringing large sets of data together to generate deeper insights Big data allows companies to store, aggregate and analyse large sets of data to generate insights that inform business decisions. Strictly speaking, big data is a term that is used to describe large volumes of data and datasets. Yet it is not the quantity of data that matters so much as the ability to aggregate, store and analyse all these data to generate insights. For the purposes of this discussion, we therefore refer to big data as both the datasets and the processing capabilities. Applying big data to Sub-Saharan African agriculture can improve farmers’ livelihoods and inform better decision-making at the macro level. Big data capabilities are allowing D4Ag actors to generate insights from the vast amounts of data now being generated. Indeed, in many of the examples we described above, big data analytical capabilities are powering enterprises’ ability to make use of the data that they are collecting (from, among other CTA sources, IoT connected devices) across all of the use cases we discuss in this report. More broadly, big data is transforming disciplines like genomics, crop breeding, climate modelling, and agronomy. By analysing new datasets in more powerful ways, we can accelerate the development of better responses to some of the most pressing challenges facing Sub-Saharan Africa: climate change, food insecurity, and environmental degradation.343 Donors and developing country governments have woken up to the imperative of bringing big data to agriculture. In 2018, a coalition including the Food and Agriculture Organization (FAO), the Bill and Melinda Gates Foundation (BMGF), and national governments launched a €449-million fund to help countries in Sub-Saharan Africa, Latin America, and Asia gather more data on small-scale farmers to help them learn and adopt better farming practices. The work will focus on expanding surveys run by the FAO and the World Bank to gather information on factors like livestock holdings and crop yields.344 CGIAR, a global partnership to advance research into food security, has also set up a platform (known as the CGIAR Platform for Big Data in Agriculture) in order to harness the power of big data for agricultural research. The platform aims to improve the use of big data within the CGIAR system, open up and share data outside the CGIAR system, and help facilitate partnerships to expand the breadth of big data capabilities in agriculture.345 CHAPTER 4 137 This momentum is supporting the growth of more solutions built on big data. Kilimo Salama (now ACRE Africa), a company launched in 2009 that offers an insurance product for smallholder farmers, is one such example. It is the product of a partnership between the large insurer UAP Insurance, Safaricom, and Syngenta Foundation for Sustainable Agriculture (SFSA). It combines real-time weather data with regional-level historical climate and crop yield data to estimate indemnities more accurately and efficiently. The project has now expanded to other countries in the region (e.g., Rwanda, Tanzania). Evidence suggests that farmers who were clients of ACRE Africa invested 20% more in their operations and generated 16% more income than did those farmers who were not insured.346 CGIAR has also supported several big data tools for farming. For example, in partnership with the International Potato Centre, it launched an online Pest Distribution and Risk Atlas for Africa – an open-access, mobile-accessible resource that combines up-to-date information on major insect threats to crop production with risk maps for each pest and predictions for future climate scenarios.347 But big data analytical capabilities in Africa are still limited. Big data analysis is often conducted by third-party private firms that offer their analytics capabilities to private and public clients. For example, MNOs like Safaricom IFAD and lenders such as Central Bank of Africa use firms such as Cignifi and Experian to produce consumer-risk profiles. Human capacity will need to be built (both in-house and among third-party firms) in order to realise the value of the data being collected today (see more in Chapter 6). Another challenge is that existing datasets are often closed. Despite donor-led efforts to create more open data public goods, there is not yet significant momentum around (or use of) these resources. Greater scale implies more widely shared data. Policymakers and lawmakers will need to make data decisions that are democratic, support the benefits of big data and still protect privacy. As we begin to share data more frequently and widely – between public and private actors and between different countries – laws will need to adapt quickly to ensure that users (in this, case farmers) can (1) consent to how their data are being collected and used; (2) access the information themselves, bearing in mind the digital literacy challenges that exist in many parts of the world; and (3) trust in systems to protect their security and privacy. Achieving these objectives will be significantly more complex than it was before the digitalisation of data, not least because now vastly more stakeholders are involved in collecting, analysing, and using this information. 138 CHAPTER 4 NEW TECHNOLOGIES POWERING THE PATH FORWARD Machine learning Unlocking the predictive capabilities of data by automating learning Machine learning is the application of artificial intelligence to allow systems to learn and improve themselves without explicit programming. If IoT is enabling the capture of billions of farm-level data points, machine learning is enabling the analysis of these data to improve automatically and continuously. As enterprises capture increasing amounts of data, machine learning can help them automatically improve the level of tailoring and precision of insights for specific smallholder farming segments, value chains, and geographies. There is also hope that machine learning may help solution providers overcome digital literacy challenges without solely relying on extension agents, e.g., through the use of interactive voice response (IVR) systems and chatbots. Machine learning solutions are currently at an even earlier stage than IoT. The reason is at least twofold. First, machine learning requires thousands Charlie Pye-Smith, CTA of data points for computers to build accurate algorithms, and the system needs to be fed with new data regularly to continue to improve its accuracy. As we discussed above, those data points are just starting to emerge in Africa, so progress in IoT and big data will fuel progress in machine learning. There have been some experiments to test solutions built for other markets (e.g., the US) in Africa, but those solutions have often fared poorly in initial trials and needed more local information before they were sufficiently accurate in the local context. Second, the talent required to build machine learning capabilities is significant (more so, even, than for big data analysis); as we discuss in the next chapter, the IT talent shortage in Africa is already acute. Machine learning also comes with important risks, i.e., because the underlying algorithms themselves may be biased or there may not be sufficient data on a specific segment, machine learning may not always offer the best solutions for specific communities. This can often be hard to notice or correct because the machine learning algorithms are rarely transparent. A number of machine learning experiments with agriculture are already underway. For example, Apollo Agriculture in Kenya uses agronomic machine learning to deliver customised and immediate advice to smallholder farmers. Farmers are able to call a local hotline and, through a conversation with an intelligent and interactive robotic system, access information about daily market prices, use of fertilisers, and expected crop yield. Even though it is still a challenge to set up an IVR system that is fluent across multiple local languages, the system has already enabled Apollo to improve its service offering in selected CHAPTER 4 139 regions of Kenya. Another interesting application of machine learning is WeFarm which uses machine learning and the power of the crowd to source the best answers culled from the platform’s network of more than 1.3 million farmers in Kenya and Uganda. Wefarm’s network allows small-scale farmers to ask each other questions on anything related to agriculture and then receive bespoke content and ideas in response. Wefarm’s machine learning algorithms then match each question to the best suited responder. Elsewhere in Africa, AI-enabled solutions are helping farmers combat plant pests and disease, likely the most mature application of machine learning in the D4Ag sector at this stage. For example, the app known as Nuru was crafted by taking thousands of photos of infected leaves. After experts diagnosed the diseases, the photos were organised into a database, which was used to train the software using machine learning to recognise the symptoms. The app is user- friendly, and farmers or extension agents simply point their smartphone camera at several cassava leaves and Nuru responds with a diagnosis. It can also work offline, getting around the challenge of limited connectivity facing many farmers. In terms of effectiveness, its developers say that the app is now twice as good at detection as extension workers.348 Similarly, Plantix, by Berlin-based PEAT GmbH, uses neural networks to diagnose plant pests and diseases via image Mergdata Services recognition. Plantix’s machine learning algorithm detects over 400 plant diseases, pests, and nutritional deficiencies and uses a learning data set of several million plant images crowdsourced from smallholder farmers. The application has over 700,000 users monthly, and is currently primarily India-focused, but already has North Africa pilots and plans for Sub-Saharan Africa entry.349 The growing success and scale of solutions such as Apollo Agriculture, WeFarm, Nuru, and Plantix, helping to draw more resources and attention to machine learning in agriculture. Four of the five innovation grants distributed in 2017 through the CGIAR Platform for Big Data in Agriculture (via its Inspire Challenge) went to machine learning projects, including pest and disease monitoring solutions and improved advisory services. As we explore later in the chapter, big tech players like IBM and Microsoft are also making major investments in machine learning for agriculture. 140 CHAPTER 4 BLOCKCHAIN Optimising for transparency, efficiency, and safety In the agriculture sector, blockchain can be applied to a wide range of use cases. At the most fundamental level, blockchain can help provide farmers with secure, portable digital identities. Using those digital identities, organisations working with farmers (from non-profits to commercial enterprises) can help create a digital footprint for farmers that includes their transaction history and a registry of their assets. This footprint, in turn, helps farmers prove that they are who they say they are, and opens the door to a range of services (particularly financial services) that they might otherwise be unable to access. Blockchain technology can also be used to trace the production and transaction journey of agricultural inputs and outputs. This provides more certainty and builds trust at each point of the supply chain, so that farmers can be confident that they are actually receiving the high-quality inputs – like seeds and fertilisers – that they are paying for. Blockchain can also help providers who are serving smallholder farmers. For example, blockchains ensure that every transaction Pasko Maksim, Shutterstock within the supply chain – from the movement of a crate to the payment from buyer to farmer – is tracked.351 These data can be used by agribusinesses and others to better understand their supply chains and take action to improve The transparency at the heart of efficiency and effectiveness – ultimately lowering costs. blockchain technology can make systems more efficient, actors more In addition, Blockchain has the potential to transform support services that farmers rely on, such as banking. For example, by making verification easier, accountable, and products and the technology can facilitate lending to farmers, insurance and other financial transactions more traceable – as services.352 At a more systemic level, blockchain could also help to quickly everyone interacts with a peer- identify the source of disease outbreaks in farming produce. A greater level of to-peer network that records all transparency would also allow buyers and sellers to work more directly with each transactions and is not controlled other rather than through intermediaries, leading to efficiency savings. by a single actor.350 Several promising initiatives are beginning to demonstrate the power of blockchains in agriculture. Blockchains are being integrated into D4Ag market linkage and supply chain management solutions to improve value chain trust and thus to maximise the uptake and ‘stickiness’ of farmers and other value chain intermediaries on such platforms, while also reducing transaction costs and speed for anyone attempting to monitor, back-trace, and verify underlying transactions. The most ambitious example of blockchain use in this context is Cellulant’s Agrikore product, which aims to register millions of agriculture value chain CHAPTER 4 141 intermediaries, such as farmers, agro-dealers, input producers, bankers, logistics companies, and warehouse receipt operators into a single transparent blockchain- based ecosystem. Users can make use of blockchain technology to transact at a low cost and with high levels of trust; in addition, the platform facilitates supply chain logistics management, traceability and access to finance for farmers as all contracts and transactions are recorded in an immutable system. Hello Tractor relies on a blockchain solution, developed in partnership with IBM, to provide a tamper-proof record of demand-side and supply-side processes ranging from tractor booking requests, to order fulfilment, payments for tractor services, distribution of proceeds to the tractor owners on the platform, and invoicing to farmers. The platform thus serves as a blockchain-enabled supply chain, finance, and logistics management ERP system. Tulaa utilises a blockchain-enabled system to track input and off-take supply chain logistics with its farmers, e.g., using the blockchain to validate hand-offs at key points across different value chain players to prevent agri-input fraud and ensure ultimate product quality. Finally, the University Cambridge Institute for Sustainability Pattyariya, Shutterstock Leadership and a corporate consortium recently deployed a solution that uses blockchain to follow the path of tea and wood products from Malawi to Sainsbury and Unilever.353 Another application for blockchains is to provide farmers with immutable identification. BanQu, based on Ethereum and tested in the Democratic Republic of Congo and several other African countries, is one notable example. It allows farmers to use their mobile phones to record their personal information and transaction history, which are then verified by a network of friends, family, and agribusiness partners. AgriLedger and AgUnity provide unique identities for farmers on their platforms and register individual transactions. This allows farmers to work in an atmosphere of trust with farmer cooperatives while also developing a ‘bankable’ transaction record that is immutable and can be made accessible to financial institutions with the farmer’s permission. In another variation on using blockchains for identification, the Government of Rwanda has teamed up with Microsoft and Wisekey, a global cybersecurity company, to create digital records of the country’s farm land registry that cannot be tampered with. The most common use of blockchain today in African agriculture is to help facilitate the speed and lower the costs of payments. Cellulant’s Agrikore was already mentioned in this regard above. 142 CHAPTER 4 BLOCKCHAIN Twiga, another example, is partnering with IBM to use blockchains to manage its loan application process for its retailers and farmers. Blockchain makes the application easier, faster, more transparent and – as a result – somewhat more affordable for counterparties to access financing. Dodore’s Agri-Wallet is a digital wallet and financial tool that creates a business account for farmers on the back of a blockchain platform.354 As farmers earn revenues, they can be paid either through M-Pesa or through blockchain- tracked tokens, which can be used to purchase inputs from vetted vendors who participate in the programme. These tokens and related blockchain verifications are then used as a form of collateral; lenders like Rabobank are willing to provide loans against the tokens in the absence of more traditional collateral. In the cross-border agiculture payments context, CropCrowd, a crowdfarming site, uses a blockchain platform to receive crowdfunding investments and to process payments back to international investors without the need for difficult and costly (or sometimes impossible) currency conversion transactions. Similarly, San Francisco-based Veem is being used by international buyers to pay farmer suppliers in countries across Africa and Asia. The Veem automated platform uses Jim Sabogal blockchain to convert payments from the source currency into the local currency in more than 80 receiving countries; it cuts payment time in half and reduces payment costs from as high as 12% to approximately 2%. For blockchain to work in developing countries, data will need to be digitised, standardised, and checked for accuracy. Most data in Africa continue to be paper-based. The trajectory of big data, IoT machine learning, and other innovations will likely determine the extent to which this remains true – and each of these technologies faces its own scale-up challenges. Once data are recorded in the blockchain ledger, they cannot be changed, so it would be essential to avoid the influence of corruption and fraud before this stage. This may prove a formidable challenge in many Sub-Saharan Africa countries. Blockchain’s future, then, rests on the ability and willingness of countries to tackle widespread governance challenges.355 Moreover, in order for blockchain to work, everyone in the ‘network’ must use the same technology, which often comes with verification structures and other auxiliary items.356 This standardisation brings high initial costs.357 CHAPTER 4 143 Increased investment Donor activity will come from donors, Based on Dalberg’s earlier analysis of D4Ag private investors, and donor flows for the BMGF, as of 2015–2016, an estimated €85–100 million annually in large corporates donor money was flowing specifically to D4Ag Based on current trends, we predict initiatives in Africa. We estimate that this that both donor and private capital number grew to €175 million by 2018, based flows to D4Ag solution developers and on estimated self-reported funding figures implementers in Africa will accelerate collected from top 15 Africa D4Ag funders significantly in the next few years. globally. These estimates exclude broader Current trends in donor and private D4Ag donor investment in connectivity and ICT investments suggest a clear upward trajectory; access or funding for small digital components our interviews with key stakeholders were (e.g., digital M&E tools, remote sensing costs) unanimous in supporting this projection of of large agriculture projects, a decentralised significant increases in funding and investment spending item that could be substantial but for volumes. Whether such increases will be which no data are available. enough to meet the needs of the D4Ag sector is an open question, however; much depends As can be seen from the figures above (and on the evidence that D4Ag players are able taking into account the very directional nature to muster for the impact and business model of all of such numbers), donor funding for sustainability of their solutions. D4Ag appears to have grown by 15–30% annually in the past few years – a pace of The total amount of ‘needed’ investment is growth that, anecdotally, felt accurate to the impossible to estimate at this stage given gaps donors we interviewed given the general rise in data and the infancy of D4Ag business in attention to D4Ag in the past few years. models, but the amount is certain to be in Using mid-range estimates of donor spending, the hundreds of millions of euros today and there were two donors who consistently spent trending toward €1 billion the next 3–5 years more than €20 million annually on the based on historical trends.358 sector, four who spent €10–20, six who spent €2–10 million and a longer tail of actors who Furthermore, it is important to distinguish provided D4Ag grants. The specific funding between the funding needs of individual figures have been anonymised at donor D4Ag players and the need for public good request, but the top donors in the sector (to the investments into D4Ag infrastructure, which best of our knowledge) appear in Figure 31. we have not been able to quantify precisely in this report, but which are likely on the order of several billion euros.359 While funders and “We recently introduced a digital-by-default policy investors may be able to meet the needs of leading individual African D4Ag enterprises across all our sectors and all our countries. We have in the coming few years, it is almost certain asked each project lead to think in terms of digital that there is insufficient funding in the first when conceptualising a new project, and to pipeline for D4Ag infrastructure public goods. Rough estimates put this gap at greater than thoroughly justify any reason not to choose digital. €1 billion.360 Representative of a leading donor ” 144 CHAPTER 4 Figure 31 Estimated annual Sub-Saharan Africa D4Ag funding, 2018 € millions, Sub-Saharan Africa, 2018 Top global D4Ag funders €300 €175 42 Earned 42% revenue 35 20% 18 16 Donor 20% grant 58% 16 funding 11 9 9 6% 7 7 443 Individual donor contributions 54% (data confidential) 200 Donors are likely to continue to grow Lately, there has been some shift in donor their investments in D4Ag in the near interests towards ecosystem building and 150 future. All the foundations, development investing in D4Ag public goods like data banks and multilateral agencies interviewed systems, AgTech incubation/acceleration suggested that they were likely to increase ecosystems, cross-sector data, data analytics, 100 their D4Ag investments in the coming 3–5 and knowledge partnerships. However, the years, but most were unable to provide specific focus on public goods and enablers is still funding commitments or targets as so many relatively new; we believe this is a critical area 50 donor D4Ag strategies were still in flux at the for future focus, as we will discuss in Chapters time of this report’s completion.361 In most 5 and 6. cases, they were looking to see results from 0 current investments before they made major Figure 33 provides an overview of major donor additional public commitments in the space. activities and priorities in D4Ag based on Thus far, most donor investments have been publicly available information and interviews. in specific D4Ag projects and solutions, with the possible exceptions of the World Bank’s Note: uses mid-range estimate for annual earned revenues (iP.e.r, iUvSDa 1t5e7 mcilliaon)pital climSaoteu-rcsme: aDratl baegrgri caunaltluysries, spuorrvtfeoililola anncaely syes tfeomr 5s major fundTehrse, sealmf-reopuorntetd oesft impartievsa btye D s4eAcgt/oICrT 4cAagp iletaadl s, interviews and agriculture data observatory investments flowing to D4Ag enterprises remains and the BMGF’s portfolio, which has always small, but has recently increased had a substantial share of D4Ag public good dramatically. We estimate that about €47 investments such as, historically, investments million of PE/VC investment flowed into into the Africa Soil Information System (AfSIS) D4Ag in 2018 (Figure 32).362 While this and GODAN and, more recently, Innovative figure represents a tenfold increase over 2016 Solutions for Decision Agriculture (iSDA) and and a nearly fourfold increase from 2017, it national D4Ag data systems in Ethiopia. still constitutes a small share (<16%) of the CHAPTER 4 145 Figure 32 Value and volume of VC/PE investments into the African D4Ag sector € millions and number of transactions 47 D4Ag investments represented ~16% of the €298M+ invested in African tech start-ups in 2018. African D4Ag investments represented only ~2-7% of the ~€620M – 2.04B invested in AgTech start-ups globally in 2018. Source: Dalberg/CTA D4Ag investments tracker, Disrupt Africa, Pitchbook and AgFunder, FT AgTech investment tracker. 13 8 5 1 3 Before 2014 2014 2015 2016 2017 2018 1 6 3 8 10 18 number of deals €335M flowing to tech start-ups in Africa in Big tech activity 2018.363 Just two companies – Twiga Foods The entrance of big tech firms will and Gro Intelligence – received nearly 2/3 of advance the data revolution in new the funding to D4Ag enterprises in 2018. An ways. Big tech firms see new opportunities additional 10+ smaller enterprises – including for themselves to play a positive role within Ignitia, Tulaa, and Cowtribe – were able to this data-driven approach to agricultural raise significant amounts of seed or Series A transformation. Some players may want to funding, ranging from €270,000 to €900,000. better understand the space itself – given More than 60% of the deals were equity based. that the majority of Africa’s over 1.2 billion people work in agriculture, understanding Despite this growth, private investment the agricultural labour force better will in D4Ag remains nascent. Total investment provide big tech actors insights into a massive of €47 million is minimal relative to the need, potential user base, one that has historically and represents only a small fraction of private been harder to get to know. Other players capital flowing into AgTech globally, estimated hope either to sell their products (e.g., cloud at approximately €1.8 billion in 2017 – a storage) or provide technology-related services roughly 30% increase over 2016 – before – from analytics services to human capacity levelling off in 2018.364 Mainstream investors building – to agribusinesses and commercial still see most African countries – with a few enterprises. Still others may simply see exceptions – as relatively risky. Those who are value in experimenting with the extent to investing in Africa tend to view FinTech and which technology can transform agriculture. perhaps InsureTech as more attractive sectors These actors invest heavily in research and than D4Ag, which shares many of the same development and are capable of launching underlying risks but is characterised by even cutting-edge applications. lower levels of regulation and greater access issues in rural areas, among other challenges. 146 CHAPTER 4 Figure 33 Major donor activities in D4Ag Donor Approach to D4Ag investment European The European Commission is a major funder of agricultural transformation in Africa and, based on our estimates, the Commission (EC) top funder of D4Ag programmes in Africa across a variety of national and regional projects. The EC has a broad set of objectives for D4Ag, which cut across different EC (DG DEVCO) units involved (i.e., Sustainable Agriculture, Digital4Development), with a primary focus being to promote D4Ag programmes and solutions that strengthen food and nutrition security, and advance the climate-smart agriculture agenda, while also contributing to sustainable development and job creation in Africa’s agri-food sector and rural economy. The EC is the principal funder of CTA, which operates within the framework of the ACP-EU Cotonou Agreement. The EC also supports a number of projects in D4Ag across the continent via country delegations, ranging from digitally- enabled advisory services to market linkages to digital financial services and innovative climate-smart agriculture programmes focused on the use of remote sensing, drones and weather surveillance systems. Bill & Melinda BMGF’s major priority is agricultural transformation, with an emphasis on smallholder farmers. The foundation has Gates Foundation a multi-billion dollar agriculture development portfolio of which a small but substantive portion is focused on D4Ag (BMGF) solutions and agriculture data projects. Since 2008, BMGF has spent over ~€400 million on D4Ag grants, typically averaging 5–15 D4Ag grants annually, with a focus on both global D4Ag public goods and country-level D4Ag programming centred on India and three countries in Africa (Ethiopia, Tanzania, and Nigeria). BMGF has maintained an ongoing commitment to the D4Ag sector, releasing new ICT4Ag and DFS for Agriculture strategies in 2017–2018 and continuing to grow its portfolio across digital interventions with a particular focus over the past year on Ethiopia’s D4Ag ecosystem, digital agriculture data public goods (e.g., iSDA and Agronomy to Scale initiatives), and a range of digital financial services and market linkage grants. The Foundation’s D4Ag programming is driven by its Digital Farmer Services team, which believes that digitally- enabled innovations in technologies, services, and platforms can rapidly increase the ability to scale and provide farmers with diagnoses of soil health and crop nutrition, access to financial services and inclusive markets, and learning opportunities to inform farm planning and practical field operations. The Foundation’s priorities include playing a strong catalytic role in advancing cost-effective D4Ag business models and supporting national/state-level D4Ag platforms. Dutch Ministry of The Dutch Ministry of Foreign Affairs prioritises D4Ag activities highly through its funding of the Geodata for Foreign Affairs Agriculture and Water Program (G4AW) and other country-level activities that sit at the intersection of food security, (Dutch MFA) water, climate sustainability, and digital for development. G4AW’s mission is to “improve food security in developing countries by using satellite data.” To this end, G4AW “promotes and supports private investments for large scale, demand-driven and satellite-based information services” and “provides a platform for partnerships of public organisations, research institutions, private sector operators, NGOs, farmer cooperatives, satellite data/service operators, business and transmission operators.” G4AW works via a number of partners in Africa and Asia. For example, G4AW has partnered with Alterra in Ethiopia on CommonSense, CTA in Uganda on MUIIS, SNV in Mali on STAMP, and Rainforest Alliance in Ghana on SAT4Farming – and a number of other D4Ag solutions in our database. Syngenta Syngenta Foundation’s mission is “to create value for resource-poor small farmers in developing countries through Foundation for innovation in sustainable agriculture and the activation of value chains.” Digital is not the central goal of their Sustainable investments, but rather a means to an end of helping farmers. Nevertheless, the Syngenta Foundation has invested in Agriculture (SFSA) a number of digital solutions – using its standard “pipeline approach: proof of concept, scale-up, handover.” The Foundation’s new D4Ag strategy is premised on the beliefs that (i) digital is an enabler, and not a solution in itself; (ii) agriculture field forces must be equipped to drive agriculture sector change; and (iii) commercial viability is key to driving innovation. The Foundation believes that the time is right to accelerate the use of digital tools in sustainable agriculture and that such solutions can dramatically reduce the costs of engaging and supporting smallholders, as well as better integrate a complex web of value chain stakeholders. To this end, SFSA is currently focused on supporting the better understanding and analysis of D4Ag business models, promoting the development of holistic and commercially viable D4Ag solutions that arm field forces with the tools they need to deliver value to farmers, strengthening the agricultural financial market through digital tools and approaches, and ensuring wide collaboration and good governance across the D4Ag ecosystem. CHAPTER 4 147 Figure 33 Major donor activities in D4Ag (continued) Donor Approach to D4Ag investment Deutsche With expertise in both sustainable agriculture and digital technology, GIZ has invested heavily in recent years Gesellschaft für in developing the digital side of its work on agriculture. Through its central team and country-level programmes, Internationale GIZ has worked on most D4Ag use case areas covered in this report, with a particular focus on digitally-enabled Zusammenarbeit information and advisory services, including market and climate-smart agriculture information services, digital input (GIZ) GmbH and off-take market linkages, and digital supply chain and logistics management tools. In 2018, GIZ launched both a blockchain lab and data lab, contributing to efforts around data for development and, in particular, the SDGs. Additionally, GIZ is a signatory to the Principles for Digital Development and aims to add value to the D4Ag space via sector coordination. While a good deal of GIZ’s work in the D4Ag space is focused on public good creation, new D4Ag tool development for specific projects, and innovative business model pilots, GIZ is also focusing on broader private sector partnerships to develop and promote economically sustainable approaches to D4Ag solution scale-up. As an example of such work, GIZ has partnered closely with SAP on several D4Ag projects that ultimately contributed to the development and roll-out of SAP’s Rural Sourcing Management platform. World Bank The World Bank Group is a leading global financier of agriculture, with $6.8billion in new commitments to this topic globally in 2018, typically through large multi-year national or regional agriculture transformation programmes. Very little of the Bank’s annual funding is explicitly earmarked for D4Ag overall or D4Ag in Africa, but digital and technology components are embedded in many programmes (80%+ of WB agriculture projects). In 2017, the Bank formed an internal community of practice with a focus on digital agriculture, particularly digitally-enabled extension services. The Bank also produced a major ICT4Ag report in that year. In 2018, the Bank began to develop a disruptive technology for agriculture strategy and formed an expanded central team to address this topic. The Bank’s Africa AgTech strategy (which goes beyond D4Ag to include other topics like off-grid energy for agriculture) is being finalised in mid-2019, building on the launch of a Disruptive Agricultural Technology Challenge and Conference in Nairobi in April 2019. The Bank’s new strategy will focus on supporting the development of AgTech incubation ecosystems across the continent, supporting AgTech entrepreneurs, and, critically, linking AgTech innovations to large Bank agriculture transformation programmes at the country level to ensure farmer impact, starting with Kenya in 2019. Key areas of D4Ag focus include digital solutions for agricultural productivity (advisory services, mechanisation, input linkages), market access, financial services, and data collection and agricultural intelligence. USAID USAID has been a long-time thought leader on the topic of ICT in agriculture. Until 2018, USAID’s work on this topic was coordinated by a Digital Development for Agriculture Team within Feed the Future, which focused on advancing the knowledge agenda on topics such as the use of data for agriculture, digital financial services for smallholder farmers, AgTech innovations (remote sensing, drones, field sensors), case studies of digitalisation business models, and overall tracking of D4Ag impacts. In support of this mission, in 2016, USAID launched an annual DC-based ICT4Ag summit that remains one of the central global events for this sector, with a global agenda but a strong Africa focus. Country-level D4Ag programming at USAID is highly decentralised at the mission level, with limited central visibility into D4Ag spending, project-level tools, data, and partnerships. In 2019, USAID is developing and launching a new ICT4Ag strategy under the leadership of a small central team that will focus on the following priorities: (i) understanding D4Ag trends and impacts (i.e., knowledge management and market intelligence); (ii) supporting effective use of D4Ag tools in the field (i.e., central D4Ag expertise function for USAID missions); (iii) working on innovative D4Ag data analytics projects with the USAID analytics division; and (iv) working with development partners to foster open, inclusive, and secure D4Ag data ecosystems. 148 CHAPTER 4 Big tech players currently seem to The impact of these large players on focus on 1) gathering various kinds of D4Ag will be significant. Given their in- agricultural data; 2) experimenting with house capabilities, reach, and wallets, big new uses of advanced technologies; tech players are capable of accelerating this and 3) partnering with other (often data-driven phase. Additionally, we expect local) organisations to do so. Big tech that their activity and investment will likely actors have deployed tools to assist with spur additional investments in other layers data collection – for example, IBM is of the ecosystem, such as connectivity and assisting Hello Tractor’s efforts to compile tech infrastructure. In some cases, big tech a transaction database while SAP is helping companies may be inclined to build out the develop farmer databases. Big tech players necessary infrastructure themselves (to some are also launching programmes that creatively extent, this has already begun to happen – use advanced technologies – in supporting much more is planned). This has the potential Hello Tractor, for example, IBM is using to create a virtuous cycle of improved tech blockchain, IoT, and IBM Cloud. A number infrastructure with greater reach, which will of other actors remain in test and pilot stages drive a greater number of users and more data of solutions that use advanced technologies, to better serve those customers. with launches anticipated soon. But it is important to note big tech’s Importantly, nearly all big tech activity in limits. These companies need to partner Sub-Saharan Africa’s D4Ag space involves with local players in order to respond partnerships with other actors, whether local to on-the-ground realities. Big tech can enterprises, agribusinesses, or NGOs. We equip enterprises to better serve farmers and are optimistic about this partnership model accelerate agricultural transformation, but this as it allows for a combination of expertise. support does not replace the need for very Overall, big tech players are making strong local talent. The capabilities of big tech significant inroads and could scale up pilot companies should instead be complementary programmes quite quickly. to organisations on the ground. Local players Clarissa Baldin, IFAD CHAPTER 4 149 are best positioned to understand farmer range (based on historic ratios between urban needs, design products that will serve them and rural unique subscriptions in Africa).374 well, and build business models that work in local contexts. By the same token, they may Unique subscriptions, however, likely lack the bandwidth or resources to complete underestimate the smallholder access to the more expensive, technical back-end work. phones. Country-level data from a handful Meanwhile, big tech players are well positioned of countries in Africa suggest that individual in terms of resources to do much of the smallholder farmer phone ownership is closer powerful processing. Therefore, partnership to 60% or more.375 Phone ownership at the will largely define success as advanced household level is likely even higher – closer technologies take off in D4Ag. The best models to 70% or more. There are still other ways will be those that pair localised knowledge with to measure access to mobile phones (e.g., big tech capabilities. Additionally, big tech percentage of farmers who have ever used a players have an opportunity to support human mobile phone, or percentage of farmers who capacity building themselves (e.g., training are able to access mobile phones outside of local teams on how to build and use artificial their home, rural 2G penetration). intelligence technology). Irrespective of the methodology, the critical Of course, the entry and scale of big point is that a large percentage of smallholder tech actors come with their own risks, farmers already have access to mobile phones including data breaches, misuse of data, and today, within their own homes. This figure is adverse effects on smaller and local D4Ag expected to continue to grow, e.g., GSMA enterprises. As such, their entry needs to be expects that unique subscriptions will grow to accompanied by thoughtful regulation. An 51% by 2025, likely 55%+ by 2030, and we additional risk is that proprietary technologies estimate that this will translate to nearly 80– could create walled gardens. We discuss risks 85% phone ownership at the smallholder farmer further in Chapter 5. household level, with the vast majority of these phones being smartphones by that stage.376 In The deep dive box in Figure 34 on the next addition, two-thirds of the total connection base page elaborates on specific D4Ag activities of a will be digitally connected through smartphones number of big tech players. by 2025, compared to just ~36% today. This means that not only will more farmers have An enhanced enabling access to simple feature phones, but also an increasing number will be able to engage with environment will fuel D4Ag solutions that rely on smartphones. substantial D4Ag expansion Continued improvements in phone Unreliable internet connectivity and high ownership will drive increased access to data prices will likely remain barriers in D4Ag solutions. the immediate term but private actors are racing to overcome them. For now, There are several ways to understand the challenges around connectivity and high smallholder farmer access to mobile phones, data prices confirm the continued relevance of and thereby access to D4Ag solutions. SMS/USSD solutions in the near to medium GSMA estimated unique mobile subscription term. Interviews have indicated that D4Ag penetration in Africa is 45% as of the end of enterprises are, in parallel, actively working 2018.373 Though difficult to quantify precisely, to develop applications that get around given that they predominately live in rural connectivity-related constraints for smallholder areas, the number of unique subscriptions for farmers (e.g., solutions powered by near-field smallholder farmers is likely in the 38–40% communication). In the medium term, we 150 CHAPTER 4 expect to see MNOs continue to invest in conjunction with the Internet Society), and expanding 3G and 4G coverage. Some major high-speed fibre optics (Google’s Project Link). telecoms in Africa have already begun to Overall, connectivity will become less of a explore and in some cases begin the transition barrier as the D4Ag market matures over to 5G, though this growth is expected to be the next decade. The sector will likely face a uneven and is still in its earliest stages. Finally, more practical issue of turning registration into there are several companies that are racing actual use – a challenge that we will discuss in to invest in expanding connectivity across Chapter 5. the continent, using, among other innovative technologies, satellites (e.g., Space-X’s Starlink Continued growth in digital payments initiative), balloons (Google’s Project Loon), access will increase and pave the way internet exchange points (Facebook, in for D4Ag enterprises to engage with Figure 34 Big tech making big waves in D4Ag IBM Microsoft IBM has partnered with a few of the most successful D4Ag Microsoft has entered the African market with a focus on precision enterprises across the continent, including a partnership with agriculture and AI technology. In collaboration with Techno Brain, Twiga Foods to establish a credit system leveraging blockchain Microsoft is working on a new Agriculture Data Platform in East technology. The programme is set to pilot among 220 retailers in Africa. Via Microsoft’s intelligent cloud system, the partners are Kenya, but if successful, IBM and Twiga Foods plan to roll out the seeking to collate data on rainfall, land type, and soil nutrition and platform to agriculture SMEs across Africa. During the first weeks create customised and wide-ranging farm management advice on of the pilot, the initiative extended loans averaging KES ~3,000 crops, harvest timing, and pest control. The project is expected to (€26.5) per beneficiary, which increased the profits of each pilot in Malawi and Tanzania in 2019. In addition, Microsoft’s retailer by 6% on average. FarmBeats technology, which uses IoT and AI to streamline farm operations, has moved one step closer to a public release of its IBM is also working with Hello Tractor in Nigeria to apply IBM’s innovations. Watson Decision Platform for Agriculture, blockchain, the IoT, and the IBM Cloud to Hello Tractor’s mobile app. The objective SAP is to capture an immutable record of all transactions from the first tractor request until the farmer has ploughed the field and SAP is currently focusing on applying its software technology returned the tractor. A database of transactions could improve the to develop comprehensive farmer databases and to connect efficiency and impact of Hello Tractor’s services. smallholder farmers to larger agricultural value chains. SAP has created a software system called Rural Sourcing Management, Going forward, IBM even plans to leverage image recognition which is designed to collect and share data on farm to determine the quality of the cultivation and to expand the characteristics and input/output transactions. service across Kenya, Mozambique, Senegal, and Tanzania. Most recently, IBM has entered into a major partnership with Yara In Nigeria, SAP is working with CBI Nigeria to integrate 850,000 to “build the world’s leading digital farming platform, providing small maize farmers into the agricultural value chains. In Côte holistic digital services and instant agronomic advice.”365 d’Ivoire and Ghana, SAP’s software has helped one of the world’s By the end of 2019, they plan to begin by offering hyperlocal leading chocolate manufacturers, Barry Callebaut, to develop a real-time weather forecasts along with actionable advice and supply chain management tool to onboard ~200,000 farmers recommendations based on weather data. While the partnership since 2016. And in Uganda, the company’s cloud-based solutions is global in nature, and initially plans to target Asia, Brazil have supported the efforts of Kalangala Palm Oil Grower’s Trust and Europe, Yara has said that it plans to reach African farmers (KPOGT) to improve the income of its 2,000 farmers. SAP’s “very soon.” software enables KPOGT to both communicate market prices for palm oil to its farmers and to inform local oil palm companies of when deliveries are expected. CHAPTER 4 151 smallholder farmers in a more cost- 15 countries, Paga has begun tapping into the effective way. According to recent GSMA Nigerian market, and the Bank of Kigali has data, 135 mobile money services supported added more than 1.5 million users to its mobile more than 120 million active accounts in money platform in Rwanda. Looking ahead, Africa in 2017, representing a growth of 18% these national initiatives will be accompanied compared to 2016. Much of this growth came by a new joint venture between MTN and in rural areas and will continue to do so in Orange (with support from the BMGF), called the years to come. While Kenya has long been ‘Mowali,’ which has the potential to reach an African leader – and world leader – in beyond the African powerhouses and extended mobile money, with solutions such as M-Pesa digital financial services to millions of rural and Equitel, over the past few years, MTN households across the continent. Mobile money has expanded to more than Google kinds of information (e.g., on soil moisture, weather, prevalent diseases) that can help farmers.370 TCS’s best-known solution is Google has partnered with ISRIC World Soil Information to mKRISHI in India, which receives questions from farmers via IVR, make soil maps widely accessible. The BMGF-funded Africa Soil and replies via SMS and IVR. This network is used as an advisory Information Services (AfSIS) project has released maps that predict information dissemination channel as well. Given mKRISHI’s “more than 20 soil properties at six standard depths at 250 meter success in India and TCS’s expansion into Africa, it would be a resolutions.” AfSIS created them with “new analysis, statistics, field natural step to launch a solution similar to mKRISHI in Southern trials and crowdsourcing.” The public can explore these maps for Africa, perhaps fuelled in part by the aforementioned analytics free via Google Earth.366 Furthermore, Google Maps and FAO are platforms.371 collaborating on climate change resilience and mitigation. Google has brought big data, cloud computing, and mapping capabilities Alibaba to the table and partnered with FAO “to make remote sensing data more efficient and accessible.” Satellites can track a host of Alibaba has already played an important role in transforming climate change-related metrics (e.g., deforestation, land usage).367 Chinese agriculture through its Rural Taobao business (profiled in Through its foundation, Google is currently exploring its options depth in Chapter 2) and other innovations such as ET Agricultural for engaging on African smallholder agriculture, but has no formal Brain, which uses artificial intelligence and machine learning programming announced at this stage. (using a combination of visual recognition, voice recognition, and real-time environment monitoring) to help farmers care for their livestock and crops. Bosch Alibaba has already made Africa a clear priority for its growth. Bosch’s technologies are currently helping support the creation of It has invested in several projects to help improve the ecosystem value-additive activities in different markets. Bosch’s packaging for e-commerce, including the Netpreneurs network (which is technology has enabled the growth of the processing sector building entrepreneurial capabilities on the continent), the new for coffee in Ethiopia and cassava in Nigeria, value that was economy initiative (targeting policy markets) and a partnership previously being left on the table.368 Looking forward, Bosch is between the Alibaba Business School and University in Rwanda evaluating the possibilities of big data and artificial intelligence to develop commerce-oriented curriculum, among others. In late in transforming agriculture. Bosch has begun to develop digital 2018, Rwanda joined Alibaba’s Electronic World Trade Platform applications that will allow algorithms to assess plants, insects, (eWTP), which “provides small and medium-sized enterprises and weeds (i.e., via photographs) and inform farmers on better with operational infrastructure, such as commerce logistics, input usage, agricultural practices, and likely much more.369 cloud computing, mobile payments and skills training.” These initiatives highlight a clear vision for how Alibaba plans to build the enabling environment and the level of importance it is placing on national-level partnerships. Given the company’s broader TCS aspirations in Africa, and its success with cutting edge D4Ag TCS has two agricultural analytics platforms that have expanded solutions in China, it is quite possible that the company will make or piloted in Southern Africa. These platforms compile various a major foray into D4Ag in Africa in the near future.372 152 CHAPTER 4 The expected increase in mobile More recent incubation efforts have focused money potentially will serve as an specifically on agriculture and agribusiness. For entry point for new digital solutions example, SmartHectar and enpact launched and is important to facilitating D4Ag an innovation hub for agriculture technology, transactions. In Rwanda, Kumwe Harvest food technology and water technology in West highlighted that it was able to drive down its Africa (based out of Ghana) in 2019. The transaction costs, as all the farmer cooperatives World Bank is in the process of setting up it worked with relied exclusively on digital a new AgTech incubator and accelerator in transactions. Going forward, mobile money Kenya as part of the broader WB Disruptive will enable more D4Ag enterprises to develop Technology for Africa strategy and is sustainable business models. considering replicating this approach in other African countries, such as Nigeria. A larger An improved start-up scene will likely and more diversified tech start-up ecosystem result in greater talent. Since 2016, the will likely bring improved technology and number of tech hubs across Sub-Saharan catalyse greater investments in local start-ups, Africa has nearly doubled from 239 in 2016 to including in agriculture. Equally important, a more than 440 in 2018.377 Equally exciting are richer ecosystem could bring in new talent and the players entering the scene: Google recently develop local talent. announced the first 12 start-ups participating in its Launchpad Accelerator in Africa, Adding it all up… Facebook has entered the Nigerian start-up Extrapolation from historical trends scene by partnering with CcHub to establish suggests that the the D4Ag sector could the new tech-focused NG Hub in Lagos and grow to nearly 100 million registered MTN ramped up its involvement in developing farmers by 2022. Our D4Ag survey local tech products and services with its Y’ello respondents self-reported that the number of Startup hub in Côte d’Ivoire. farmers registered for their D4Ag solutions CTA CHAPTER 4 153 grew 44% annually over the three-year into the realism of these figures, it is important Georgina Smith, CIAT period ending in 2018. Several of the biggest to note that these numbers refer to the overall D4Ag enterprises in Africa did not respond number of farmer registrations for D4Ag to the survey, but follow-up interviews with solutions, rather than the unique number of large players such as Digital Green, PAD, farmers registered for D4Ag, and certainly WaterWatch Cooperative, and Digifarm not the number of farmers engaged with or indicated that that this sort of growth rate is actively using such solutions on a regular basis. broadly in line with the overall African D4Ag sector and, if anything, is slower than the For the baseline, our analysis concluded that registered farmer growth rate of some of the there were ~33 million registered farmers market leaders. When asked for their three- today (see Chapter 3). Based on interviews year projections for the path forward, survey and smallholder survey data from countries participants reported, on average, that they like Kenya, we estimated 20% duplication expected an annual growth rate of 55% in (i.e., users registered to multiple D4Ag registrations through 2022. Large D4Ag sector solutions), which would mean ~26 million actors not included in the survey data each unique users today. Our database indicates reported plans to digitise low millions and in that roughly 42% of those registered for D4Ag one case tens of millions of smallholders over (or approximately 11 million farmers) are the next five years. ‘engaged’ to the extent that they have used the solution to even a moderate extent after Using the more conservative historical growth registration. Other users have registered but do rate of 44% leads to ~100 million registered not use the solution. farmers by 2022, or triple the farmers registered for D4Ag solutions today. Applying the 44% historical growth rate for registered farmers to unique users yields a The number of unique users actually projected ~80 million total unique users and engaged with D4Ag solutions is far more 33 million engaged unique users by 2022 modest in this projection. Before delving (see Figure 35). 154 CHAPTER 4 Even with such adjustments, the number of unique registered farmers to grow from still appears aggressive. It implies that ~54 26 million to 47 million in 2022. This would million farmers – 18 million per year – will be mean adding ~20 million farmers over the registered over the next three years, up from next three years, roughly the same absolute 11 million unique farmers registered in 2018, number of new farmer registrations as the pace leading to a total penetration rate of roughly of farmer registration over the past three years a third of all smallholder farmers in Africa by (2016–2018). 2022. In some ways, these numbers are not unprecedented – for example, Cellulant took We propose this 47 million farmer figure just a few years to register 17 million Nigerian as the very conservative scenario for farmers for its e-wallet as part of the Nigeria potential unique farmer reach and 80 SES subsidy scheme. In the absence of such million unique farmers registered in national schemes, however, this pace of farmer 2022 as a highly optimistic figure. acquisition appears hard to sustain. What these figures reveal more broadly, Even if one believed that the growth rate for however, is that farmer registration is D4Ag registrations was likely to slow down not the binding constraint for the sector. dramatically after 2019, an annual growth Looking forward to 2030, we believe every rate just half of what was seen in the past farmer with a cell phone will use at least one few years (22%) would still lead the number D4Ag solution. If we assume the number of Figure 35 Projected unique registrants and engaged users, 2019–2030 number of users, by year 2019 2022 2030 200M Unique (non-engaged) users Unique engaged users 116M We assume that penetration of D4Ag solutions among smallholder farmers will reach 80M 80%+ as connectivity improves and cell phone usage expands 33M 46M 7M 26M True challenge in 2030 will likely not be ‘reach’, but rather 84M 15M ensuring higher levels of 33M engagement among registered 11M Registered Double Unique Unique Unique users counted users users users 1 20% haircut to 2 42% engaged 3 44% historical user rate from growth rate from 4 Projected number of smallholder de-duplicate farmers based on UN the reach figure survey data survey data and Dalberg analysis CHAPTER 4 155 smallholder farmers in 2030 to be 250 million (i.e., the same as today) and the connectivity “These trends suggest that the next 3–5 years rate to be 80% (per the discussion in Chapter 3), we would expect around 200 million unique are likely to be transformative for D4Ag. users. Based on current engagement levels, we ” expect only ~84 million engaged users in 2030. The number of truly active users is likely much and impact. This growth will likely not occur lower – perhaps half of all engaged users based evenly across all segments of Sub-Saharan on our desk research and interviews. The Africa’s smallholder population, however. greatest challenge over the next decade will Smallholders, particularily men, in countries not be reach but rather increasing levels of with stronger enabling environments will engagement among registered users. likely enjoy significantly improved access to D4Ag solutions, while access for others These trends suggest that the next 3–5 may expand at a slower rate or – in certain years are likely to be transformative for environments – not at all. The ability of the D4Ag and will build the foundation for D4Ag sector to surmount such accessibility even more dramatic changes through barriers, particularly among more marginalised 2030. D4Ag success stories are just beginning populations, will depend on the concerted to emerge, and we believe the sector could go efforts of all sector actors to overcome the much farther – especially in use, inclusivity, D4Ag challenges outlined in the next chapter. Tamiru Legesse, FAO 156 CHAPTER 5 WHAT IT WILL TAKE TO ACCELERATE GROWTH AND IMPACT Image to go here CTA In our efforts to build a strong, foundational D4Ag ecosystem that will support sustained, inclusive growth, too much focus has been placed on experimentation and short-term success. Progress toward a strong D4Ag infrastructure because D4Ag will not eliminate ecosystem is promising, but the sector the need for it. Digital tools can improve still faces a number of challenges. market efficiency, transparency, aggregation, Some of the challenges are specific to the and integration, but parallel investments D4Ag ecosystem while others – e.g., national in physical infrastructure (e.g., roads and agronomy R&D systems, agricultural policies, electricity) are still needed to deliver inputs and rural land tenure – apply to agricultural to farmers, to deliver farm products to transformation more broadly. We discuss these markets, and to power production and post- challenges in this chapter. harvest agricultural equipment. Governments, donors, and others must invest directly in We do not address connectivity because the necessary non-digital infrastructure in order for broader market is already making significant agricultural transformation to occur. Similarly, progress toward overcoming this issue. the significant investment and ongoing costs Additionally, we do not address non-digital required for human infrastructure (e.g., “Successful D4Ag solutions are evolving faster than the ability of the enabling environment to support them. ” CHAPTER 5 157 extension agents, financial agents, and agro- input dealer networks) are crucial to achieving “49% of D4Ag enterprises surveyed for this report real agricultural transformation and impact. reported human capital as a key growth challenge. Four main challenges significantly limit ” the role D4Ag is currently able to play constrain the growth of D4Ag solutions in advancing inclusive agricultural on the supply side. Despite the efforts of transformation in Africa: (1) there is African-focused tech staffers like Andela and insufficient tech-savvy human capital to technology hub communities like Nairobi and support D4Ag solution development and Lagos, local skill development for software support, matched by the problem of low end- and product creators, data analysts, product user digital literacy, (2) the sector underinvests implementers, and monitoring support remains into D4Ag infrastructure, particulary enabling largely insufficient. Even in countries with agriculture data systems at the national level, more advanced technology ecosystems like (3) poorly calibrated government policies Kenya and Tanzania, one out of three firms hinder or fail to encourage D4Ag ecoystem described ‘inadequately skilled workforces’ as development, (4) companies still struggle to a key business constraint.378 In Kenya, where develop viable business models and (5) D4Ag nearly one out of five formal sector positions is growing unevenly across the continent. is ICT intensive, the agricultural sector may If we overcome these challenges, the sector struggle to attract and retain workers with could very likely grow faster and become more strong technical skills. Forty-nine percent of inclusive in the coming decade. surveyed D4Ag enterprises reported human capital as a key growth challenge. The failure Successful D4Ag solutions are evolving of private, public, and non-profit actors to faster than the ability of the enabling cultivate a large volume of workers with ICT environment – skills, policy, and skills can compound development challenges middleware – to fully support and take for D4Ag enterprises, particularly in markets advantage of them. While some enabling that struggle to attract funding due to their factors such as connectivity and mobile money small size or instability. have improved, others lag behind, even as recognition of their importance grows among In the absence of established donors and policymakers. In order to meet the start-up supports like prize demands of D4Ag, the enabling environment competitions, university incubators, must improve human capital, develop and and formal networks, local tech enact supportive agricultural technology policies, entrepreneurship in much of Sub- CTA and fund and build out D4Ag infrastructure, Saharan Africa remains weak. For particularly agricultural data systems, that will example, Senegal has not invested in the enable D4Ag solution scale-up and impact. development of local digital skills, and as a result, few D4Ag enterprises exist in the Insufficient human capital country. Moreover, because they lack access to continued funding and human capital, development among D4Ag the start-ups that do exist there struggle to creators and consumers advance, much less to succeed. Senegal is limits the range of solutions a relatively unattractive market for external offered and the uptake of private investors (see the Senegal case study in the Annex for more details) and without the ones that do exist investment in local skills the whole country’s The low concentration of refined ICT D4Ag space remains underdeveloped. Broadly skills in most African countries can speaking, in the absence of human capital, 158 CHAPTER 5 FAO local enterprises struggle to scale and, in their and intermediaries – to familiarise themselves place, foreign enterprises, likely with a weaker with useful technologies and share the benefits understanding of context, control what D4Ag with others. To reduce the need for such space exists. investments, 80-28 has begun to investigate how artificial intelligence and machine learning D4Ag enterprises report that low approaches might inform IVR solutions that levels of digital literacy and comfort overcome digital literacy challenges. However, among farmers and agricultural agents these approaches may not sufficiently tackle the constrain demand, adoption, and use digital literacy gap for another 5-10 years. of offerings. Architects of Ethiopia’s wide- reaching 80-28 programme reported that, Enterprises without the time and resources initially, users often did not understand how necessary to confront digital illiteracy may to dial the hotline number or cycle through find it difficult to grow, but actors can support call menus. Businesses around the continent digital education across multiple solutions. cite farmers’ lack of trust in phone-based CTA worked with enterprises to develop transactions as a key barrier to the adoption of a curriculum for user training, which was their market linkage solutions. Overall, 28% piloted by Farmerline in Ghana, Ensibuuko in of surveyed enterprises cite consumer- Uganda, and FarmDrive in Kenya. level barriers as a top-three challenge to D4Ag adoption and use. Gaps in D4Ag Enterprises with the time and resources infrastructure, particularly to do so have either invested heavily in in terms of under- digital education or sought to design investment into agriculture around literacy barriers. For example, Digital Green relies on a vast network of data systems extension workers to facilitate video displays Agriculture data ‘middleware’ for farmers, while 80-28 allots time for staff infrastructure – e.g., farmer registries, to respond to non-topical calls. Over time, digital agronomy data, soil mapping, these investments in digital education help pest and disease surveillance, and some farmers – particularly model farmers weather data infrastructure – enabling CHAPTER 5 159 layer for D4Ag solutions. The lack of and disease surveillance, livestock surveillance, agriculture data infrastructure in most contexts and advisory data systems. can significantly hamper D4Ag solutions, while the presence of high-quality agriculture Kenya recently partnered with the World data ecosystems can increase their efficiency Bank to build a national agroclimatic data and effectiveness. surveillance system, the Kenya Agricultural Observatory Platform (KOAP). The Ugandan Agriculture data infrastructure in the government with donor partners is working form of national farmer registries, for with Dalberg Data Insights, Dalberg’s data example, can play a highly useful role. science team, to build out and scale the Registries can support farmer identification CubicA platform, a set of big data tools and and verification, reduce the cost and effort data repositories (e.g., national scale crop maps of data collection, help simplify agribusiness and yield forecast maps) for monitoring key and government processes and inform policy agriculture and food security trends in the making. Comprehensive and regularly update country. Rwanda has a national agriculture national-scale digital farmer databases--such as data roadmap that goes well beyond the SNS Ethiopia’s input subsidy e-voucher and 80-28 farmer registry and is seeking to build other databases (4 million farmers), Rwanda’s Smart important agriculture data systems. Namibia Nkunganire System (1.5 million farmers), and eSwatini have invested heavily into Zambia’s ZIAMIS (1.15 million), and Nigeria’s national livestock traceability systems. partnership with Cellulant (17 million farmers at its peak, of which 7 million were receiving Government-led digital agriculture data subsidy payments) – can provide governments initiatives are, however, very few in and D4Ag enterprises with the necessary data number today. The vast majority of African to tailor extension services to farmers’ needs, countries lack the resources and the technical increase access to customised farm inputs and capacity to build comprehensive digital farmer strengthen value chains through increased registries, let alone more complex agriculture traceability and transparency. These types surveillance systems that feature remote sensing of government-affiliated initatives or social data layers, weather data, or soil data. Beyond enterprise farmer digitalisation plays like resource constraints, some governments CTA’s MUIIS solution in Uganda (250,000 discourage agriculture data infrastructure farmers), can also facilitate smallholders’ access development in response to legitimate data to financial services, including insurance, policy concerns (noted later in this chapter) or savings, and – most important – credit products, due to less valid considerations since increased by allowing smallholder farmers to formally transparency and availability of information register their farms. In addition, farmer registries may not always be welcome. Georgina Smith, CIAT of this type can provide a better understanding of D4Ag’s impact on women, youth, and other marginalised groups by tracking resource flows and outcomes at the individual level. Beyond national digital farmer registries, working closely with donors, a number of countries have launched efforts to build other types of agriculture data systems. Ethiopia, for instance, via its Agriculture Transformation Agency (ATA), is working with the Gates Foundation in 2019 on an ambitious plan to build out national pest 160 CHAPTER 5 As a result, D4Ag infrastructure Foundation is working on a Agronomy-to-Scale initiatives tend to be primarily donor-led (ATS) data platform concept, which would at the moment. A number of such initiatives build on iSDA’s soil data assets but develop are currently picking up momentum with a much broader Africa-focused geospatial regional or sub-regional lens. agronomy data sets and tools (e.g., crop maps). For high-quality soil data, for example, The World Bank and organisations like the the Gates Foundation has already invested UNDP have growing portfolios of investments extensively over the past decade into building across the continent into climate-smart out the digital infrastructure for soil data agriculture data systems and related ‘hydromet’ collection, analysis, and dissemination systems weather surveillance and early warning under the umbrella of its AfSIS programme, services, of which the Kenyan KOAP data which has now been transitioned into a new observatory, mentioned above, is one social enterprise, Innovative Solutions for advanced instance.379 Decision Agriculture (iSDA). The programme made extensive progress in generating national- Such endeavors demonstrate a significant level digital soil maps in partnership with opportunity for donors to better balance African countries such as Ethiopia, Tanzania, funding for specific innovations with Nigeria, and Ghana, where these maps are investments in public goods. Along with this increasingly being utilised by D4Ag actors good news, however, comes a general sense to build value-added tools and applications. from all of our sector interviews that such Through a recently launched partnership public good D4Ag infrastructure investments with the Islamic Development Bank, BMGF are very limited at the moment relative to is seeking to scale this national soil data the scale of the challenge and, furthermore, infrastruture to another 8+ countries in the are overly concentrated in just a handful of Sub-Sahara Africa region over the next African countries. several years. Government policies The CGIAR system, via efforts coordinated by that stifle innovation the CGIAR Big Data4Ag initiative, is in the midst of scaling up digital agronomy platforms, or expose consumers which include agronomic data repositories and to security risks hinder Georgina Smith, CIAT systems that track field trial data. The Gates inclusive D4Ag expansion Policy frameworks that stifle innovative approaches or fail to clearly stipulate regulatory requirements discourage D4Ag innovation and investment. D4Ag solutions rely on coherent business procedures, strong financial systems, and clear regulations of digital and data processes. Inconsistency in the interpretation or implementation of policies in these areas can disincentivise innovators and entrepreneurs from entering the D4Ag space. For example, Ethiopia’s conservative banking regulations forced mobile money operator M-BIRR to engage in five years of redesigns. This substantially slowed the growth of mobile money in Ethiopia. In Senegal, unexpected CHAPTER 5 161 Figure 36 An illustration of country-level D4Ag readiness Malabo Montpellier Panel’s country-level ranking MCI index score 70 Strength of mobile internet 60 Egypt Morocco Ghana 50 Kenya Cameroon Nigeria Zimbabwe Ivory Coast 40 Sudan Rwanda Benin Ethiopia Tanzania Senegal Uganda 30 Mali Liberia Zambia Burkina Faso Mozambique Burundi Malawi 20 Niger 10 0 0 1 2 3 4 5 6 7 8 EBA index score/regulatory framework Low EBA & low MCI Low EBA & high MCI High EBA & low MCI High EBA & high MCI changes to financial regulations forced mobile political climates. D4Ag enterprises have money provider Wari to shift business models seized upon farmer data as a viable revenue after an initial period of business model source. This has encouraged the collection success. and dissemination of increasingly specific pieces of farmer information – incomes, A lack of regulatory guidance can prove crops, vulnerabilities to climate change, soil equally discouraging to investment. A types, water access, etc. As a result, farmers, Rwandan agricultural drone company hesitated particularly those in politically volatile to expand into neighbouring Uganda due environments, are left susceptible to risks to Uganda’s lack of clear drone policies. ranging from unscrupulous business practices In Senegal, e-commerce platform Sooretul to violence. These risks are not unique to the struggled to formalise its business due to D4Ag space. CGAP conducted a study of 11 the absence of a policy framework. While digital financial service providers and each of policymakers may find it challenging to them experienced a cyber attack in 2017 that design regulations for emerging, experimental risked troves of customer data. Unlike digital technologies, doing so can attract investment finance, however, agricultural technology lacks and encourage D4Ag innovation. governing data standards and principles, rules around data sharing and selling, informed At the same time, the lack of policies consent, data security, and mechanisms for around privacy, security, and customer accountability and redressal, among other protection brings unique risks to protections. High-profile cases in Europe and farmers, particularly in less stable the US illustrate the dangers of leaving this 162 CHAPTER 5 work undone. As D4Ag evolves, such systems converting customer reach to actual use in must be prioritised more quickly, particularly order for these types of models to yield returns given the vulnerability of smallholder farmers and to achieve scale. and the risks of losing their trust should data or security breaches occur. Companies with business models that remain works in progress may Most companies are still deprioritise or miss important issues working to develop a like impact, data stewardship, etc. They may believe such issues are secondary to viable business model proving their business model. For example, While some companies have started several companies mentioned during interviews to reach scale and turn profit, the vast that focusing on women was too challenging majority of D4Ag enterprises still rely to make it an immediate priority. Donors can heavily on donor-funding. In recent play an important role in ensuring the right years, as discussed in great detail in Chapter balance between impact and business model 4, the D4Ag sector has learned a lot about viability, e.g., by incentivising a focus on use what models do not work, but we are still in and impact and targeting specific marginalised the relatively early stages of understanding segments. Similarly, donors might consider what models do work for most D4Ag use extending time horizons. Currently, most cases. For example, as noted in Chapter 4, investments are made with 3–5 year time experience from several businesses suggests horizons in mind, but realistically, impact will that farmers are unlikely to pay for D4Ag take longer to achieve. services (especially advisory services) and that data is quite challenging to monetise. As such, companies are beginning to experiment with Private investment may new approaches, e.g., taking a cut of the not be reaching the value created for customer segments, and in countries and segments many cases moving to bundled service ‘super platform’ models. This may have strong that need it the most ESADA promise, but companies will have to focus on High degrees of country-level and regional variation in investment expose uneven growth across the continent. While the progress in countries like Kenya serves as a strong inspiration for others, the level of variation across countries highlights some important challenges. First, it highlights that not all countries have sufficiently strong enabling environments in which D4Ag can thrive. For example, the Malabo Montpellier Panel’s recent report developed a country-level index to explore the variation in enabling environments across Africa using two primary criteria: the strength of regulatory environment and the ability to adopt and use mobile internet.380, 381 The report found a very uneven landscape overall, with most of the countries on the continent requiring a lot of support and enabling environment progress to truly move their D4Ag ecosystems forward. CHAPTER 5 163 But variations in investment patterns and volumes also indicate that donors, investors, and – to a somewhat lesser extent – enterprises are still risk-averse and likely prioritise the easiest-to-reach markets. This also occurs within individual countries, where companies largely target the easiest-to-reach customers. This kind of uneven growth results in uneven outcomes and could further the divide between the haves and have nots. Private investment often fails to target the poorest farmers, on whom D4Ag could have the highest impact. D4Ag solutions that attract investment tend to work through aggregators – including cooperatives, financial service providers, input providers, off- takers, MNOs, and others – that touch higher- which cooperatives and farmers they worked CTA income farmers in larger markets, despite with. Through such methods, aggregators the fact that lower-income farmers would help interconnect the otherwise fragmented benefit the most from these solutions. Data agricultural sector. suggest that farmers with access to financial services, cooperative memberships, and tight Financially viable opportunities for value chains fare better across a variety of aggregation often exist in large stable metrics than farmers outside these aggregator markets. Over 80% of the solutions that networks. Subsistence farmers, who have the received the most investment were active in the lowest incomes, lack access to such services. top eight most populous Sub-Saharan African Financial service providers are unlikely to countries (Nigeria, Ethiopia, DRC, South touch smallholders and women who would Africa, Tanzania, Kenya, Sudan, Uganda). likely benefit most from D4Ag. In contrast, Lesotho, Gabon, Guinea-Bissau, Mauritius, and eSwatini saw far less. There The most financially viable opportunities are a number of possible reasons for this utilise aggregators, particularly those discrepancy. For example, large stable markets in large stable markets. Since most have a larger potential use base and more enterprises do not charge farmers, aggregators expansive physical infrastructure. often comprise the largest revenue stream in financially sustainable business models. Increasing investment only in large For example, one enterprise that focuses stable markets could widen the on financial inclusion derives revenue from disparity between the poorest farmers charging financial institutions per farmer and those with access to aggregators. who uses the product and per loan given. It could further create regional divides or Aggregators can also provide a route through discourage regional integration between which to reach scale. For example, one small and large national markets. In a report notes that for ACRE Africa, “strong worst-case scenario, these inequities could lead partnerships with MNOs, input manufacturers, to community unrest, food insecurity, and and local agricultural vendors ensure scalability violence. As the sector continues to mature, of the product and wide reach of coverage at donors, investors and enterprises alike will a low cost of service.” A digital platform in need to work toward more equal access to Nigeria found farmers by asking agribusinesses D4Ag solutions across the continent. 164 CHAPTER 6 RECOMMENDATIONS FOR FURTHERING A SUSTAINABLE, INCLUSIVE D4AG AGENDA Image to go here Agrocenta Over the past 15+ years, the digital agriculture sector in Africa, mainly driven by donors, has launched a multitude of D4Ag enterprises and initiatives. Despite many failures and setbacks, these efforts have built a foundation of increasingly commercial D4Ag solutions – a growing number of which have promising business models and are starting to show meaningful scale. The D4Ag sector is still highly fragmented, however, the evidence base for D4Ag’s impact on smallholders is early stage for many use cases, and many other challenges to more rapid progress abound. Efforts of digital agricultural services Together, enterprises, donors, investors, to become sustainable and scalable agribusinesses, and governments must create continue to face challenges. How does the an environment in which digital agricultural sector transition from short-lived, donor-funded solutions can thrive and produce impact. In projects to self-sustaining, business-driven this chapter, we lay out seven priorities that initiatives that create demonstrable impact for will help the D4Ag sector succeed in a way smallholder farmers – and how does it do so that is impactful, sustainable, and inclusive. equitably? These are not wildly provocative investments CHAPTER 6 165 or ‘silver bullets’ for D4Ag. Rather, they are important foundational steps that will help “Together, enterprises, donors, investors, build a sustainable D4Ag ecosystem in Africa agribusinesses, and governments must create – one that can support the mainstreaming of D4Ag efforts going forward. Political an environment in which digital agricultural will, commitment, and engagement are solutions can thrive and produce impact. fundamental to the implementation of these ” recommendations and need to flow across those areas of the developer ecosystem most government institutions, not just agricultural capable of boosting human capital, i.e., start-up ministries. ecosystems, incubators, accelerators, etc. Efforts must also be made to increase the capacity of Much greater investment – on the order government workers in relevant ministries to of several billions of euros annually understand how to use and deploy D4Ag tools rather than a couple hundred million in various government initiatives. euros – is also needed. For instance, in the US, the government spends We recommend that governments: ~€1 billion annually, on top of billions • Invest in ongoing training to build the spent over the decades on underlying digital and D4Ag skills of individuals (from infrastructure, supporting the climate and legislators and ministers to IT leads and weather surveillance systems that provide local extension agents) throughout their essential services to the agriculture agricultural ministries and in other relevant community. In Africa, in comparison, ministries. investments into weather infrastructure are an order of magnitude lower in any • Implement farmer digital literacy and D4Ag given year for the entire continent. training programmes (with the support of the appropriate ministries, where applicable). In this chapter we focus on • Support the start-up ecosystem and recommendations for donors, investors encourage youth participation in incubators, and governments given they are the accelerators, and local university initiatives. primary audiences for this report. As with the prior chapter, we do not, herein, discuss • Participate in knowledge transfer important enablers that are not specific to programmes across departments and with D4Ag, like investments in rural connectivity, other countries. given how well understood and covered such efforts already are in other reports. We recommend that donors: 1. Develop human capital • Increase support for initiatives such as incubators, hackathons, prize competitions, at every level of the university classes, etc., to foster local digital D4Ag ecosystem skill development. Developing human capacity will be • Earmark funding for capacity building critical to building D4Ag readiness initiatives as a standard condition of grants across the ecosystem, from farmers to D4Ag enterprises. to government ministers. The necessary • Help create partnerships with D4Ag growth in human capital includes increased enterprises and non-profits experienced in awareness of D4Ag, improved digital literacy, digital literacy training. and greater digital skill building among smallholder populations. Such growth will • Offer technical assistance to government require deeper investment across Africa in capacity building initiatives. 166 CHAPTER 6 CTA We recommend that investors: We recommend: • Bring in developers from other geographies • Increased funding for a more diverse set of to share knowledge with and build skills business models rather than just for those among investees. models that have already attracted funding. • Support incubators and accelerators, • Greater focus on improved product design especially those with a strong focus on young and consortium/platform-based approaches entrepreneurs. to drive greater value for farmers. • Insist that investees incorporate strong digital • A continued push toward B2B models so literacy and consumer-training programmes that enterprises can attract paying clients. into their business plans. • Deeper research on D4Ag business models (see recommendation 6 for additional 2. Drive greater business details). model sustainability While a handful of companies are We recommend that governments: starting to see positive returns, the • Make direct investments in promising D4Ag vast majority still struggle to achieve models, where appropriate, in partnership economic and operational sustainability. with private investors, particularly for those Most start-ups are unlikely to succeed. While agriculture value chains where governments this is consistent with other sectors and in are already active in market support or other geographies, Africa needs to prove that public procurement. D4Ag deployments can be sustainable in order • Serve as paying clients for promising D4Ag to drive greater investment. solutions, especially at the proof of concept stage. Governments, donors, and investors can help achieve greater sustainability of D4Ag • Promote the creation of consortia that take a businesses. more holistic approach to value creation. CHAPTER 6 167 We recommend that donors: • Consider more flexible investment approaches (patient capital, innovative • Fund high impact studies on successful – funding models, etc.) that are better suited to and failed – business models and share best the needs of investees. practices. • Help build partnerships between investees, • Require investees to share and communicate private actors, and technology providers in financial results (anonymously as appropriate) order to reduce technology and operational with the broader D4Ag community. costs. • Share lessons learned and best practices • Share lessons learned and best practices from investees (anonymously, as appropriate) from investees (anonymously, as appropriate) with the broader D4Ag community. with the broader D4Ag community. • De-risk investments in high-impact models for investors through co-funding and increased grant/subsidy period of projects 3. Create greater impact to 5–7 years for products to be ready for market. by bringing D4Ag to • Promote bundling and consortium-based less-served populations approaches among investees. Today, D4Ag solutions primarily reach the lowest-hanging fruit – farmers in tight value chains – and many enterprises fail to prioritise We recommend that investors: outreach to women and other marginalised • Channel greater investments into D4Ag by segments. To achieve equitable growth, D4Ag building upon and scaling up viable models needs to be more inclusive. supported by donors. • Shift focus from companies that have already We recommend that sector actors: attracted significant investment to those • Offer greater support for enterprises in that have attracted less investment but have geographies that have historically attracted promising business models. less investment but enjoying strong enabling • Allocate greater funding for product design environments. and prototyping. V. Atakos, CCAFS 168 CHAPTER 6 • Incentivise D4Ag enterprises to target enterprises for the development of product marginalised segments, especially women, offerings tailored to the needs of women. who are systematically left behind. • Investing in gender-disaggregated data that both governments and enterprises can use We recommend that governments: to build more appropriate solutions and models. • Attract new investors by publicly supporting D4Ag and highlighting the benefits of local • Directly funding and focussing attention enabling conditions. on organisations in geographies that have traditionally received minimal funding. • Incentivise impact-oriented investments by entering public–private partnerships with • Shift expectations toward a slower return D4Ag enterprises that are committed to on investment than the typical three-to-five- impact. year window. With patience comes greater opportunity for these enterprises to reach • Prioritise and take into account the needs of beyond the low-hanging fruit. marginalised segments as part of their D4Ag investments. We recommend that investors: We recommend that donors: • Invest in promising D4Ag businesses even if they are not located in the most obvious incentivise D4Ag enterprises to engage the target markets. hardest-to-reach smallholder farmers segments, especially women by: • Support organisations that may be less known but that are equally as promising as • Incorporating gender targets as part of their those that have already received support. investment portfolios and explicitly fund grantees who prioritise women. • Consider incorporating specific impact metrics related to marginalised segments into • De-risking the cost of designing for specific their investment criteria. segments – e.g., by offering grants to Georgina Smith, CIAT IMAGE REQUIRED CHAPTER 6 169 • Take on the role of a catalytic investor that We recommend that donors: can help unlock funds for D4Ag in Africa from others. (Note: not all investors need to rebalance portfolios to include a greater share do this, but even a few investors taking on of investments in the D4Ag data infrastructure this role could have outsized impact). layer. Specifically, we recommend that they: • Fund investments in D4Ag data infrastructure alongside governments. 4. Invest in the missing • Offer technical assistance and advisory middleware infrastructure support to governments as they design and Successful D4Ag solutions require make use of D4Ag data infrastructure. access to a wide range of data (from remote sensing data to farmer-specific • Help identify strong implementation data) in order to deliver high-quality partners. services to farmers. This data needs to be • Share best practices from prior efforts. accurate, precise, and, in many cases, available in real time. However, it is neither efficient Investors, for their part, are likely to play a nor effective for each D4Ag enterprise to relatively smaller role in the creation of these individually collect, store, and analyse all the public goods. Still, they can help open new data it would like to access. markets by investing in ecosystem enablers while or even before making direct investments in We therefore recommend investments in a enterprises. robust D4Ag middleware layer that includes, among other items, farmer registries, digital We recommend that investors: agronomy data, soil mapping, pest and disease surveillance, and weather data infrastructure. • Partner with technology companies to build These public goods would immediately common solutions for their investees. impact side actors and could eventually • Invest in public–private partnerships (PPPs) benefit smallholder farmers directly. A strong, that offer revenue-generating (perhaps with coordinated effort – rather than one-off, small- the help of subsidies) public goods, e.g., scale efforts – by multiple actors is critical to weather services, soil and crop diagnostics, the success of such initiatives. etc. We recommend that governments: 5. Invest in good data • Make investments – in partnership with research agencies and donors – toward stewardship and design the creation of D4Ag data infrastructure for the risks and limitations and ensure that data about/for the most of digital systems marginalised groups is captured as part of The need for good data stewardship will these efforts. only grow. Actors in the sector increasingly • Deploy the data infrastructure for high rely on algorithms. As greater investment priority uses within their own efforts (e.g., flows into the middleware layer and as ever national soil cards). more significant volumes of data are captured, aggregated, and analysed, clear, conscientious • Promote open standards and modular standards will be necessary. systems so that other government agencies and other actors can plug into and use the new D4Ag infrastructure. 170 CHAPTER 6 We recommend the creation and similar to legislation but with shorter incorporation of strong D4Ag data lead-time. policies and practices across Africa. • Invest in strong data protection measures Data policies should incorporate the values and abide by their own policies as part of of good data stewardship (e.g., protections for their data infrastructure investments and digital ID, user privacy, etc.) writ large and data collection efforts. should span multiple sectors. Such values are exemplified by the emerging digital principles Donors can play an important advisory and for development and can be augmented with technical assistance role in these efforts. recommendations that focus specifically on D4Ag (e.g., farmer registry guidelines). We recommend that donors: • Help governments and legislators develop Governments must lead the way on strong data policies by offering technical assistance data stewardship efforts. and funding for such initiatives. We recommend that governments: • Consider the balance of risks and returns in data privacy/security regulation. Support • Work in conjunction with regional bodies to market development policies that ensure develop and enact strong privacy, security, consumer protection while managing the and consumer protection laws tailored to the downsides of overregulation. local context and in line with regional needs. • Familiarise government decision makers • Incorporate best practices and lessons with the issues and risks inherent to capacity learned from other geographies into the building and then assist their efforts to design and implementation of these laws. build actual capacity that attends to the • Work with technology actors to ensure that technological and legal aspects of data they understand and will abide by these privacy, data regulation, and cyber security. principles. Expertise in this field is often absent today. CTA • Consider developing shorter-term ‘codes • Share best practices and lessons learned of conduct’, which can achieve outcomes from other geographies. CHAPTER 6 171 • Invest in research that will promote the creation and adoption of good data policies. This could include behavioural research that explores D4Ag user experience and willingness to share data in order to establish a business case for company adoption of strong privacy practices. • Advocate for and promote greater transparency among enterprises to help fight against algorithmic bias against specific segments. Investors serve as stewards of good data policies. We recommend that investors: • Prioritise privacy and consumer protection as key elements of their diligence processes. • Help build shared infrastructure for their underserved and marginalised groups like FAO investees, e.g., through partnerships with women. It should be part of day-to-day cyber security firms, to help investees protect product design, so that enterprises build their data. solutions rooted in the needs and preferences of their customers. While this may sound 6. Invest in the D4Ag obvious, companies often overlook this step. research agenda • Better market and business model The D4Ag space is evolving rapidly. intelligence. Case studies on successful New approaches, business models, and ideas actors – e.g., how they were set up, their are continually being tested. Yet, broadly revenue models, the pivots they made speaking, stakeholders have focused more along their journey – will provide valuable on experimentation than on sharing insights insights into the key factors that drive and lessons. As the sector matures, there is success in D4Ag. Case studies on less a valuable opportunity to develop both a successful examples are equally important stronger set of indicators, best practices and and will allow the sector to also learn from lessons learned and a stronger community with shortfalls and mistakes. Similarly, we need which to share these practices. continued investment in market intelligence that regularly updates and builds upon the We recommend knowledge baseline developed in this report. investments in three major areas: • User-centric research and design. • Systematic research on impact. We Immersive, farmer-centric research will need more evidence about the impact on enhance the sector’s understanding of what the ground. Impact metrics should be more farmers want, how farmers are responding to standardised so we can make stronger existing products, what drives the adoption comparisons across use cases and business and use of such products, and ultimately, models. Also, in many cases, the evidence how offerings can evolve to increasingly needs to be more rigorous (e.g., driven generate value for farmers. This kind of by a third party, rather than purely in- research can help address the needs of house metrics). When collecting evidence, 172 CHAPTER 6 CTA it is crucial that we better understand the • Fund or co-fund investments in data contribution of digital vs other business collection efforts, especially those that involve model enablers in creating the impact in large-scale data collection at the level of question. individual farmers. • Integrate emerging lessons and findings into Donors should take the lead in advancing the their own plans and programmes. research agenda. We recommend that donors: We expect that investors will primarily be consumers of knowledge products, but they can • Fund the proposed knowledge initiatives still play an important role in generating and in conjunction with governments, D4Ag sharing knowledge. enterprises, researchers, and others as appropriate. We recommend that investors: • Facilitate sharing of best practices and • Fund or co-fund market-building research lessons learned. initiatives, for example by partnering with governments that test and bring new • Promote greater standardisation of impact technologies to market. metrics and data collection practices. • Contribute to broader sector efforts by Governments have an important role to play sharing (even confidentially) important in contributing to the research agenda. information about their D4Ag investments – including information that may not be We recommend that governments: public, e.g., amount and mix of funding and • Open their own databases for research strategic plans. purposes, especially as they invest in and • Transfer knowledge across and between expand their D4Ag data infrastructure. regions in which they work. CHAPTER 6 173 7. Create an alliance commitment to this initiative, inform its of key stakeholders mandate and priorities, offer resources for its operations, and serve as active participants and to promote greater contributors to its agenda and activities. They investment, knowledge should also back and support the priorities sharing, and partnership and recommendations of the alliance – where possible and in line with their own priorities building – and serve as champions for its efforts. The Strong leadership and improved success of similar alliances in other sectors, partnerships between sector actors are e.g., in health, highlights the promise of such needed in order for the opportunities an approach. identified in this report to come to fruition. Given the fragmented nature of We also recommend that the alliance existing initiatives, this is not likely to happen invest in building a deep membership automatically. Rather, D4Ag needs a strong base that is excited about its mission alliance and a knowledge clearing house to and offering. Beyond the core group of drive the sector. sponsors, the alliance will need to attract the We recommend establishing such a D4Ag interest of the broader sector: non-sponsors, alliance with the following key objectives: enterprises, farmer organisations, etc. These • Attracting greater investment in the D4Ag groups will play important roles as active sector, for example by supporting pipeline participants and contributors to the alliance’s generation and facilitation. efforts and will serve as consumers and • Facilitating deeper relationships and beneficiaries of its knowledge products and collaboration amongst D4Ag actors. convenings. • Helping connect various aspects of the In order to ensure its relevance for the ecosystem together, for example by linking sector, the alliance should maintain a agricultural technology innovation to big deep understanding of D4Ag, the needs technology players or helping link agronomy and perspectives of farmers, and the insights to various actors’ D4Ag efforts. priorities of the full ecosystem of actors, • Building knowledge and producing periodic especially regional and local priorities. reports about the state, progress, and It should incorporate those priorities as it challenges of the D4Ag sector. defines its mandate and should revisit these priorities on an ongoing basis so that its efforts • Developing capacity – especially among remain complementary to existing efforts on governments, farmers, and young the ground. As such, we recommend that the entrepreneurs – to realise the potential alliance be nimble in its approach and capable of D4Ag. of adjusting to the dynamic needs of the space. • Developing indicators for monitoring/ tracking progress and reporting to the key stakeholders through regular convening. “ For the alliance to be successful, we Strong leadership and improved partnerships between recommend a partnership between sector actors are needed in order for the opportunities governments, donors, investors and identified in this report to come to fruition. other value chain actors who are ” dedicated to advancing inclusive, sustainable D4Ag across Africa and beyond. Members must make a public 174 ANNEX 1 ANNEX 1 COUNTRY CASE STUDIES Image to go here CTA ANNEX 1 175 ETHIOPIA382 Ethiopia has shown that a state-led development model for D4Ag can deliver rapid scaling. In the long term, however, the sector will likely require greater private sector involvement to realise its potential. Key D4Ag statistics: Total users of solutions headquartered in Ethiopia383 5 million Number of solutions 4 (headquartered); 29 (with a presence) Proportion of users that are women384 17% Most common primary use case of solutions Advisory services Government role Sole operator. All solutions are government-provided. Snapshot of D4Ag solutions: Advisory services Market linkage Supply chain management Financial access Macro agriculture intelligence 176 ANNEX 1 Context: Agriculture The state of D4AG in Ethiopia in Ethiopia today More than 80% of Ethiopia’s population live Ethiopia’s state-led approach to the in rural areas, where agriculture serves as their introduction of D4Ag offers important main source of income. The sector accounts for advantages for scaling. Ethiopia established 45% of GDP, almost 90% of exports, and 85% the Agriculture Transformation Agency (ATA) of jobs. The vast majority engage in agriculture in 2010 as a strategy and delivery-oriented as subsistence farmers. The country’s main government agency to help accelerate the crops are coffee, pulses, oilseeds, maize, wheat growth and transformation of their agriculture and teff. Yields remain low, even by Sub- sector. Through the ATA, the government Saharan African standards. The government designs and in some cases implements has made tackling the country’s food insecurity interventions. The government is the sole a core development priority. As part of this distributor and price-setter of inputs to farmers effort, it introduced a series of reforms meant (e.g., fertiliser and seeds), and employs the to increase agricultural yields and put an largest network of extension workers in Africa. end to unsustainable farming practices that Ethiopia’s only mobile network operator, lead to environmental degradation and affect Ethio telecom, is state-owned, as are major agricultural productivity.385 D4Ag solution providers (others work in close partnership with the ATA). Ethio telecom CTA ANNEX 1 177 responds to Ethiopia’s digital and agricultural operators to enter Ethiopia’s telecom market. transformation agendas and helps them New rules will permit firms that are not 100% coordinate D4Ag efforts. A centrally-organised government-owned to issue SIM cards and approach to some degree also helps Ethiopia man operation towers. More such reforms encourage farmers to align their farming are needed to reshape policies and laws that practices and outputs. This leads to crop discourage competition. Government and intensification and efficiency gains through donor-backed investments have supported the economies of scale. rapid and substantial development of D4Ag in Ethiopia, but such investments will not likely Solution providers have developed be sufficient to build a competitive, sustainable effective ways to work around the sector in the long term. Policy reform could country’s digital challenges. The 8028 encourage more private operators to enter Farmer Hotline is a prime example. This the fray. This would increase the number of platform offers farmers free advisory services solutions on offer and, in turn, the breadth of via interactive voice response (IVR)/short products that farmers could access. message service (SMS).386 Three factors buoy its success. First, by utilising text and IVR, the service offers a much wider reach Lessons than internet-based solutions. In Ethiopia, just 4% of the population has access to the u Simple workarounds can circumvent internet and digital literacy among farmers is digital barriers to D4Ag scale-up. For nominal. To address these challenges the 8028 example, phone-based solutions can service invested heavily in agents and in the overcome low internet penetration, and deployment of its lines. They also developed low digital literacy rates among farmers a platform with information that digitally- can be counteracted by heavy investment savvy intermediaries can share via existing in agents who can address their queries. networks (e.g., extension workers, teachers, health workers, or just popular farmers in their u State-controlled D4Ag advisory services region). can help align farmer activities and in doing so, achieve economies of scale. Outlook u Central coordination of D4Ag scale- More flexible regulation could further up can help align digital and agricultural expansion of the D4Ag industry. So development agendas, as demonstrated far, the government has achieved impressive by the ATA, which controls all aspects results under their growth strategy. However, of digital agriculture in Ethiopia. This is at present, providers are barred by law distinct from countries like Senegal, where from charging farmers for advisory services. digital and agricultural decision-makers in Similarly, tight strictures regulate who can government work less collaboratively. provide financial services. This constrains the growth of mobile money in the country.387 More generally, businesses that offer D4Ag solutions via official channels report that the government’s deep involvement in the sector constrains private sector development. Recognising the limitations of public-only models, in February 2019, Ethiopia announced that it will privatise their state-owned telecommunications company and allow private 178 ANNEX 1 GHANA Ghana has created an environment that is well suited to rapid D4Ag scale up, but existing solutions must be tweaked before their full impact potential will be achieved. Key D4Ag statistics: Total users of solutions headquartered in Ghana388 1.6 million Number of solutions 28 (headquartered); 57 (with a presence) Proportion of users that are women389 30% Nearly even mix across four use cases: Most common primary use case of solutions advisory services (7); market linkage (7); supply chain management (6); data intermediary (5) Government role Active promoter of D4Ag via agricultural and digital policies. Snapshot of D4Ag solutions: Advisory services Market linkage Supply chain management Financial access Macro agriculture intelligence ANNEX 1 179 Context: in 2018;393 (2) an electronic, agricultural Agriculture in Ghana input distribution system with barcodes that allows the government to more quickly Agriculture accounts for 18% of Ghana’s gross detect problems like low-yield seeds and poor domestic product (GDP).390 The percentage of fertiliser. Policymakers have also set themselves agriculture’s contribution to GDP is expected the ambitious target of registering every cocoa to decrease, while non-agricultural services farmer in the country. and other industrial sectors are projected to expand. At present over half of the country’s A number of companies have taken workforce (52%) engages in agriculture.391 advantage of the supportive environment Crop farming is economically more important and built valuable D4Ag services in than livestock production, with cocoa, oil Ghana. palm, coffee, and rubber ranking as the most significant crops.392 In recent years, growth in Outlook non-agricultural services and other industrial The priority now is to ensure D4Ag sectors has outpaced that in agriculture. solutions reach underserved populations Agriculture, nonetheless, continues to grow to deliver real impact. Many farmers are at a strong pace (e.g., 8.4% in 2017), thanks, illiterate, so providers are starting to roll-out in part, to government support via a number services that work around this. Similarly, of interventions, including, as an example, Farmerline and Esoko now provide IVR the 2017–2019 ‘Planting for Food and Jobs’ services that cater to the country’s linguistic Campaign (PFJ). diversity by offering services in local languages. The state of D4Ag in Still, some regions remain too unproductive for D4Ag solutions providers to enter, either Ghana today because soil quality is too poor, transportation Ghana’s government created an infrastructure is weak, or insecurity is high. environment that helps D4Ag thrive. Between 2013 and 2015, Ghana introduced D4Ag should be used to address a series of regulatory reforms intended to barriers to access to credit that farmers, help expand the use of mobile money in the particularly low-income farmers, face. country. These reforms led to a rapid rise in Such farmers enjoy using D4Ag to access the adoption of related services and helped financial services, but few use services beyond open people to the use of digital products mobile payments. Credit remains too expensive and services. Since then, large agribusinesses for most farmers – 28% interest rates were like Yara have encouraged farmers to adopt quoted as recurrent by one expert we spoke mobile money by requiring farmers who work to. Although some D4Ag providers have with them to open mobile money accounts. encouraged banks and investment companies In addition, the recent insecurity of some of to help reduce this cost of debt, little progress Ghana’s trade neighbours (e.g., Burkina Faso, has been made. D4Ag may be able to promote Niger) has pushed more farmers to use mobile farmers’ access to credit indirectly. For money – a safer alternative to in-person cash example, by improving land rights data, D4Ag payments. Moreover, the government has can increase farmers’ ability to use their land introduced a range of initiatives intended to for collateral when borrowing. At present an support the use of innovative technologies initiative funded by the Omidyar Network is specific to agriculture. These include the supporting capacity building at government launch of: (1) ‘Planting for Food and Jobs’, levels in the use of drone technology for land an e-registration platform for farmers with tenure adjudication. This project also extends 577,000 farmers registered and with 202,000 to the Philippines and Colombia. It aims to farmers participating in 2017 and 677,000 build evidence in each country for the effective 180 ANNEX 1 utilisation of drones for property mapping massive impact gains for farmers, but if only and seeks to demonstrate how drones can be landowners, who tend to be men, and not deployed for cadastral surveying on a global other household members are recorded, it scale. In addition, Meridia, the leading D4Ag may also reinforce gender inequality in land innovator in digitally-enabled land registration, ownership. Similarly, women who work in has digitally surveyed and mapped thousands agriculture in Ghana tend to participate more of smallholder farms in Ghana starting in as retailers in local markets. D4Ag solutions 2017, and has helped issue over 5,000 legal could render many of these jobs obsolete. land documents – documents that are crucial Firms like Esoko Ghana are demonstrating to helping smallholders leverage the economic how to counteract the potentially negative potential of their land, a model with potential gender effects of such innovations by, for across Africa. example, actively hiring women to be call centre operators. Forthcoming D4Ag solutions have the potential to help or harm women’s Market linkage solutions are likely to empowerment in the country. Solution be most useful for farmers positioned providers must design products that to service multiple markets. Because are gender positive. For example, the they lack transport options to reach a wider Farmerline digitisation of land rights records will lead to range of potential buyers, most smallholder farmers deal with one local produce buyer only. The impact potential of market linkage D4Ag solutions is therefore limited to larger players and those dealing with multiple markets (e.g., aggregators), who benefit from having a better understanding of when and from where products are coming. Lessons u Mobile money is a key enabler for D4Ag service providers, because it helps farmers and the broader population trust and understand digital products/services. u D4Ag has the potential to increase farmers’ access to credit, for example, by improving their ability to use their officially adjudicated land as collateral. u Market linkage products are most useful to farmers with the means of transport to work with a range of markets. ANNEX 1 181 NIGERIA Nigeria provides an example of how the private sector can drive an innovative digital transformation of agriculture, but it also illustrates how this development can leave more rural and vulnerable farmers behind. Key D4Ag statistics: 0.5 million (another 7 million in Cellulant database via Total users of solutions headquartered in Nigeria394 former Cellulant/SES e-wallet subsidy programme). Number of solutions: 46 (headquartered; 83 (with a presence) Proportion of users that are women395 20% Most common primary use case of solutions Market linkage Supportive, but private sector plays a heavy role in steering Government role the direction of D4Ag. Snapshot of D4Ag solutions: Advisory services Market linkage Supply chain management Financial access Macro agriculture intelligence 182 ANNEX 1 Context: are smallholder farmers.397 Nigeria’s primary Agriculture in Nigeria crops are rice and cassava, but the country is also well suited to become a leading exporter Agriculture accounts for 20% of Nigeria’s of more valuable commodities like cocoa, GDP, compared to an average 16% of GDPs groundnut, and palm oil. Yet, according across Sub-Saharan Africa more generally. to a recent report from Nigeria’s National The sector employs approximately 26 Agricultural Extension and Research Liaison million people, representing about half of Service (NAERLS), the overall farm yield in all jobs.396 More than 80% of these people Arne Hoel, World Bank Nigeria is well below the African average. ANNEX 1 183 Nigeria is still a net importer of some of the Outlook crops the country is best suited to produce, The scalability of these existing solutions such as rice and tomatoes. Even though the remains uncertain. The D4Ag providers government is anxious to slow Nigeria’s import that have emerged in recent years tend to of rice, these imports are expected to increase focus on more specific points of the agriculture by 13% in 2019 making Nigeria the world’s value chain: Hello Tractor aggregates demand second largest rice importer.398 In response for tractor services across Nigeria through a to the increasing imports as well as to the wide network of extension agents; AFEX set insufficient infrastructure in rural areas, the up storage facilities for grains across Nigeria government, in recent years, launched policies and uses Binkabi’s blockchain technology to to liberalise the sector and attract more private improve farmers’ access to credit through the investment. This encouraged many businesses, provision of electronic warehouse receipts; including D4Ag solution providers, to enter the and FarmCrowdy provides capital for specific market.399 on-farm projects. Such focused approaches do not offer the same opportunity to generate The state of D4AG multiple revenue streams for their companies, in Nigeria today but, because the service offerings are simpler, Despite increasing investment in they are easier to adapt to new markets and Nigeria’s agriculture, most D4Ag could, therefore, be easier to scale. These players remain small. Nigeria has one D4Ag firms among others have set ambitious of the most active D4Ag markets in Africa growth targets for themselves – Hello Tractor as measured by count of solutions but few seeks to have 15 million users within five years. players, even those that are well-known and Yet, scaling at this pace will depend heavily regarded – have reached scale. For example, on the sectors’ ability to attract private sector FarmCrowdy serves around 7,000 farmers and capital, and, according to experts in the sector, the Crest Agro-processing project, supported most companies will be hard pressed to deliver by CardinalStone, accounts for about 5,000 returns and risk profiles that can compete with registered farmers. In our data analysis, only similar investments, especially those outside of one solution provider headquartered in Nigeria Africa (e.g., in Southeast Asia). had more than 50,000 users (Hello Tractor, with 250,000 registered farmers). This may be Growth in D4Ag will concentrate on because the players are mainly focussing on solutions that serve farmers who are larger farms in tighter value chains or because more profitable in the short-term, expansion among rural farms is difficult. Few leaving vulnerable populations behind. companies in Nigeria approach the digital Over the next 3–5 years, the most successful transformation of agriculture through a lens D4Ag firms in Nigeria will likely be the ones of inclusion and more vulnerable farmers that target the larger farms, the more well- face fundamental challenges that make them connected farms, and the farms closer to large unattractive for most private sector players. But offtakers. Small farmers in Nigeria, such as this is also true everywhere. What differentiates those in less fertile northern regions and those Nigeria from other D4Ag ecosystems, like farming perishable goods, are more likely to Rwanda and Ethiopia, is that the government be left behind than their counterparts in other and NGOs/foundations have yet to play a countries such as Ethiopia and Rwanda where major role in filling service gaps to promote governments have stepped in to fill the gaps inclusion. left by private sector players. 184 ANNEX 1 Weak fundamentals prevent many achieve scale – other sectors demonstrated smallholder farmers from benefiting the potential of PPPs to do this, for example, from D4Ag. Agriculture experts in Nigeria, in 2017, the government partnered with the including agribusinesses, investors, and D4Ag Venture Garden Group to launch the Health companies, say there is a need to ‘get the Pay Platform; and (2) fill investment gaps in basics right’ before many of the smallholder D4Ag left by private sector investors – most farmers are ready for a digital revolution led private capital is not patient enough to support by the private sector. Until their essential needs investment in inclusive solutions that can serve are addressed, these farmers will not be able vulnerable and less lucrative customer groups, to deliver the quality and scale of output that but philanthropic and government investors is attractive to private sector companies.400 are better positioned to do this. Barriers include: (i) irrigation levels that are well below the average in Africa with less than 2% of all cropland in Nigeria under irrigation; Lessons (ii) last-mile infrastructure that ranked among the worst in the world in the 2018 Global u Service providers who focus on less Competitiveness Report from the World sophisticated D4Ag solutions with only Economic Forum;401 and (iii) poor digital one revenue stream may be able to scale connectivity – despite high levels of mobile more easily, since it is easier to adapt their penetration, the vast majority of farmers are offerings to new markets. But scale does either disconnected or only have access to an not equal use and it may be challenging unstable 2G connection. to make money without offering greater value add to farmers. The government can address these challenges by investing in agriculture’s u While important to encourage fundamental necessities and partnering innovation by the private sector, with the private sector to drive more governments also need to play a role inclusive investments. Public infrastructure in promoting impact and sufficiently for agriculture remains in dire need of regulating the sector. Otherwise, the government funding. The government recently result is a landscape littered with many collaborated with IFAD on a major roads- solutions that do the same thing, with building project in the north of the country. many providers paying limited attention With just 15% of roads in the country paved, to impact. more investments like this are urgently needed. The government also needs to partner with private sectors operators to: (1) use PPPs to ANNEX 1 185 SENEGAL D4Ag could accelerate Senegal’s agricultural transformation but greater policy support and help from incubators/early stage investors is needed for it to take off. Key D4Ag statistics: Total users of solutions headquartered in Senegal402 400,000+ Number of solutions 15 (headquartered); 43 (with a presence) Proportion of users that are women403 10% Most common primary use case of solutions Advisory services and market linkages Government role Government has yet to put its full weight behind D4Ag. Snapshot of D4Ag solutions: Advisory services Market linkage Supply chain management Financial access Macro agriculture intelligence 186 ANNEX 1 Context: Agriculture date (MyAgro is a rare success story). Ninety in Senegal one percent of farmers own less than 10 hectares. D4Ag solutions are less affordable The government has made agriculture to farmers working on this scale. Moreover, a central priority of its development because of Senegal’s underinvestment in plans for the country, but has yet to cell towers and other infrastructure, rural throw its weight behind D4Ag. The populations lack solid access to 2G/3G agricultural sector is of critical importance coverage, mobile phones, or internet. Farmers’ to the economy – it employs over half the attitudes towards digital products and services workforce (53%). However, it accounts for just pose another barrier for D4Ag. Loose data 16% of the GDP.404 Senegal’s primary crops privacy laws have eroded their trust in these are rice and maize, which are organised in solutions. Furthermore, after years of donors loose value chains, and millet and fish, which providing these solutions at no cost, farmers’ have value chains that are slightly tighter (but willingness to pay is low, even if they do still not ideal).405 Compounding the challenge recognise the value in them. of insufficient value chains, land cultivation is lacking – less than 5% of the country’s arable Through their unwillingness to fully land is irrigated.406 Policymakers are focused support D4Ag, policymakers hinder broadly on agricultural transformation (i.e. its ability to scale-up. Tight regulations mechanisation and commercialisation) as a way discourage private actors from choosing to drive economic growth. However, while the to locate in Senegal rather than in a more government has made large public investments favourable environment. Corruption and lack into agriculture (~10% of GDP per year), little of transparency have held back the digital of this has gone toward D4Ag.407 transformation of several aspects of agriculture in the country. Reforms are badly needed – The state of D4AG for example with regard to the management in Senegal today of land rights and the state’s distribution of D4Ag has yet to take off in Senegal, fertiliser to farmers. Country experts say the because farmers are fragmented and lack of government action to support D4Ag is have low levels of access to and trust in partly due to a limited awareness of the long- digital products. Few successful examples term efficiency gains it could yield for state- of D4Ag solutions have emerged in Senegal to funded projects. Xaume Olleros, RTI. ANNEX 1 187 Outlook potential investors. Specifically, better data Senegal’s D4Ag start-ups need a more privacy laws are needed to reassure D4Ag supportive ecosystem of incubators users that their personal data will be kept safe. and early-stage investors to help them Also, more investment in the country’s ICT get off the ground. Mentorship and seed infrastructure (e.g., mobile towers) is required funding are in short supply in the sector. to lay the much-needed foundations upon There are limited incubators and few angel which the private sector can build. investors or VC firms focused on Senegal. In addition to this lack of support, language barriers discourage many would-be investors Lessons (who are often primarily English-speaking) from entering Senegal and other francophone u D4Ag is harder to scale up in countries markets in the region. The shortage of capital where farmers are highly fragmented, has prevented many high-potential firms and this is exacerbated by limited country (e.g., Mlouma and Monobi) from growing support beyond the start-up phase. Additionally, several solutions have failed because their u Cooperatives are a good stakeholder to designers tried to introduce ‘copy-and-paste’ work with when looking to build trust with models from other markets, rather than farmers. investing sufficiently in customisation for local u Expertise is needed to translate Senegalese contexts. A few, rare success stories successful solutions from one market to the show that incubation or early stage investment next. Incubators and early-stage investors can work well. For example, Orange are often well placed to provide this. incubated Bayseddo, a platform that facilitates Translation is also not a game of pure agricultural production by crowdsourcing replication and can require significant finances in Senegal, which CTA recognised effort into learning and adapting to local as one of the winners of the CTA-sponsored market conditions and strong investments Pitch Agrihack awards in 2017. in user-centric design. Cooperatives could provide a good network through which to grow D4Ag. Cooperatives are well coordinated and have deep relationships with their local communities. They are trusted intermediaries, so farmers are “ more likely to use products, including D4Ag, Cooperatives are well coordinated and have deep provided by them. relationships with their local communities. They are trusted intermediaries, so farmers are more likely to The private sector will only be able use products, including D4Ag, provided by them. to achieve so much alone. Advocacy ” and policy reforms are needed to drive more D4Ag momentum within government. NGOs and other organisations focused on social impact must make a clearer case for D4Ag scale-up to decision-makers in government. With greater political will, reforms can follow the example of Nigeria and other countries in the region with policy environments that are more welcoming to 188 ANNEX 1 KENYA D4Ag has flourished in Kenya. This success will continue if ecosystem players works together to manage risk. Key D4Ag statistics: Total users of D4Ag solutions headquartered in Kenya408 9.0 million Number of solutions: 64 (headquartered); 114 (with a presence) Proportion of users that are women409 28% Market linkage (22); Advisory services (19); Financial Most common primary use case of solutions Inclusion (22) Government role Supportive and forward-looking. Snapshot of D4Ag solutions: Advisory services Market linkage Supply chain management Financial access Macro agriculture intelligence ANNEX 1 189 Context: KCB/Mobigrow (0.4 million users), and Agriculture in Kenya PAD (0.4 million users). Agriculture accounts for 34.6% of Kenya’s Kenya’s digital-friendly environment GDP. There are 16 million smallholder has helped D4Ag flourish. D4Ag benefits farmers in the country. More than three- from Kenya’s high levels of connectivity, quarters of Kenyans make some part of their mobile phone usage, and data transparency. living in agriculture. The sector’s primary Safaricom’s M-Pesa and the rise of mobile crops are: maize, coffee, and tea. Yields in money over the last decade has made the country are about 12% higher than Kenyans more comfortable with digital Sub-Saharan African averages but agricultural products, particularly for transactions. An productivity has stagnated in recent years, adaptable regulatory environment enhances maximum yields have not been achieved, the relatively quick uptake of mobile money.411 and only 20% of land is suitable for farming. Nairobi’s emerging community of ICT Moreover, drought and disease continue entrepreneurs has also strengthened growth. to pose a risk to food security for many Additionally, Kenyans have relatively high vulnerable populations in the country.410 levels of basic literacy, especially among youth. This allows enterprises to use SMS The state of D4Ag rather than more-expensive IVR when in Kenya today communicating with users. Kenya has more D4Ag enterprises and users than any other Sub-Sahara African The presence of mobile money has country. Over 100 solutions are in the market increased interest in D4Ag among – 31% of operators on the continent have businesses. Our research found that half locations in Kenya. And 20–30% of Kenyan of venture capital/private equity investment farmers are touched by more than one digital in AgTech in Sub-Saharan Africa occurs in solution. The projected revenues of D4Ag Kenya. The ability to move money digitally is players in Kenya is €18-35 million in 2019. important for most revenue-seeking enterprises Large and fast-growing examples include and private investors. Donors/NGOs tend to WeFarm (1.4 million users), iCow fill the gaps by supporting those solutions that (0.8 million users), Pula (0.6 million users), do not focus on mobile money. Neil Palmer, CIAT 190 ANNEX 1 Outlook enable this. For example, the agricultural The outlook for D4Ag in Kenya looks supply chain, iProcure, is partnering with good, with bundled services best existing agricultural dealers in Kenya. positioned to grow. Commentators are Meanwhile, the growth and expansion of optimistic about the growth potential of D4Ag such platforms as iKilimo and iCow has been in Kenya. Private investment and donor hampered by the lack of strong partnerships support are expected to continue. Broader among stakeholders and by weak evaluation trends are also positive. For example, more and monitoring.413 Intermediaries can play an young people – who drive Kenya’s increase important role in encouraging partnerships. in digital literacy – are expected to stay in AgriFin has become an early leader in this rural areas. Amid such trends, more providers effort, hosting networking opportunities for entities active in agriculture finance.414 will follow enterprises like DigiFarm, which provides farmers with bundled services.412 Additionally, as in some other countries we Providers that offer more than one solution will profiled, policies around data privacy and likely capture more revenue in a competitive customer protection have yet to be developed market where farmers have limited expendable fully.415 Given the size of its D4Ag space, this income. deficiency could present a bigger problem for Kenya than other countries and should be a While the overall forecast is positive, focus area in coming years. experts in the field have advised caution. Some experts on Kenya’s agriculture are concerned about the speed at which extension Lessons services have decreased in recent years. This view is informed by, for example, the u Mobile money and a digitally savvy fact that farmers respond much better to population enable rapid scale-up of D4Ag extension workers using digital tools, rather solutions. than digital-only services. To mitigate risk, the roll-out of new D4Ag technologies should be u Bundled services are better positioned to accompanied by strong human intermediation capture revenue opportunities in consumer along with close monitoring and evaluation. markets primarily consisting of farmers with low expendable income. Collaboration between D4Ag u Farmers are wary of fully digitalised stakeholders can build a thriving D4Ag services. Kenya highlights the sector that works for all users. To continued value of human intermediation help coordinate ecosystem actors and avoid (agent networks) in D4Ag. duplication of effort, solutions must combine familiar faces, technology, and business knowledge. Partnerships between enterprises, agribusinesses, NGOs, banks, and others can ANNEX 1 191 RWANDA416 Rwanda’s government has led remarkable growth in D4Ag. It is now shifting toward a more market-driven approach to scaling up solutions. Key D4Ag statistics: Total users of solutions headquartered in Rwanda417 3.5 million Number of solutions 8 (headquartered); 44 (with a presence) Most common primary use case of solutions Advisory services Active promoter and now moving from market-player toward Government role market-enabler. Snapshot of D4Ag solutions: Advisory services SMART NKUNGANIRE SYSTEM Market linkage Supply chain management Financial access Macro agriculture intelligence 192 ANNEX 1 Context: solutions in agriculture. Physical infrastructure Agriculture in Rwanda has also contributed to this enabling environment. For example, the government has Agriculture accounts for a little more than prioritised the installation of fiberoptic network 30% of Rwanda’s GDP.418 Out of Rwanda’s connections in all districts.422 population of more than 12 million people about 70% are dependent on subsistence farming.419 Due to the high population density CTA’s ICT4Ag international of the country, the average size of farms in conference in Kigali Rwanda is small – between 0.30 and 0.70 CTA hosted an international conference hectares.420 Tea and coffee are the country’s in Kigali, Rwanda, in November 2013, major export products, while plantains, that focussed on the use of ICT in cassava, potatoes, sweet potatoes, maize, and agriculture. Over 400 people attended, beans are among the crops with the highest ‘to explore the possibilities that ICT yield. Government agricultural policy has can provide in agriculture and to focused on a number of priorities in recent develop new solutions that can improve years: low productivity in the agriculture the day-to-day operations of Africa’s sector, the risk posed to Rwanda’s subsistence millions of farmers’.423 The conference farmers by their high-reliance on rain-fed included a number of sessions on produce, and the high fragmentation of crops ICT4Ag-related topics, a hackathon, across the county.421 and a “plug and play day” – during which numerous digitally-enabled The state of D4Ag in solutions for agriculture were presented Rwanda today to attendees.424 This conference set the Rwanda has supported remarkable stage for the ICT4Ag sector in African, growth in D4Ag by investing in large- Caribbean and Pacific countries to grow scale digital hardware and systems. The and attract international attention.425 government has digitised its national identity The subsequent advancements have card system, land titles, platforms to access now equipped Rwanda to move from government services (Irembo), and social registry ICT4Ag to D4Ag and to transition from (Ubedehe). Rwandans’ participation in these government reliance to sustainability. programmes has increased familiarity with digital technologies, priming them to use digital P. Kimeli, CCAFS ANNEX 1 193 To attract D4Ag investment, the government has begun to consolidate Rwanda’s fragmented agriculture sector, but this may only help larger farmers. The government has consolidated farms based on agro-climatic positioning, which has significantly increased the average farm size (previously it was just 0.2 hectares). It also organised farmers into cooperatives and sub-national markets. For example, 350,000 farmers were divided into 300 districts, each of which has a designated coffee aggregator who purchases coffee. D4Ag enterprises tend to reach farmers via such aggregators so these government-led steps make Rwanda a more attractive country for D4Ag activity and allow D4Ag firms to serve larger groups of aligned farmers who have shared paths to market. We have yet to see clear evidence of the impact of this consolidation on farmer productivity, but some experts assert that it tends to help only In response to this ecosystem-building, a Simona Siad, IFAD farms that are above average in size.426 few D4Ag firms have located operations in Rwanda but private investment Donors and NGOs have also supported remains low. N-Frnds records farmer efforts to scale-up D4Ag in Rwanda. transactions to incentivise soft loans from FAO chose to pilot their new initiative, banks, charging the bank for each loan Agricultural Services and Digital Inclusion in obtained by leveraging its data. Kumwe Africa, in Rwanda and has developed four developed internal digital tools to track market smallholder farmer-focused digital products transactions and optimise transportation from and services to launch in 2019. One Acre farm to market. Both Kumwe and N-Frnds Fund created and is beginning to trial a digital are generating healthy revenues and running enrolment system that runs on USSD. This sustainable businesses, but they need capital application is intended to increase adoption and broader markets to scale. Private investors, by allowing farmers to self-enroll with namely, venture capital (VC) and private limited assistance from a field officer. This equity (PE) firms, have not yet demonstrated could dramatically increase the field officer’s much interest in this space. On the other management capacity from an average of hand, Charis Unmanned Aerial Solutions 300 farmers to as many as 2,500 farmers. (UAS) Ltd., a youth-led startup incorporated One Acre Fund also collaborated with the in 2014 and now employing 15 youth, offers Rwandan government in farmer mobilisation drone-based services to various industries, and registration in the Smart Nkunganire including agriculture, and is growing fast. It System, ‘a supply chain management system now provides services to private sector and built by BK TecHouse Ltd in collaboration government agencies in Rwanda, opened with Rwanda Agriculture and Animal a satellite office in Côte d’Ivoire, and also Resources Development Board to digitalise executes contracts in neighbouring countries. the end-to-end value chain of the agro-input The company attracted foreign investment subsidy programme’.427 which allowed further expansion. 194 ANNEX 1 Rwanda has pledged to address the populations are less digitally savvy; and their need for greater investment from the agricultural sectors are fragmented. The result private sector. It introduced tax exemptions is that many companies are hesitant to expand on ICT and agriculture imports, access to into these neighbouring geographies (Uganda land that favours agribusinesses, and access and Zambia perhaps more so than Tanzania). to extensive data about farmers. In late 2018, the Rwandan Ministry of Agriculture and To help its D4Ag firms, Rwanda should Rwanda Development Board announced the look to coordinate its D4Ag policy creation of a ‘one stop centre’ for investors with other countries in the region. committed to increasing annual investment Regional integration has served Rwanda’s in agriculture to €80 million.428 Toward the economic growth well in the past decade. same end, the government also strengthened its Rwanda now has an opportunity to promote focus on the expansion of innovation and skill digital technology as part of this regional building in Rwanda. Knowledge Lab (kLab) integration, and given their sharp dependence is an ‘open technology hub’ that supports on agriculture, D4Ag should be a central entrepreneurs with mentorship, networks, and component. more.429 In 2014, CTA collaborated with kLab and others on the Rwanda National ICT4Ag Hackathon.430 Carnegie Melon, Andela Lessons University, and African Leadership University have talent centres in Rwanda that build u The Rwanda example highlights that needed local skills. Additionally, the active government investment in the €90 million Rwanda Innovation Fund plans to broader enabling environment has strong “support between 20 and 25 ICT companies, impacts on innovator interest in building of which at least 10 will grow into $50 million D4Ag businesses in country. Strong, public worth of corporation in 10 years.” 431 The declarations of commitment to building government will contribute 30% of the capital out ICT infrastructure and PPP models needed for this fund.432 can stimulate investor demand, as well. To become viable, Rwanda’s D4Ag firms u D4Ag players operating in small may need to expand into new countries countries will likely need to expand across with less receptive markets. Rwanda’s borders to reach financially sustainable small size makes it difficult for firms operating scale. That likely requires more regional there to hit the scale needed to become cooperation. profitable. The natural response is to expand into nearby countries. Uganda, Zambia, and u The consolidation of farms helps attract Tanzania are likely targets, but these markets D4Ag investment but may increase the are likely to present new barriers to overcome productivity of large farms only, rather – they are mostly cash-based economies; than smaller farms and more marginalised their governments are less pro-D4Ag; their groups. ANNEX 1 195 SAHEL The analysis of the G5 countries (Niger, Burkina Faso, Mali, Chad and Mauritania) was not done with the same level of detail as the Senegal case study. Nevertheless, the intention is to give a flavour of the specific challenges in these countries based on desk study, interviews, and responses to a survey. The Sahel countries face unique challenges to D4Ag scale-up, making them different from neighbouring countries.433 Solutions, however, could make a large impact in the region, and a few early movers have provided precedents to potential entrants, even under difficult conditions. Key D4Ag statistics: Total users of solutions headquartered in Sahel434 5.7 million Number of solutions 28 (headquartered); 92 (with a presence) Most common primary use case of solutions Advisory services. Snapshot of D4Ag solutions: Niger Advisory Market services linkage Mali Advisory services Financial access Chad Advisory services Burkina Faso Advisory services Market linkage Financial access 196 ANNEX 1 Context: Agriculture in was much higher than to commitments made the Sahel in Chad and Mauritania (€8.5 million and €7.3 million, respectively). It is not possible to As mentioned above, agriculture in estimate what proportion of these commitments the Sahel region faces a number of is designated for digital, but it is believed to significant challenges that make D4Ag be very low. The G5 Sahel group is launching scale-up and agricultural transformation multiple agricultural and infrastructural potentially more difficult. Various factors efforts through a rolling, three-year Priority make farming in the region less profitable Investment Program (PIP); many of these and, in turn, reduce the viability of D4Ag projects have allocations for agriculture and for solutions. Loose commodity markets do not telecommunication, but the majority of funding lend themselves to the implementation of still needs to be secured.436 Still, broadly standardised digital solutions but may benefit speaking, there have been some important most from the price transparency they could improvements in IT and communications.437 create. One of the main questions the region The implementation of appropriate D4Ag faces is how Sahelian agriculture can innovate could catalyse agricultural development in and develop to meet the vital needs of a Sahelian countries. growing population in the face of climatic hazards. The State of D4Ag Governments in the Sahel have made in the Sahel today agriculture a central priority of their Because of the level of market development, but D4Ag is not yet dysfunction in the Sahel, the potential a priority for all. Recent funding for impact of D4Ag solutions could be vast. agriculture in the Sahel by the Organisation Isolated farmers would benefit most from for Economic Co-operation and Development digitally-enabled information sharing and (OECD) countries exhibits significant advisory services but struggle to find affordable variance.194 Similarly, bilateral commitments and available connectivity. It will be difficult to during the last five years by the Development make D4Ag work in Sahel’s loose commodity Assistance Committee (DAC) countries to markets, but this is the kind of environment the agriculture sector of Senegal, Mali, Niger where the price transparency offered by D4Ag and Burkina Faso (€85 million, €95 million, solutions could offer the largest benefits. €55 million, and €58 million, respectively) Other solutions like digitally-enabled climate Figure 37 Characteristics of Sahelian countries, by country Country Population % rural Mobile obile $ total/ (WB, 2017) (WB, 2016) penetration M (GSMA) rural438 Niger 22M 80% 29% (2018) 8%/6% Burkina Faso 20M 69% 44% (2018) 29%/27% Mali 19M 59% 61% (2016) 24%/20% Chad 15M 77% 30% (2016) 13%/13% Mauritania 4M 40% 65% (2016) 3%/1% ANNEX 1 197 insurance, soil mapping, water availability, and Rambaldi Giacomo, CTA grazing guides also hold particular promise for the region. Some positive experiences illuminate the way forward (see below). The level of D4Ag development varies considerably across the region. Burkina Faso and Mali are significantly ahead of other countries, with 36 and 35 solutions present, respectively. Niger has less than half this amount, 14, whereas Chad has six and Mauritania has only one.439 These figures largely mirror how connected each country’s rural populations are. For example, almost 40% of Burkina Faso’s rural population has access to a mobile phone or the internet, but less than 15% of rural populations in Niger, Chad and Mauritania have such access. Nevertheless, interviews indicate that connectivity is not perceived as a huge issue, even for those working in remote areas. However, the state of IT infrastructure at Burkina Faso has laid the D4Ag groundwork, government ministries – dated systems that not only through investments in connectivity lack internet connection and have weak but also through the development of security features – presents a significant issue. middleware. D4Ag started emerging in the country about 15 years ago.440 More recently, Several promising D4Ag solutions e-Burkina, a World Bank-supported platform in emerged in recent years that offer Burkina Faso, helps digitalise land registrations lessons to those entering the market. The and farm profile systems. This service provides following include some of the multiple actors farmers with more information about how that are already present in the Sahel countries much land they have, how they should use it, and deploying such solutions on a broad scale. and how they can protect themselves against SNV launched two Geodata for Agriculture drought. Burkina Faso is also leading the and Water (G4AW) projects: Sustainable way in the field of open data for agriculture, Technology Adapted for Mali’s Pastoralists working on a coalition in the Sahel gathering (STAMP) located in Mali and Mobile Data various actors including the Ministry for the for Moving Herd Management (MODHEM) Country Population % rural Mobile (WB, 2017 (WB, 2016) penetration Mobile $ total/ Development of Digital Economy, the ) (GSMA) rural438 based in Burkina Faso. Espace Geomatique Ministry of Agriculture, Global Open Data société anonyme à responsabilité limitée For Agriculture and Nutrition (GODAN), Niger 22M 80% 29% (2018) 8%/6% (SARL), Georisk Afric SARL, and Cargitech Akvo, and the Permanent Interstate Committee SARL have all introduced drone-based D4Ag for Drought Control in the Sahel (CILSS). Burkina Faso 20M 69% 44% (2018) 29%/27% efforts. Afrique Verte, Manobi, and Esoko Drone technology is well represented in provide market linkage and supply chain the Sahel countries with several companies Mali 19M 59% 61% (2016) 24%/20% services. Akvo, Viamo and others act as data offering services. intermediaries and provide data intelligence. Chad 15M 77% 30% (2016) 13%/13% Below is an overview of D4Ag use cases However, the benefits of D4Ag have yet to be Mauritania 4M 40% 65% (2016) 3%/1% identified through interviews and survey are fully realised because solution providers are presented for Burkina Faso and Mali as well as still struggling to feed highly localised data some examples for the different countries. into their IVR services. Although the solutions 198 ANNEX 1 Rambaldi Giacomo, CTA require further refinement, the tactics used to consortium of Dutch NGO SNV, Orange scale can provide inspiration for others. For Mali, Malian NGO TASSAGHT, and Satellite example, open platforms allow large farmer data processor Hoefsloot Spatial Solution. federations to contribute directly, rather than In a second phase, Garbal will roll out work through government authorities. This financial access services (leveraging Orange’s increases efficiency and participation. mobile money platform) and input access (in partnership with regional input providers). In Mali, the initial successes of STAMP’s Working with these aggregators and value Garbal services441 exemplify the potential chain actors, the Garbal team believes it can for carefully cultivated partnerships, build upon existing infrastructure, while also programme flexibility, and commercially- developing a long-term sustainable business focused programme design to address the model. As STAMP’s Garbal demonstrates, needs of climate-vulnerable and conflict- models that achieve significant impacts while affected pastoralists in the Sahel. This project linking pastoralists and farmers to value chain provides Malian pastoralists with satellite- actors to ensure commercial viability will driven insights about the location of grazing underpin D4Ag’s contributions in the Sahel. grounds and water, crowdsourced information about grazing quality and availability, and locally relevant market price information. It is funded by the Dutch government through the G4AW (Geodata for Agriculture and Water) programme and implemented by a ANNEX 1 199 In Niger, the “Tele-Irrigation” (from TECHINNOV) is a technological process Lessons that allows a farmer to remotely control the irrigation system of his farm and follow u To bring greater benefit to the an intelligent distribution of water (needs, agricultural value chain actors it is quantity, time, type of speculation), regardless paramount to better understand their of its geographical position and time, by means needs and the needs of smallholder of his mobile phone and solar. Tele-Irrigation farmers and to develop relevant/adapted/ can also collect and disseminate real-time and gender-sensitive services. Examples in the remote meteorological and hydrological data Sahel highlight that it is possible to serve including temperature, soil moisture content, even highly marginalised segments with rainfall, solar radiation and wind speed. success. This process allows the farmer (i) time and u It is not enough to focus on registration. energy savings; (ii) increased irrigable area; Impact is only achieved when a service is (iii) increased production and income; and (iv) utilised: important work should be done to controlled water management. increase service use. In Chad and Mauritania, few companies are u Data quality and accessibility must offering market linkage and advisory services in be improved to aid actors in making the agriculture sector and the number of D4Ag informed, evidence-based decisions. use cases identified through the survey and This need is particularly prescient given interviews is low. the context of climate change, in which experience no longer serves as a reliable Outlook barometer. The success stories in the Sahel highlight that it is possible for D4Ag solutions to make an u Actors recognise that data impact even in challenging conditions. Still, intermediaries/aggregators442 and data for D4Ag to truly take off, there needs to be storage systems improve agricultural much more political will for D4Ag across the value chains. Different datasets should be region. It is strong political will that will set the brought together to increase value. Data stage for countries to make the requisite policy sharing is paramount. changes and enabling investments for D4Ag to u Developing human capital at every take off. level of the D4Ag ecosystem is crucial: All actors in the agricultural value chains (from smallholder farmers to extension officers and policy makers) must build digital skills and literacy in order for D4Ag to expand. u Various kinds of business models are explored by agri-preneurs and, to create jobs for youth and women, their efforts require specific support, such as incentives for small-business and market development assistance. 200 ANNEX 2 ANNEX 2 STAKEHOLDER CONSULTATIONS Name Entity Name Entity Ademola Akinyemi FarmCenta Catherine de Come STAMP/MODHEM and SNV Aimable Ntukanyagwe IFAD Christian Merz GIZ Alex Calvin Gbetie Profish Clara Colina MasterCard Foundation Alex Sanderson Kumwe Codou Ndiaye Dimagi Aliyu Suleiman Dangote Daniel Asare-Kyei Esoko Amadou Ba World Bank David Muwonge NUCAFE Amare Mugoro CommonSense project Diouf Mamadou Coumba PRODAC Amsata Niang ANIDA Eli Pollak Apollo Agriculture Ananth Raj Farm to Market Alliance Elias Gossaye Apposit André Laperrière Godan Elias Nure Agricultural Transformation Agency Andrew Gartside DFID Elisa Minischetti Yara International Andrew Nevin Binkabi Ethan Laub MOSS ICT/M-BIRR Anelyia Muller World Bank Farah Dib World Bank Angelique Uwimana FAO Filippo Brasesco FAO Angus Keck AgUnity Florien Habinshuti PSDAG (USAID) Anne Bastin Ndiaye IFC Getamesay Demeke Interaide Arnaud de Vanssay EU Delegation Girma Meki Batu Union Asaye Asnake Farm Africa ET Hamza Rkha Chaham SowIt Awa Caba Sooretul Harriet Blest VIAMO Ayodeji Balogun AFEX Heiner Bauman PAD Ayokanmi Ayuba Technoserve Hillary Miller-Wise Tulaa Bekure Tamirat Gebeya Hussain Suleman SigFox Belinda Bwiza OneAcreFund Ifeanyi Anazodo FarmCrowdy Ben White VC4Africa Ikenna Nzewi Releaf Benji Meltzer Aerobotics Ilisa Gertner Chemonics Bernhard Kowatsch WFP Innocent Mudenge NYAB Bolaji Akinboro Cellulant Jasper Spikker Agriterra Bolaji Akinboro Cellulant Jean Louis Uwitonze PSDAG (USAID) Brook Ashinne Viamo Jeehye Kim World Bank ANNEX 2 201 Name Entity Name Entity Jeroen van der Sommen Akvo Onyeka Akumah FarmCrowdy (Nigeria) Jonas Chianu AfDB Oswald Jumira Liquid Telecom Joshua Ayinbora Groital Farms Papa Samba Diop APIX Joshua Thompson AAIN Paul Wechuli WaziHub Jovani Ntabgoba N-frnds Peter Githinji AAIN Judy Payne USAID Ranveer Chandra Microsoft Justine Mucyo Holland Greentech Reha Yudarkal IBM Karin Lion Digital Green Rob Fuller AgDevCco Karl Wurster USAID Robert Berlin Syngenta Foundation Katie Hauser USAID Ruud Grim G4AW Kebede Ayele Digital Green Sandi Roberts AgDevCco Khalifababacar SARR GIS Association Saskia Vossenberg FMO Laurent Cochet Interaide Selina Kim IBM Levon Minassian Arable Labs Serge Moungnanou UNCDF Liisa Smits Ignitia Shirley Somuah Cardinalstone Partners Luda Bujoreanu World Bank Shreya Agarwal Digital Green Mamadou Sall Bayseddo Simon Pierre Jules Duchatelet World Bank Marc Schut IITA Stephane Devaux EU Delegation Marco Streng Agriterra Stephen Ibaraki REDDS Capital Marise Blom ScopeInsight Stewart Collis aWhere International Livestock Sylvie Nirere IDH Masresha Taye Research Institute/Index Based Livestock Insurance Tiphaine Crenn IFC Melat Mebtratu MOSS ICT/M-BIRR Tomaso Ceccarelli CommonSense project Mikael Hook MasterCard Foundation Van Jones Hello Tractor Mr MBAYE Birame Seck Institut Sénégalais de Recherches Agricoles (ISRA) Venkat Maroju SourceTrace Mutembei Karakui GIZ Waly Clement Faye UNCDF Natalia Pshenichnaya GSMA mAgri Yaron Cohen Mareco LTD Ndubuisi Ekekwe Zenvus Nicole Ihirwe Agriterra 202 ANNEX 3 ANNEX 3 DETAILED METHODOLOGY In this report, the Dalberg team and CTA sought D4Ag solutions database to analyse the state of D4Ag in Sub-Saharan Africa The CTA-Dalberg D4Ag solutions database (the and to construct a current-state baseline as well as ‘D4Ag database’) currently contains information on projections for key D4Ag sector characteristics and 410+ active D4Ag solutions, of which the data trends (historical and future-facing) on the basis of set used for all analyses in this report focused primary data collection, secondary research, and on 390 active D4Ag solutions. The others forecasting models. (typically very small or very early stage enterprises) were launched in recent months or were discovered This methodology provides an overview of the during late stages of the report editorial process. We overall approach, the key tools used, and critical will include and analyse these additional solutions in assumptions for a few select areas of analysis. the next edition of the report. Where not covered in this methodology, relevant information on assumptions and sources is embedded D4Ag solutions for the purposes of the database in the endnotes section of the report. include both specialised D4Ag enterprises with a single D4Ag solution and individual D4Ag services/ The data collected for this report is the intellectual solutions developed and distributed by a third- property of CTA and Dalberg, but our hope is to party parent organisation such as an NGO, MNO, make additional elements of the underlying data agribusiness, or technology company (both big and available in future publications and via the selective small). All of these solutions are either headquartered release of data sets for researchers. Additionally, the in Sub-Saharan Africa or focus a substantial CTA and Dalberg teams – in collaboration with portion of their activities on the region if they are other sector knowledge leaders and funders and as incorporated or led from other geographies. part of our commitment to open agriculture data – are exploring opportunities to develop an open, We estimate that the D4Ag database likely represents digitalised, publicly available, and regularly updated 90%+ of all existing and functioning D4Ag version of the D4Ag solution database which will solutions in Africa. While we attempted to make serve as a knowledge tool for the entire sector. our database of solution providers as comprehensive as possible, it is not by any means exhaustive, owing Advisory Council to the time constraints facing the report’s production An Advisory Council was convened enlisting team and the rapidly evolving nature of the D4Ag experts from public and private sector actors, sector where new D4Ag solutions get launched thought leaders, foundation representatives and almost weekly in Africa. leading implementers. The individuals are detailed in the Acknowledgments section of the report. Beyond missing some of the newest start-ups, The Council was invaluable in informing the for several use case categories in this report, the development of the report, in particular around boundaries between D4Ag solutions and out of scope refining the strategic framework used to investigate enterprises were not always clear. For instance, in D4Ag’s role in Africa’s agricultural transformation; the financial access use case, traditional banks and reviewing, and providing feedback on the report’s MFIs are increasingly digitising their operations and various drafts and insights; providing input on data incorporating digital features into their products sources and advising on how to tailor key report and services even if such products are not explicitly messages to its multi-sectoral audience. branded as being ‘digital’. In the macro agri- intelligence use case, a growing number of donor- funded initiatives and private sector solutions are Data collection exploring various uses of data for agriculture sector To gather the required information, we relied on intelligence but have not yet fully productised such sector interviews, a large-scale survey of tools, or are exploring them within the context of a solution providers, and desk research. These broader technology category (e.g., satellite imaging data collection activities fed into the development of intelligence) and not limiting their activities to a large database of D4Ag solutions, which was a agriculture. It is almost certain that a number of core analytic tool for the effort and is meant to serve such financial access and agri-intelligence solutions as a refreshable baseline data set for the sector for are not in the database. years to come. ANNEX 3 203 In addition to ‘live’ solutions, the database tracks Expert interviews >70 defunct solutions which have ceased Between October 2018 and February 2019, the operations due to business model failure, the end of Dalberg team conducted ~120 semi-structured donor funding, or business model changes that have interviews of leading experts and D4Ag solution taken them out of the D4Ag sector (e.g., moving leaders in the fields of agriculture technology and from D4Ag financing to an urban fintech focus). The digital services, agriculture and food markets, data set of defunct solutions is far less comprehensive donor initiatives, and government programmes. than that of the active players. Based on data from In many cases, CTA provided connections to other early D4Ag solution databases in the sector, interviewees, while in other cases the Dalberg most notably GSMA’s mAgri tracker (active until team sourced contacts through its global network ~2014), we estimate that there are at the very least of consulting professionals or through external 50 and possibly as many as 100 other defunct D4Ag connections. Interviews generally ranged from 30 to solutions that are currently non-operational but were 60 minutes. Where possible, the team corroborated in business at some point over the past 15 years. the interviewees’ statements with secondary data acquired through desk research. To generate the list of >480 total solutions in the database (460 analysed specifically in this The interview insights then fed into a variety report), the Dalberg and CTA teams drew on a of the analyses for this report including the wide range of sources including old data (2013- D4Ag database, country case studies, use case 2014) from the no longer functioning GSMA mAgri segmentation, business model analyses, and general tracker, CTA’s ICT4Ag solution database, Dalberg’s perspectives on sector trends. ICTAg database (developed in support of the Bill and Melinda Gates Foundation’s ICT4Ag strategy D4Ag solution survey in 2016), Dalberg digital agriculture landscaping Drawing on an early version of the D4Ag solution studies for select African countries (developed database, CTA and Dalberg collaborated to design jointly with MercyCorp’s AgriFin Accelerate the D4Ag solution survey during the autumn of team), the Global Open Data for Agriculture 2018. The survey launched in mid-November 2018 Network (GODAN) membership list of >920 and remained open for data collection until the first member organisations, MasterCard Foundation week of February 2019. Rural Finance Learning Lab’s data sets on digital agricultural finance providers and intermediaries, Dalberg distributed the survey to all solutions and and Africa AgTech startup landscape maps from enterprises it had identified up to that point (430) organisations like Disrupt Africa and Briter Bridges. via extensive desk research prior to the survey’s Less systematically, we supplemented this data with design, which included all CTA-supported or additional D4Ag solutions surfaced through desk affiliated solutions. research on specific use cases, expert interviews, and country case study field trips. Dalberg sent several follow-up emails, collecting 175 responses by February 2019. Of these 175 responses, The database tracks ~20 data fields for each solution 35 were highly incomplete or otherwise flawed. Once that cover factors such as geographic location and these were removed, Dalberg proceeded to analyse focus, year of launch, organisational type/sub-type, the final ‘clean’ dataset of 140 survey responses use case type (all use cases covered, primary use (~32% response rate). Dalberg then supplemented case, primary use case sub-type), reach (registrations, analysis of these data points with extensive secondary engaged/active users), revenues, profitability, data collection. inclusion (e.g., gender and youth disaggregated data), impact (i.e., yield and income), and contact Desk research information. For factors like reach, revenue, and impact only a subset of all solutions have data. We supplemented our primary research with analyses of publicly available knowledge Alongside this central database we collected resources published by international development a few other datasets used for the analyses organisations such as CTA, USAID, GSMA, including (i) a tracker of D4Ag transactions (based World Bank, FAO, CGAP, AGRA, GIZ, and the on press releases, PE/VC specialist reports on Africa, MasterCard Foundation. In addition, we conducted and our expert interviews); (ii) a small database searches of academic literature through academic on D4Ag donor funding based on desk research research databases, consulted the official reports of and funder interviews; (iii) a D4Ag impact tracker solution providers where available, and reviewed capturing yield, income, and other impacts of D4Ag relevant news coverage. In sum our team reviewed solutions (based on the USAID ICT4Ag impact hundreds of sources, ~250 of which are captured in tracker and extended with data points found through the report’s Bibliography. desk research and interviews). 204 ANNEX 3 Country case study field work yield and income) for solutions in our database Between November 2018 and March 2019, the and the broader academic literature on D4Ag Dalberg team conducted five in-person country case impacts based on peer reviewed publications, study field visits and two ‘light touch’ case studies via publicly available publications, and proprietary phone interviews or brief in-person conversations. M&E materials shared by a few large players Fieldwork in Ethiopia, Rwanda, Nigeria, and • Investments: Analyses of volumes, number of Senegal leveraged the local knowledge of Dalberg’s transactions, and investment instruments for consulting professionals based in-country, while PE/VC transactions focused on African D4Ag fieldwork in Ghana engaged local resources with start-ups and non-African D4Ag start-ups that strong knowledge of the local context. The team have an exclusive or major focus on Africa conducted in-person interviews in these countries and conducted supplementary interviews with local • Donor funding: Analyses of the volume, experts by phone. composition, and trends over time of the development sector (DFI, bilateral, private The case studies of Kenya and the Sahel region foundation) funding for D4Ag relied on remote conversations with experts with The methodology and key assumptions for all D4Ag experience in these regions. Additionally, of these analyses are discussed in the endnotes for Kenya, our team drew on interviews and data throughout the report, tied to the relevant report collected during the World Bank’s Disruptive sections. Below we delve into a few of the more Agricultural Technology Challenge and Conference critical analyses and assumptions. in Nairobi in March 2019. Solution landscaping and segmentation Data analysis Our team categorised all solutions captured in the The report looked at a large number of issues related D4Ag database into five broad categories of use to the D4Ag sector and relied on both qualitative cases (advisory services, market linkages, financial and quantitative data. access, supply chain management, and macro agri- intelligence). In addition, we collected information Among other variables, quantitative data on D4Ag infrastructure players – typically referred to analytics focused on key elements such as: as D4Ag or agriculture data ‘intermediaries’ in the report. These are essentially D4Ag data, software, • Solution landscape: The number, and analytics vendors who work across multiple use segmentation, and dynamics over time of D4Ag cases on a B2B (and occasionally B2C basis) but are solutions not aligned to any individual farmer facing use case. • Reach and use: The reach of D4Ag solutions – including different definitions of reach, The categorisation of solutions relied on self-reported ‘engagement’, and ‘active’ use; as well as the responses for survey participants and then expert- segmentation of the number of registered based judgments by the Dalberg team for other farmers, the most accessible reach variable, along organisations in the database. dimensions such as use concentration, case, geography, and organisation type Reach and inclusion – registered, engaged/active, women and • Penetration analysis: Assessment of D4Ag youth users penetration in Africa along different definitions of the addressable smallholder farmer market Our team collected total reach information in terms of the number of farmers registered or self-reported • Revenues: Sizing of current earned revenues of ‘active’ users on the basis of the survey, interviews, the D4Ag sector, split by organisation type and and desk research. For the largest players in the solution use case, as well as self-reported data on database, every attempt was made to validate the revenue sources numbers by interviewing representatives of the organisation or by talking to their peers and sector • Addressable market: Sizing of the addressable experts. Active women user information was based market (both in terms of the number of client on the solution provider responses to the survey and potential revenue pools) supplemented with interviews and desk research. • Profitability: Estimates of the share of the sector that are break-even/profitable based on Definitions of ‘active’ or ‘engaged’ users lack self-reported data, triangulated across a number standardisation or consistency across use cases and of survey questions for survey respondents they are not transparent or comparable; an ‘active’ financial user might have money in a savings account • Inclusion: Inclusivity of D4Ag solutions with a while an ‘active’ market linkages user might report particular emphasis on the share of users who are prices each day. Surveyed solutions reported both women and youth (<35 years) self-defined ‘active users’ and ‘users active at least • Impact: Self-reported impact data (particularly once a month’; the self-defined figure was less than ANNEX 3 205 the monthly figure, suggesting that solutions define The total number of addressable farmers, in itself, ‘active’ reasonably, but still subject to tremendous is a figure on which there is no clear consensus in methodological and terminological ambiguity and the sector (or the broader agriculture development variation. literature on Africa). As noted in the body of the report, to deal with these For the purposes of this report we estimate a total inconsistencies, we have created a new definition of of 63 million smallholder (<2 hectare) farms in ‘engaged’ users as a catch-all category to differentiate Africa based on the latest estimates from a systemic farmers who use D4Ag solutions, to at least some review of global smallholder farmer estimates.443 The extent, from those who are registered but are in number is derived by multiplying what we believe is reality non-users. the most recent and credible estimate of the number of Sub-Saharan African farms (77 million) by the Estimated revenues share of those farms that are under two hectares We calculated revenues by (1) establishing average in size (82%). Using an average of three adults per annual revenue per user (ARPU) from solution smallholder farm from the literature, we estimate providers that publish both user and revenue that the total number of smallholder farmers in information or shared such information with us via Sub-Saharan Africa is 190 million. The figure below interviews and the survey; (2) we mapped ARPU shows these numbers and the underlying sources. from (1) for solution providers that publish numbers We use the top of the range for our estimate as that of users but not revenues to estimate their total reflects more recent and granular data sets. revenue; (3) for solution providers that publish neither number of users nor revenues, we used See Figure 38: Smallholder farmer estimate an averaged number of users from (1) and (2) and In addition to the number of smallholder farmers, average ARPU from (1). Adding the three analyses we also estimate the number of pastoralists in together produces a minimum, maximum, and Africa, small agriculturalists engage in livestock average estimate of total D4Ag revenues. production who do not have land and therefore cannot be estimated from smallholder farm data. Extrapolation across organisations with unknown There are a range of estimates for the number of revenues was done for commercial enterprises, African pastoralists in the literature (25-80 million), NGOs, and MNOs, as revenues flowing to other complicated by the paucity of data and definitional organisation types are difficult to isolate and challenges (e.g., distinction between pastoralists quantify. and agro-pastoralists). We believe the most reliable data, with granular country level estimates, comes D4Ag market penetration and total from a UNECA study in 2015, which we have addressable market analyses supplemented with research on additional countries This analysis was based on two key inputs: (i) ARPUs (e.g., Tanzania) that have pastoralists but were not across each use-case – retrieved from estimated included in the data set to estimate a total of 60 revenue figures; (ii) the expected total number of million pastoralists in 2018. See Figure 39. farmers in Africa that could theoretically receive a D4Ag product or service. Figure 38 Smallholder farmer estimate Statistic Figure Source Year Underlying source years Farms in Africa 51M Lowder, et al. 2016 1960-2008 77M Lowder, et al. 2016 1970-2014 Share of African smallholder 82% FAO 2001, 2013 1990; farms <2ha 1996-2005 Smallholder farms in Africa <2ha 42–63M Calculated - - Number of adults (14-60) per ~3 Deininger, et al. 2017 2010-2012 African smallholder farm Number of adults (14-60) on 125–189M Calculated - - smallholder farms <2ha 206 ANNEX 3 Using a pastoralist household size of six based on To estimate household penetration of D4Ag country survey data, we estimate a grand total of solutions, we looked at the estimated number of 8-10 million pastoralist households in the region. registered farmers for each use case in comparison to the total number of smallholders and smallholder Combining across farm-based smallholder farmers households in Sub-Saharan Africa. and pastoralists, and using the top of the range based on our interpretation of the numbers, we See Figure 40 below with D4Ag registered user estimate a grand total of 250 million smallholders penetration of the market, overall and by use case. and 72 million smallholder/pastoralist households in the region. Figure 39 African pastoralist estimates444 Statistic Figure Source Year Underlying source years Pastoralists 25M Bonfiglioli 1992 40M Cervigni, et al. 2016 Pastoralists (Sahel and Horn) 58M UNECA 2015 Agro-pastoralists 80M Cervigni, et al. 2016 (incl. some smallholder farmers) 50-200M Bayer & Bayer 2015 Number of pastoralists 50-60M Dalberg estimate using – – existing ranges Number of adult equivalents per 6 ElHadi, et al. 2012 2012 African pastoralist households Number of pastoralist households ~8–10M Dalberg estimate using – – existing ranges Figure 40 D4Ag registered user penetration of the market, overall and by use case EOY 2018 Assumes per SHF (190M SHFs plus 60M pastoralists) Assumes per SHF household (63M SHF HH plus 10M pastoralist HHs) Upper bound Lower bound Total reach 33.1M 13% 45% Advisory services 22.6M 9% 31% Financial access 5.6M 2% 8% Market linkages 2.5M 1% 3% Supply chain 2.4M 1% 3% ANNEX 3 207 The analysis in the preceding figures shows of solutions, as that depends on how fast the market that, taken as a share of all smallholders and consolidates. In this analysis, it is assumed that 20% momentarily assuming no duplications between of users are double-counted in 2019 and 2022 and farmers registered for different categories of D4Ag the same number in 2030 (a simplifying assumption solutions, the total reach (33.1 million) represents which is unlikely since duplication in use will grow as 13% of all smallholders (250 million). With a more farmers register for services). Further, based on duplication assumption of 20% (as explained survey data, ~42% of all unique users are ‘engaged’. in the body of the report), the estimated reach figure of ~26 million farmers represents an overall Survey participants reported a historical (three year) penetration of smallholders of ~10%. We believe annual growth rate of 44% in terms of their number that the actual penetration very likely sits in this of registered farmers, a figure also triangulated 10-13% range today. with a few large D4Ag actors who were not survey respondents. From a forward-looking perspective, Viewed from the perspective of smallholder survey participants projected an average growth households, the penetration figure could be a lot rate of 55% over the next three years in their client higher. If one assumes, for instance, that households base. We also looked at the absolute number of new only subscribe to one solution, penetration could farmers that were registered over the past three years be as high as 45% in 2018, but we view this to derive a more conservative scenario in which assumption as being highly improbable based on farmer acquisition by D4Ag enterprises does not observed behaviour in the field, particularly for accelerate but instead proceeds with the same pace advisory solutions which can easily have many in terms of the absolute number of farmers registered subscribers or subscriptions per each smallholder each year. farmer household. Because data on the average number of solutions per household is unavailable The three scenarios (55% CAGR for aggressive today from smallholder surveys, we anchor the growth, 44% CAGR based on historical growth, discussion in the report on the overall number of and 22% CAGR, derived, for conservative growth), smallholder farmers rather than the number of then yielded our estimates of 60/100/125 million households for the penetration estimate (i.e., we farmers registered by 2020 from a 33 million farmer propose the 10-13% penetration figure). base. We dismissed the top end of this projection as being too aggressive and the report then used The total addressable market analysis (covered in the 60-100 million registered farmer range in 2022 depth in the body of the report and related endnotes to also derive the market size based on unique and in Chapter 3) draws on this same data for overall engaged farmers. population sizing and then multiplies it by estimated ARPU ranges for each solution. As one nuance Estimated investments in that analysis, we assume that the addressable and donor funding market (in terms of revenues) for advisory solutions We calculated investments based on desk research is bounded by the number of all smallholders data (supplemented with interviews) of relevant whereas for use cases like market linkage and yearly PE/VC investments to D4Ag enterprises financial access, the more relevant metric is the operating in Africa. We triangulated the resulting number of households as the solution (e.g., credit, estimates with data reported by organisations like insurance contract, digitally-enabled market off- AgFunder as well as players like Disrupt Africa who take arrangement) is tied to the farm rather than to track start-up investments in the region on an annual the number of individuals on that farm. To derive basis by theme and sector (in this case, AgTech). the final addressable market figures, the resulting potential revenue pools are adjusted based on the For donor funding volumes, building on earlier connectivity constraint for the market (e.g., share of analyses of donor trends in the space developed by households with mobile subscriptions or share that Dalberg, we sourced estimates of donor funding have access to phones). from ~15 known active funders in the sector. In some cases the number was a directional estimate While our survey and interviews only focused on derived from interviews, in others (e.g., EU, smallholder farmers as users of the digital solutions, BMGF) our team had access to underlying project we also believe that there are other users within the databases which were generously shared by some agricultural ecosystem such as traders, extension of the Advisory Council members for the purposes workers, researchers, policy makers but are not of this report. specifically referenced in this report. Future reach and revenues The future growth rate is based on self-reported historic growth rate and expected growth rate from survey respondents. We did not make any projections regarding the growth rate of the number 208 ENDNOTES ENDNOTES Executive summary 1 Food and Agriculture Organization of the United Nations (FAO). 2017. ‘The future of food and agriculture: Trends and challenges’ (www.fao.org/3/a-i6583e.pdf). 2 The baseline number of smallholders in Sub-Saharan Africa used in this report is estimated at 73 million agricultural households, including smallholder farm and pastoralist households, and a total of 250 million smallholder farmers and pastoralists (i.e., adults engaged in these agricultural activities). For details on sources for this estimate, please see the Methodology section in the Annex. 3 For details on how this – and all other figures in the executive summary – have been calculated, please refer to the main body of the report as well as the Annex, where we present a detailed Methodology. Chapter 1 4 FAO. 2017. ‘The future of food and agriculture: Trends and challenges’ (www.fao.org/3/a-i6583e.pdf). 5 van Ittersum, M.K. et al. 2018. ‘Can sub-Saharan Africa feed itself?’ PNAS (www.pnas.org/content/pnas/113/52/14964.full. pdf). 6 FAO. 2017. ‘The future of food and agriculture: Trends and challenges’ (www.fao.org/3/a-i6583e.pdf). 7 These factors bring a number of other risks, such as poor educational outcomes and susceptibility to communicable and non-communicable diseases. See FAO. 2017. ‘Regional Overview of Food Security and Nutrition’ (www.fao.org/3/a-i7967e. pdf). 8 FAO and UN explain, “Nutrition security differs from food security in that it also considers the aspects of adequate caring practices, health and hygiene in addition to dietary adequacy.” See FAO. 2018. ‘The State of Food Security and Nutrition in the World’ (www.fao.org/3/I9553EN/i9553en.pdf). We believe that agricultural transformation has the potential to most directly contribute to dietary adequacy. 9 Ibid. See also FAO. 2018. ‘Africa Regional Overview of Food Security and Nutrition: Addressing the Threat From Climate Variability’ (https://reliefweb.int/sites/reliefweb.int/files/resources/ca2710en.pdf); Wiebe, K. et al. 2016. ‘The Effects of Climate Change on Agriculture and Food Security in Africa’. International Food Policy Research Institute (IFPRI), book chapter (www.ifpri.org/publication/effects-climate-change-agriculture-and-food-security-africa). 10 Goedde, L. et al. 2019. ‘Winning in Africa’s Agricultural market’. McKinsey and Company (www.mckinsey.com/industries/ agriculture/our-insights/winning-in-africas-agricultural-market). 11 McArthur, J. and McCord, G.C. 2014. ‘Fertilising Growth: Agricultural Inputs and Their Effects in Economic Development’. The Brookings Institution (www.brookings.edu/wp-content/uploads/2014/09/fertilizing-growth-final-v3.pdf). 12 High-end estimates suggest that women constitute 60–80% of African smallholder farmers. More recent research suggests that female share of crop production in Africa is closer to 40-50%, but women are likely overrepresented in the most marginal and informal subsistence agriculture value chains in the continent. See Doss. 2018. ‘Women and Agricultural Productivity: Reframing the Issues’. Development Policy Review (www.onlinelibrary.wiley.com/doi/epdf/10.1111/dpr.12243). 13 MasterCard Foundation. 2018. ‘Young Africa Works’. 14 FAO. 2014. ‘Contribution to the 2014 United Nations Economic and Social Council (ECOSOC) Integration Segment’ ( www.un.org/en/ecosoc/integration/pdf/foodandagricultureorganisation.pdf). 15 Twenty out of 47 reporting African Union countries were on track for their 2025 CAADP/Malabo Declaration commitments by the end of 2017, according to the Africa Agriculture Transformation Scorecard. See 2017 scorecard data in Alliance for a Green Revolution in Africa (AGRA). 2018. ‘Africa Agriculture Status Report’ (www.agra.org/wp-content/ uploads/2018/10/AASR-2018.pdf). 16 African Development Bank (AfDB). 2016. ‘Feed Africa: A Strategy for Agricultural Transformation in Africa 2016-2025’ (www.afdb.org/fileadmin/uploads/afdb/Documents/Generic-Documents/Feed_Africa-_Strategy_for_Agricultural_ Transformation_in_Africa_2016-2025.pdf). 17 See, e.g., African Centre for Economic Transformation (ACET). 2017. ‘African Agricultural Transformation Report’ (www. acetforafrica.org/acet/wp-content/uploads/publications/2017/10/ATR17-full-report.pdf); Alliance for a Green Revolution in Africa (AGRA). 2018. ‘Africa Agriculture Status Report’ (www.agra.org/wp-content/uploads/2018/10/AASR-2018.pdf). Chapter 2 18 It is important to clearly define not only what the D4Ag sector is, but also what it is not. One such critical distinction important for the scope of this report is to separate D4Ag solutions from the much broader category of agriculture technology (AgTech), which includes many important technologies that are either not digital (e.g., basic farm machinery and tools) or where digital elements play a secondary role to other innovations (e.g., farm robotics and automation, biotech and biochemistry, innovative food and farming systems such as indoor “vertical” farms, off-grid energy solutions for agriculture that lack digital business model components, etc.). For a helpful visualisation of digital agriculture vs. AgTech, see USAID Feed the Future. 2018. ‘Policy Brief #5: ICT Solutions for Inclusive Agricultural Value Chains’, available at (https://www. agrilinks.org/sites/default/files/brief_5_-_ict_solutions_for_agricultural_value_chains.pdf). ENDNOTES 209 19 For instance, digital advisory services are sometimes categorised as digital farmer information or digital extension services (e.g., GIZ, USAID), farm management software is sometimes separated out into its own use case area as distinct from digital advisory services (e.g., BMGF, AgFunder), digital market linkages to inputs and market linkages to off-take markets are sometimes split out into separate use case areas (e.g., World Bank), digital tools for data collection and M&E are sometimes seen as a separate end-use case (e.g., GIZ) rather than as underlying data collection and data analytics tools that support other use cases. For alternative D4Ag frameworks reviewed as source materials for this report see, e.g., USAID Feed the Future. 2018. ‘Policy Brief #5: ICT Solutions for Inclusive Agricultural Value Chains’ (https://www.agrilinks. org/sites/default/files/brief_5_-_ict_solutions_for_agricultural_value_chains.pdf); GIZ. 2017. ‘Use of ICT for Agriculture in GIZ Projects’; World Bank. 2017. ‘ICT in Agriculture e-Sourcebook’; Bill and Melinda Gates Foundation. 2017. BMGF ICT4Ag Strategy; World Bank. 2019. ‘Africa Smallholder Agriculture Digital Disruption Conference Proceedings’ (full report forthcoming); USAID. 2018. ‘Digital Tools In Agricultural Programming’; FAO. 2013. ‘ICT Uses for Inclusive Agricultural Value Chains’. See also AgTech investment and innovation ecosystem maps from AgFunder (www.agfunder.com) and BriterBridges (www.briterbridges.com). 20 USAID. 2018. ‘Data Driven Agriculture: The Future of Smallholder Farmer Data Management’ (https://www.usaid.gov/ digitalag/documents/data-driven-agriculture). 21 See, e.g., GFAR, GODAN, CTA. 2018. ‘Digital and Data-Driven Agriculture: Harnessing the Power of Data for Smallholders’, available at https://f1000research.com/documents/7-525; USAID, 2018. ‘Data Driven Agriculture: The Future of Smallholder Farmer Data Management’ (https://www.usaid.gov/digitalag/documents/data-driven-agriculture). 22 See GSMA’s case study of the earlier version of Esoko’s business model, available at https://www.gsma.com/ mobilefordevelopment/wp-content/uploads/2016/02/Case_Study_-Esoko.pdf; currently the Esoko model has evolved substantially with several products including a farmer information services solution (with ~1 million registered farmers), a data collection tool (Insyt), and – the organisation’s current primary focus – a market linkage platform solution called the Digital Farmer Service (DFS) (see https://esoko.com/). 23 See an overview and assessment of the CKW program in Van Campenhout (2016), available at https://www.tandfonline. com/doi/full/10.1080/1369118X.2016.1200644. 24 For an overview of several such early-stage MNO mAgri solutions see the GSMA mAgri case studies, available at https:// www.gsma.com/mobilefordevelopment/resources/mfarmer-case-studies/. 25 The 80-28 Farmer Hotline is an SMS/IVR-based farmer information system with roughly 4 million registered farmers today, making it the single largest D4Ag solution in Africa. For more information, see http://www.ata.gov.et/programs/ highlighted-deliverables/8028-farmer-hotline/. 26 The Zambia Integrated Agricultural Management Information System (ZIAMIS), which was launched in 2017 and has 1.5 million registered farmers in the country, was initially a platform for real-time management of payments and monitoring of agricultural transactions but is increasingly being used as a mass SMS service for smallholders. 27 Since 2017, Kenya’s Agriculture and Livestock Research Organisation (KALRO) has launched 17 D4Ag applications which offer step by step information to manage chickens, crops like avocado, banana, garlic, and cassava, or how to diagnose and manage specific plant diseases and pests (e.g., army worm, maize lethal necrosis). See https://www.scidev.net/sub-saharan- africa/agriculture/news/kenya-mobile-apps-transform-agriculture.html. 28 The Smart Nkunganire System (SNS) was developed in 2018 by BK Techouse, a sister company to Bank of Kigali. In partnership with the government of Rwanda, by mid-2019, over 1.4 million farmers and all agro-dealers in the country have been registered and validated within SNS and actively use it to receive advisory messages and market information. See https://ktpress.rw/2019/05/bank-of-kigali-launches-ikofito-boost-agriculture-financing/. 29 See Figure 18 in Chapter 3 for a more in-depth discussion of MNO D4Ag business models in general and Viamo and Orange in particular. 30 iShamba is an SMS and call centre-based farmer information service in Kenya that provides smallholder farmers agricultural advice, crop management best practices, weather updates, and market price information. Started in 2015, the solution has ~350,000 registered farmer clients. See https://ishamba.com/. 31 iCow is a mobile phone agriculture advisory platform, which utilises push SMS services and a call centre to offer farmers advice on dairy, poultry, and soil management practices. Started in 2012, iCow currently has over 820,000 registered farmers. 32 While Verdant includes SMS-based farmer information advisory service in Nigeria (see https://verdant.ng), the solution is much broader in nature with market linkage and macro agri-business intelligence elements. 33 FarmerLine, launched in 2013 and currently reaching ~200,000 registered farmers, has a number of D4Ag services in its portfolio; the 399 Farmer Information Service, which is an extension of Farmerline’s original business model, provides smallholder farmers weather forecasts, market prices, and information about cultivation methods and quality farm inputs via SMS and voice message in nine West African languages. See https://farmerline.co/. 34 The Regional Agricultural Trade Information Network (RATIN), a service of the Eastern Africa Grain Council provides SMS-based market price and volume information to smallholders at large scale; ~400,000 are farmers registered for the service in 2018 in Kenya, Uganda, Tanzania, Burundi, and Rwanda. See www.ratin.net. 35 Since 2011 ECX features an SMS/IVR market data dissemination service (see http://www.ecx.com.et/?AspxAutoDetectC ookieSupport=1); the ECX itself is a commodity exchange, with increasingly digitalisation of trading features (e.g., e-auction functionality), so this example only highlights the SMS market info service. 36 For some attempts in the literature to define precision agriculture and emerging precision advisory services for smallholders see World Bank. 2019. “Future of Food: Harnessing Digital Technologies to Improve Food System Outcomes”; USAID. 2018. “Digital Farmer Profiles: Re-Imagining Smallholder Agriculture”. 37 See http://www.climark.org/. 38 Ignitia is a Swedish social enterprise currently focused on West Africa (Mali, Côte d’Ivoire, and Ghana) whose product is a 48 hour weather forecast, including monthly and seasonal predictions, delivered daily via SMS to smallholder farmers’ phones in partnership with African MNOs (http://www.ignitia.se/iska). 39 See further details in https://www.apcam.org/index.php/documents/rapports-divers/165-les-lecons-apprises-du-projet-ewea- fis-au-mali-cercles-de-kolokani-et-de-diema/file 40 Weather Impact is a Dutch enterprise founded in 2014 which focuses on innovative solutions to manage the risks of extreme weather and climate change. The company has four weather-based solutions for Africa smallholder farmers deployed jointly 210 ENDNOTES with partners, Rain4Africa in South Africa, CropMon in Kenya, AgriCoach in Burundi, and CommonSense in Ethiopia, which combine weather, satellite, and – in some cases – soil data to deliver customised SMS-based advisory and early warning weather services to farmers. See https://weatherimpact.com/about-us/. 41 See the proceedings of the 2018 Fall Armyworm Tech Prize challenge (https://fallarmywormtech.challenges.org). 42 Plantwise, launched in 2012, is a global donor-funded network of health plant clinics and plant doctor agents that advise farmers on how to diagnose and treat pests and diseases; Plantwise has been digitalising its model with an online Plantwise Knowledge Bank, a number of D4Ag solutions for plant doctors, and is also experimenting via partners (e.g., Plantix) on delivering pest and disease management directly to farmers’ phones (www.plantwise.org). 43 Waterwatch Cooperative is an NGO which is scaling an AI-enabled pest and disease surveillance and advisory system in East Africa, reaching 500,000 registered farmers in 2019 (see https://waterwatchcooperative.com). 44 WeatherSafe is a UK enterprise that is scaling a pest and disease and weather risk management advisory product for coffee farmers in Rwanda and Tanzania (see http://weathersafe.co.uk). 45 Agripredict is an AI-aided pest and weather risk management solution in Zambia (http://www.agripredict.com/). 46 Sat4Farming, launched in 2017, is a Netherlands G4AW (Geodata for Agriculture and Water) funded consortium of Touton (Mars’ cocoa trader), Satelligence, and Grameen, to deliver customised advice and individualised seven-year Farm Development Plans to small-scale cocoa producers with the help of satellite imagery. See https://utz.org/corporate-news/ ghanaian-farmers-benefit-new-sat4farming-program/. 47 ACCORD is a donor-funded pilot that has been specifically developed to help smallholder coffee farmers in Africa improve crop quality and yield by combining Earth-i’s very high-resolution satellite imagery with WeatherSafe’s data platform, to provide extensive crop, weather and pest analysis, and share the information via a mobile app. See https://earthi.space/ accord/ https://earthi.space/accord/. 48 Orange Garbal, a service privately operated by telecom company Orange Mali in partnership with SNV and with funding support from the Netherland Space Office (NSO), was established in 2017 and aims to improve the resilience of pastoralists to climate change through the access and use of geo-satellite data (see http://www.snv.org/update/garbal-information- service-increases-pastoralists-resilience-mali). 49 Started in 2017 by Agrics, Geodatics is a precision advisory service that integrates satellite imaging and farmer data to deliver geospatially tailored advice (see http://geodatics.net/). 50 Market-led User-owned ICT4Ag-enabled Information Service (MUIIS), launched in 2015, is one of the Dutch Ministry of Foreign Affairs initiatives called G4AW (Geodata for Agriculture and Water) implemented by CTA and now transitioned into a sustainable business – https://muiis.com/. 51 CropIn, which has roughly 2 million farmer clients globally of whom the majority are in India but several hundred thousand are also in Africa, targets agribusiness clients but one of the main sources of value that CropIn delivers to its clients are remote-sensing based advisory services for smallholders (see https://www.cropin.com/). SatSure likewise relies primarily on satellite data for its farmer focused advisory services, financial risk assessment tools, and macro intelligence offering (see https://www.satsure.co/). 52 PAD, launched in 2015, works in Kenya, Ethiopia, and Rwanda and has significant scale, with 650,000 smallholder farmers across these three countries registered for PAD applications and services in 2018 (see https://precisionag.org). 53 CTA and Dalberg are tracking ~30 drone agriculture solution providers in the region, with headquarters in 13 African countries and operations and/or discrete projects in several dozen more. While these drone enterprises offer a variety of solutions for smallholder agriculture, the majority have an advisory component or are working with digital advisory partners. For more information on examples mentioned here, see AgrInfo/Jembe (http://www.agrinfo.co.tz/), Ziongate Geospatial/AirborneAgric (https://airborneagricsolutions.com/), ThirdEye (http://www.thirdeyewater.com/), Astral Aerial (http://astral-aerial.com/agriculture/) , AcquahMeyer Drone Tech (https://amdronetech.com/), Charis (http://charisuas. com/#home), and WeFly Agri (https://www.weflyagri.com/en/). 54 Yara’s ImageIt is a farming application designed to measure nitrogen uptake in a crop (e.g., oilseed, wheat, and barley) and to generate a nitrogen recommendation based on the resulting photo using machine learning (see https://www.yara.us/crop- nutrition/tools-and-services/imageit/). 55 For more details on the application, developed jointly by PlantVillage and IITA, see https://plantvillage.psu.edu/. 56 For more details on Yiri Drotro, see http://grainotheque.ci/. 57 Plantix, a mobile advisory application for farmers and extension workers, developed by PEAT, a Berlin-based D4Ag startup in 2015 is an image-based diagnostic tool for plant diseases and nutrient deficiencies that is able to detect more than 240 plant pests and diseases automatically. It is used by over 700,000 smallholder farmers monthly, 80% of them in India. While Sub-Saharan Africa has not been a focus to date, Plantix has already expanded to North Africa last year and Sub-Saharan Africa expansion is part of the enterprise’s strategy. Other solutions utilising a similar image processing and machine learning approach are likewise on the way with funding from donors like BMGF. 58 See https://cropnuts.com/portfolio-item/smallscale-farmers/. 59 AgroCares, launched in 2013, currently focuses on 7 African countries, expanding to 11 in 2019 for precision advisory and diagnostics services (see https://www.agrocares.com). 60 PlantVillage (see note 55) is currently experimenting with using the Croptix sensor in Africa for integration with PlantVillage’s diagnostic application (see https://plantvillage.psu.edu/solutions). 61 Yield Sky is designed for smallholder farmers to mount on a stick and walk around the farm to generate a detailed farm health scan via a Normalised Difference Vegetation Index (NDVI) that shows stressed crops, pests, diseases, and nutrient deficiencies. See https://www.zenvus.com/products/yield/. 62 UjuziKilimo, launched in 2015, uses sensor technology to measure soil characteristics, relay the information in real time to an analysis centre comprising a comprehensive database; and relay the information with the crop breed, fertiliser required, pest control, markets and other farm management tools to the farmer, in real time, through his/her mobile phone. See https://www.ujuzikilimo.com/. 63 Lentera is a Kenyan agriculture technology start-up (2016), which combines field sensors and satellite imaging to delivery precision agriculture advisory services to smallholders over their phones (https://lenterafrica.com/). ENDNOTES 211 64 SunCulture’s (http://sunculture.com/) soil sensors that are deployed alongside the company’s off-grid solar irrigation pumps and feed into the enterprise’s digital advisory platform. 65 Zenvus Smartfarm is an intelligent electronics sensor which when inserted in a farm soil collects pertinent data like humidity, temperature, pH, moisture, nutrients etc. and wirelessly transmits the data to a cloud server where advanced computational models translate this data into advisory recommendations via the Zenvus application. See https://www.zenvus.com/ products/smartfarm/. 66 See https://www.ibm.com/case-studies/t869341z93257n45. 67 See https://microsoftcaregh.com/2019/05/08/ai-edge-iot-agriculture-microsoft-farmbeats-farmers-kenya/. 68 See https://www.itu.int/en/ITU-D/Regional-Presence/AsiaPacific/SiteAssets/Pages/E-agriculture-Solutions-Forum-2018/ TCS%20Digital%20Farming%20Initiatives_Shankar%20Tagad_ESF%202018_v0.2.pdf. 69 See http://kitovu.com.ng/. 70 For a profile of iShamba see note 30. For ATA’s 80-28 Hotline see note 25. Mlimi Hotline is a multi-modal farmer call centre established by Farm Radio Trust in 2016 in Malawi to provide affordable, actionable and timely on-demand advisory services to farmers. The approach uses inbound and outbound calls combined with IVR, SMS services, and internet messaging, working in sync with participatory radio to raise awareness and encourage farmers to use the service. See https://www.facebook.com/FarmRadioMw/posts/mlimi-hotline-is-a-multi-modal-farmer-call-centre-established-to-provide- afforda/1562864670478907/. 71 We estimate that at least 20% and likely as many as 30-35% of D4Ag advisory solutions in Africa today have some IVR functionality which reach over 10 million smallholder farmers registered for associated D4Ag solutions. For an overview of how IVR solutions function in the agriculture space, see https://www.agrilinks.org/sites/default/files/resource/files/ Presentation_final_4.13.pdf. 72 Awaaz.De is a technology company specialising in last mile communications to base of pyramid populations; in the D4Ag space Awaaz.De helps manage a number of agriculture advisory solutions in India, but also works as a technology partner with a number of D4Ag enterprises in Africa like Digital Green. See https://www.awaaz.de. 73 Arifu, which has nearly 1 million registered farmer users as of Q1 2019, positions itself as a smart personal learning companion and content marketplace that helps farmers (along with others) access free educational content over SMS and chatbot interfaces. Arifu, a B2B model that works with agribusinesses, financial institutions, and NGOs to support their farmer clients, initially focused on financial literacy education for smallholders but has over time developed a broader set of content on smallholder agronomy techniques for partners and clients like Syngenta Foundation and Safaricom’s DigiFarm. While not a magic bullet for farmer engagement (depending on underlying quality of content and Arifu partners’ business models for adding value to farmers), evaluations have shown that Arifu’s model significantly improves farmer engagement and retention of content given the interactive design, behavioural nudge techniques, and participatory features (e.g., learning proceeds at farmers’ pace and content is customised/adapted based on farmer responses in the chatbot). See https://www. arifu.com/ and http://pubdocs.worldbank.org/en/211611556636989321/2-Arifu-Overview-for-WB-DAT-Challenge.pdf. 74 See https://farm.ink/# and https://www.facebook.com/africafarmersclub/. The enterprise reaches over 100,000 farmers today in Kenya; half of whom are active users on the platform. 75 MyAgriGuru uses natural local language interface text and voice chatbots to respond to farmer queries for advisory information and to facilitate plant disease diagnostics. In mid-2018 MyAgriGuru was being used by ~400,000 Indian smallholders and the solution is targeting 3 million users by the end of 2019. See https://www.myagriguru.com; see also https://dribbble.com/shots/6474077-MyAgriGuru-India-s-1st-Agri-Advisory-Chat-Bot. 76 Digital Green, founded in 2008, currently reaches nearly 2 million farmers globally of whom ~500,000 are in Africa, primarily in Ethiopia. For a selection of case studies and evaluations of Digital Green’s participatory model, please see https://www.digitalgreen.org/case-studies/. 77 Started in 2015 in Kenya and now active in Kenya, Uganda, and Tanzania with more than 1.4 million farmers on the platform as of late 2018, WeFarm is built around the principle that rural farming communities in developing countries have generations worth of knowledge to share, but lack the tools to do so. WeFarm provides an SMS service based around peer-to-peer, crowdsourcing of knowledge. Users ask a wide range of questions regarding farming techniques and share information around business ideas, or how to improve livelihoods. See https://wefarm.org/. 78 N-Frnds, founded in 2014, and initially focused on Rwanda now has over 15 million registered users globally of whom a significant number (in the millions) re African smallholder farmers. The solution is a cloud-based digital distribution platform which utilises technology innovation (USSD 2.0) to enable farmers with feature phones to gain access to sophisticated interactive features even in the absence of mobile data, including group chat, email, and interactive agriculture advisory content (Nfrnds mAgri) which is deployed to local markets via MNO, agribusiness, and other B2B clients. Beyond its mAgri advisory application, Nfrnds is also used by agribusiness to manage agent networks and farmer interaction, such as, for instance, 200,000 farmers in the Rwanda potato value chain. For more details see https://www.nfrnds.com. 79 See, e.g., https://www.businessdailyafrica.com/corporate/enterprise/WhatsApp-farmers-chatting-their-way- to-profits/4003126-5041800-131klw5z/index.html and https://ict4dblog.wordpress.com/2018/12/11/how- whatsapp-strengthens-livelihoods-of-women-farmers-in-rural-zimbabwe/; see also https://www.researchgate.net/ publication/326958580_WhatsApp_Model_for_Farmer_Led_Extension_Linking_Actors_and_Generating_Localized_ Information_for_Farmers. 80 By some estimates, the number of Whatsapp users in Kenya is 8-10 million monthly vis-à-vis a total population size of 50 million (see https://techweez.com/2018/11/02/kenyas-communication-authority-looking-into-whatsapp-regulation/ and Hootsuite, Digital 2019: Kenya report, available at https://cnyakundi.com/state-of-the-internet-number-of-kenyas-active- twitter-monthly-users-drop-by-half-after-censorship/). 81 See the GeoPoll Kenya smallholder survey of 900 farmers with phones chosen from a nationally representative 18,000 farmer panel (see https://www.geopoll.com/blog/data-farming-kenya-mobile-phone/). The survey results suggest that while 15% of Kenyan farmers were using “Farming Apps” a full 7% where using Whatsapp groups for the “farming needs” (likely including both information and market linkage uses). 82 Launched in 2018 by Intersoft Eagle, the SmartCow app offers the usual advisory features but also enables farmers to monitor their expenditure and income and to capture and analyse the history of each and every animal including the production levels for milk. See http://farmbizafrica.com/machinery/1895-nairobi-company-launches-mobile-app-to-help- dairy-farmers-maintain-records. 212 ENDNOTES 83 DigiCow, launched in Kenya in 2014 by Farmingtech Solutions, is a simple record-keeping app for dairy farmers which targets smallholder farmers and enterprises engaged in dairy farming enabling the farmer to increase their profits through data driven decision-making. The app’s functionality is currently being expanded to enable farmer-to-farmer chat groups, market linkages (e.g., to vets), and linkages to financial providers. See http://digicow.co.ke/. 84 AgroInnova’s AkokoTakra (2017) is a farm management software application for phones, tablets, and PC that enables Ghanaian poultry farmers to record, monitor, keep track and analyse all their farm operations easily including feed, drugs, birds, eggs collection, sales, and input purchases. See https://www.akokotakra.com/. 85 Launched in 2018 in Senegal, Sen Ngunu offers the solution to manage the entire production chain of one’s poultry farm, adapted to poultry farmers at small scales. With their partners they offer a management solution, coaching and training and a management smartphone app with budgeting, planning, record-keeping, and advisory features. See http://senngunu.com/. 86 Launched in 2017, Probity Farms is a simple advisory solution for smallholder farmers. It helps them plan their farm management, inventory management, and also their accounting. The solution is specially targeted towards those who are new to farming. The platform helps them make a business out of farming and guides them through the everyday activities of farm operations. See https://probityfarms.com/. 87 AgriGo, founded in 2016, is an advisory platform with some farm management components including recordkeeping of all farmer purchases and activities (tracked with USSD) and the ability to calculate costs of production and track expenses. AgriGo to date has signed up 30 cooperatives in Rwanda, through which they serve a total of 90,000 individual farmers and supports rice, maize, and potatoes. Revenue comes from account management fees (paid by cooperatives) or user subscription fees (paid by independent farmers). See https://agrigo.rw/ and https://i2ifacility.org/system/documents/ files/000/000/069/original/AgriGO_-_A_farmer’s_financial_tool_to_grow_greater_finanical_harvest_i2i_July_2018. pdf?1532604835. 88 Launched in 2017, BudgetMknoni is a farm budgeting and recordkeeping application for smallholder farmers launched by the iShamba team. See https://budgetmkononi.com/. 89 See https://www.agrivi.com/en. 90 See a general discussion of these factors at https://dev.meas.illinois.edu/wp-content/uploads/2015/04/Ferris-et-al-2014- Linking-Farmers-To-Markets-MEAS-Discussion-Paper.pdf. 91 For the most relevant overview to date of digital smallholder market linkage and e-commerce models, see Mercy Corps AFA & Dalberg. 2018. ‘Benchmarking E-Commerce Models for African Smallholders’, at https://www.findevgateway.org/ sites/default/files/publication_files/afa_ecommerce_benchmark_slideshare_9.17_fnl.pdf; for other perspectives on the digital opportunity for input and off-take market linkages, see World Bank. 2016. ‘Will Digital Technologies Transform Agriculture in Developing Countries?’, available at http://documents.worldbank.org/curated/en/481581468194054206/pdf/WPS7669. pdf, and USAID. 2018. ‘Where and How Digital Tools Impact the Value Chain’, available at (https://www.usaid.gov/sites/ default/files/documents/15396/Why_Where_and_How_Digital_Tools_Impact_the_Value_Chain.pdf). 92 Likely the best existing attempt to characterise digital market linkage business models are the MercyCorp AFA & Dalberg (2018), cited above, and AGRA. 2016. ‘Digital Harvest’, with the report available at https://www.raflearning.org/sites/ default/files/20161024_digital_harvest_final_report.pdf and case studies at https://www.raflearning.org/sites/default/ files/20160929_digital_harvest_case_studies_final.pdf. 93 See Ibid; see also the forthcoming research from MasterCard Foundation on digitally-enabled integrated value chain players like Tulaa and Safaricom’s Digi-Farm (see https://www.raflearning.org). 94 Farmers Pride (https://farmersprideafrica.com/), with ~10,000 smallholder farmers today leverages technology and franchising to give Kenyan farmers access to high quality inputs via an online mobile platform that connects farmers to the nearest verified vets, agronomy, inputs and insurance service providers, as well as real time climate information. The app platform also digitalises and links together existing village-level input shops thus combining the benefits of both digital and human linkages (see http://pubdocs.worldbank.org/en/622751556637637102/Farmers-Pride.pdf) 95 CowTribe(https://www.cowtribe.com/) is a Ghana-based for-profit organisation focused on supporting livestock farmers via a mobile platform that aggregates demand for livestock farming inputs and services, starting with vaccinations and veterinary services (DRK Foundation). CowTribe’s service connects cows to vaccines and veterinarians. It is unique in West Africa and has attracted 30,000 users and substantial investment. 96 myAgro (https://www.myagro.org/) started in 2011 and now working with more than 50,000 farmers, is a mobile layaway commitment savings model for agri-input financing. myAgro operates by linking the aggregated farm input demand from smallholder farmers to high-quality input suppliers via local agro-dealer stores. 97 Agrics (http://www.agrics.org/), started in 2014 and initially supported by the Dutch G4AW programme, is a for-profit enterprise that is currently serving 35,000 farmers and generates revenue by procuring farm inputs in large quantities and selling them, on credit, with a target gross margin above 30%. 98 iProcure (https://iprocu.re/) is a digital B2B start-up working on optimising the agricultural input supply chain in Africa. The enterprise has established a network of 5000 farm input agro-dealers, which it supports with technology tools that help them improve their operations through business intelligence, improved inventory management, and streamlined distribution efficiency. iProcure is currently linking >25k farmers to inputs and has ambitious plans for scale and big scaling partners like Safaricom’s DigiFarm. 99 For One Acre Fund’s integration of digital technologies into their value chain approach, see USAID’s 2017 case study of the organisation, available at https://www.usaid.gov/sites/default/files/documents/15396/One_Acre_Fund_Case_Study.pdf. 100 For a recent comprehensive profile of DigiFarm, see IFC. 2018. ‘Digital Financial Services for Agriculture Handbook’, available at https://www.ifc.org/wps/wcm/connect/4ca05121-fe39-42ae-891f-76203c7b91f0/Digital+Financial+Services+for +Agriculture_IFC%2BMCF_2018.pdf?MOD=AJPERES. 101 DigiFarm has registered roughly a million farmers in Kenya, but only a relatively small share of these clients is currently receiving inputs and input financing as the organisation scales up the market linkage element of its model; our interviews with the organisation and sector experts suggest a great deal of optimism for the platform’s potential to generate more farmer engagement and scale as the value proposition gets further refined. 102 Twiga Foods (twiga.ke), founded in 2014, runs a mobile-based B2B food supply platform combined with physical infrastructure for farmer engagement, produce aggregation, and transport logistics that supplies fresh fruits and vegetables sourced from >16,000 farmers in rural Kenya to small- and medium-sized vendors, outlets and kiosks in Nairobi. Twiga is able to offer higher prices and a guaranteed market to farmers, and lower prices and a reliable supply to vendors. Twiga has raised more than $35 million to date, a record for the African D4Ag sector. See GSMA. 2018. ‘Twiga Foods’, available ENDNOTES 213 at https://www.gsma.com/mobilefordevelopment/wp-content/uploads/2018/05/Twiga-Foods-Improved-market-access-for- farmers-and-a-reliable-supply-for-vendors.pdf. 103 Selina Wamucii (www.selinawamucii.com/), has as its mission the integration of African smallholder farmers into high quality global supply chains for products like avocados, bananas, and fish, and is currently working in six African countries. Farmshine (www.farmshine.io/) helps smallholder farmers aggregate and sell their harvests directly to reliable commodity companies in Kenya with the help of field agents and a proprietary agent and buyer application. Taimba (www.taimba. co.ke) provides rural small-scale farmers in Kenya with direct linkages to urban traders. Similarly, Trade (www.tradeghana. co/) uses digital technology melded with a physical agent and storage warehouse network to play the role of maize value chain integrator in Ghana. Ninayo (https://www.ninayo.com) started as a virtual marketplace but has involved with more value additive intermediation activities. 104 See Digital Green Loop (www.getloopapp.com) model overview in https://www.digitalgreen.org/wp-content/ uploads/2017/06/Digital-Green-Loop-brief-June2017.pdf. 105 Tulaa (www.tulaa.io) has a unique digitally-enabled end-to-end value chain formalisation business model, currently reaching <5k farmers at the pilot stage. The company provides pre-screened quality inputs on credit to smallholder farmers based on a proprietary alternative data credit scoring tool, manages the logistics of input orders and delivery via its digital platform, and then brokers the sale of farmers’ crops at harvest time. See brief Tulaa profile in CGAP. 2019. ‘Fintechs and Financial Inclusion: Lessons Learned’, available at: https://www.cgap.org/sites/default/files/publications/2019_05_Case_Study_ Fintech_and_Financial_Inclusion.pdf; see also the forthcoming in-depth independent assessment of Tulaa’s business model and economics from MasterCard Foundation’s Rural Agriculture Finance Learning Lab (www.raflearning.org/), Dalberg, and IDH. 106 Launched three years ago, Akorion (www.akorion.com) has at this stage reached ~60,000 farmers with its services working with a network of ~500 digitally-enabled village agents. 107 The FtMA (www.ftma.org), an alliance of eight agri-focused organisations, including large agribusiness partners, currently supports ~150,000 East African smallholder farmers “from seed to market” with inputs, finance, and market facilitation; of these ~60,000 are now supported via FtMA’s digital platform. See MercyCorps AFA & Dalberg. 2019. ‘FtMA Digitalization Lessons Learned’, available at http://mercycorpsagrifin.org/wp-content/uploads/2019/01/AFA-FtMA_Digitization-and- lessons-learned_FIN.pdf. 108 For an overview of these models, see MasterCard Foundation RAFLL. 2017. ‘How can digital tools enable smallholder finance’, available at https://www.slideshare.net/MaliaBachesta/raf-ll-wapl-session-5. 109 See, e.g., https://www.theguardian.com/sustainable-business/2016/feb/11/internet-veg-box-schemes-africa-kenya-rwanda- gambia-farming. 110 The AfDB estimated in 2017 that the African middle class is already 350 million people out of a total population of 1 billion (35%) and is likely to grow to 43% of the population by 2030, see https://www.afdb.org/fileadmin/uploads/afdb/ Documents/Publications/AEO_2017_Report_Full_English.pdf; alternative assumptions, like those from Credit Suisse lead to much more conservative numbers, but the continued growth of the middle class (in terms of both numbers and numbers) is incontrovertible. 111 See https://www.rockefellerfoundation.org/blog/africas-farmers-ready-supermarkets-revolution/. 112 The positive farmer value proposition is largely anecdotal but attested by many of our interviewees. In terms of reach, most of these models are still relatively small. While precise figures are not publicly available, we estimate that the dozen or so such businesses in our database work with 50,000 to 100,000 smallholder farmers across Africa, suggesting that most players are still quite small in their production volumes. The continued attention to such models from investors and a steady flow of publicly announced VC or follow on deals suggests, however, that the investment community sees viable economics and potential for greater scale. 113 For more details on a few of these types of enterprises, see, e.g., IzyShop (https://izyshop.co.mz/), FarmFresh (www. farmfresh.gm), HMart (www.shop.mart.rw), Foodstock (www.foodstock.com.ng/), Farmart (www.farmartghana.com), Khula (www.khula.co.za), and Herdy (www.herdy.co/). 114 For more information on these models see Afrimash (www.afrimash.com/), FarmIT (farmit.co.ke), and eMsika (www.emsika. com/). 115 See FAO (2013), where such electronic marketplaces are also labelled as virtual trading floors (VTFs). 116 For more details on the MasterCard Farmer Network (MFN), see https://newsroom.mastercard.com/mea/press-releases/ mastercard-recognised-with-best-agtech-solution-award-in-kenya/. 117 Ninayo’s (https://www.ninayo.com) original business model was fairly typical of such solutions. The service was set up as a two-sided virtual buy/sell platform, with ~25,000 farmers registered, in which farmers could advertise their crop holdings and buyers could advertise their crop needs. The two were able to find each other through an online interface (currently available only via smartphones, but with a USSD product in development), and could link up for the sale. In recent years Ninayo have been moving away from a pure marketplace model and has started to take on middleman trading functions via its own agents (i.e., migrating to an integrated off-take value chain model). For more information on other examples, see Usomi’s Rubi (www.usomi.com/), Mifugotrade (https://livestock.herokuapp.com/), Farmster (www.farmster.co/), Animartt (www.animartt.com/), Zowasel (https://www.zowasel.com/), and eFarm (https://www.efarm.cm/). 118 See TruTrade (http://www.trutradeafrica.net/) and AgroTrade (https://agrocenta.com/about). 119 For more details, see FarmAll (https://farmallke.com/) and http://www.agromarketday.com/. 120 See Lima Links (http://www.limalinkszambia.com/) and Farmerline (https://farmerline.co/). 121 For more details, see Agrikore (www.cellulant.com/agrikore/). 122 See FAO. 2018. ‘Sustainable Agricultural Mechanization: A Framework for Africa’, available at http://www.fao.org/3/ CA1136EN/ca1136en.pdf. 123 Ibid. 124 German Development Institute (GDI). 2017. ‘Unlocking the Power of Irrigation for Sub-Saharan Africa’, available at https://www.die-gdi.de/uploads/media/BP__7.2017.pdf; see also Liangzhi Y. et al. 2010. ‘What is the irrigation potential of Africa’, IFPRI. 125 Malabo Montpellier Panel. 2018. ‘Mechanized – Transforming Africa’s Value Chains’, available at https://www. mamopanel.org/media/uploads/files/MaMo2018_Mechanized_Transforming_Africas_Agriculture_Value_Chains.pdf. 214 ENDNOTES 126 Ibid. 127 See, e.g., https://www.aatf-africa.org/programmes/mechanization-and-digital-agriculture/. 128 See an overview of this business model at https://agra.org/news/uber-for-tractor-at-work/. 129 See, e.g., https://smartagripost.com/trimble-showcases-laser-land-leveller-for-chhattisgarh-farmers-promises-up-to-30-water- savings-10-gain-in-crop-yields/. 130 Hello Tractor (www.hellotractor.com), founded in 2014 and with a reported client base of 250,000 African farmers, is an IoT platform that works across the entire tractor ecosystem from OEMs, to tractor distributors, to local tractor entrepreneurs/investments via digital applications that support fleet management, fleet monitoring, and shared economy tractor demand-matching services for farmers. HelloTractor allows farmers to rent tractors from owners for a predetermined amount of time and also stacks functionalities to increase value for its customers: it is a booking agent platform, offers alerts for maintenance and technicians to service the tractor, and utilises remote sensing to offer more in-depth analytics. 131 For further details on these models, which tend to focus more on digitalising the shared economy elements of mechanisation rather than the more B2B fleet management and IoT dimensions of Hello Tractor’s model, see TroTro Tractor (www. trotrotractor.com), Kobiri (www.kobirigroup.com/), E-Tinga (www.e-tinga.com), and Farmall (farmallke.com). 132 See Trringo (www.trringo.com); several thousand farmers already using the service in Tanzania. 133 See M-KOPA Solar (http://www.m-kopa.com/), Fenix International (https://www.fenixintl.com/), BBOX (https://www. bboxx.co.uk/), Zola Electric (http://zolaelectric.com/), and PEG Africa (https://pegafrica.com/). 134 PAYG solutions had an estimated reach of 2 million African households in early 2018; see overview of sector in the GOGLA/Dalberg. 2018. ‘Off-grid Solar Market Trend Report’ (https://www.lightingglobal.org/2018-global-off-grid-solar- market-trends-report/). 135 For more details on SunCulture, see http://sunculture.com/. 136 For more information, see Azuri (https://www.azuri-technologies.com/), SimuSolar (https://www.simusolar.com/), AgSol (https://agsol.com/), and ColdHubs (http://www.coldhubs.com/). 137 See the forthcoming WB & Dalberg report on Productive Use Leveraging Solar (PULSE) in mid-2019. 138 See, e.g., discussion in USAID Feed the Future. 2018. ‘Policy Brief #5: ICT Solutions for Inclusive Agriculture Value Chains’. See also World Economic Forum (WEF). 2019. ‘Innovation with a Purpose: Improving Traceability in Food Value Chains through Technology Innovations’ (http://www3.weforum.org/docs/WEF_Traceability_in_food_value_chains_Digital. pdf). 139 African agribusiness surveys consistently highlight growing investment into technology solutions as a major area of challenge and opportunity. The 2016-2017 PWC Africa Agribusiness survey, for instance, highlighted technology access as the top challenge for the sector. Top priority technology innovations targeted by agribusinesses for investment in the survey included digital tools for demand forecasting, inventory management, digital track and trace methods to improve food safety, and digital communication and monitoring tools that can facilitate greater connectivity with smallholder farmers and field agent forces. PWC. 2017. ‘Africa Agribusiness Survey (2016/2017)’ (https://www.pwc.co.za/en/assets/pdf/agri-businesses-insights- survey-may-2016.pdf); the 2017/2018 survey is available at https://www.pwc.co.za/en/assets/pdf/africa-agribusiness- insights-survey-2017-2018.pdf. 140 This discussion draws heavily on GSMA. 2018. ‘The role of digital in improving traceability and certification in the agricultural last mile’ ( https://www.gsma.com/mobilefordevelopment/blog-2/the-role-of-digital-in-improving-traceability- and-certification-in-the-agricultural-last-mile/). 141 For example, 10 of 16 digital Africa-centred “tracking and traceability” solutions reviewed last year by USAID were export focused. See USAID Feed the Future, “Policy Brief #5: ICT Solutions for Inclusive Agriculture Value Chains” (2018). 142 The importance of sustainability and thus track and trace digital solutions highlighted in PWC 2016 and 2017 surveys (see note 139). 143 Ibid. 144 For more information on these players, see SourceTrace (www.sourcetrace.com), SourceMap (www.sourcemap.com), EProd (www.eprod-solutions.com), and FarmForce (https://farmforce.com/). 145 See https://www.sap.com/products/agriculture-supply-chain-mgmt.html. 146 See https://blog.chainpoint.com/blog/the-rainforest-alliance-selects-chainpoint-as-central-data-collection-platform-in- sustainable-supply-chains. 147 NamLITS was launched by the Namibian government in 2006 for commercial farmers and extended to communal livestock farmers in 2014, which proved to be a fortuitous bit of timing. A recent evaluation has found that during the 2015 foot and mouth disease outbreak in the country, the worst such outbreak in 40 years, NamLITS was used to minimise the impact of this outbreak and made free trade possible once again by using its advanced functionalities illustrating its effectiveness. See Prinsloo et al., “The role of the Namibian Livestock Traceability Systems in containing the recent foot-and-mouth disease outbreak,” NextComp (2017), available at https://ieeexplore.ieee.org/document/8016172. 148 See Ashour et al., “An Evaluation of the Impact of E-verification on Counterfeit Agricultural Inputs and Technology Adoption in Uganda”, IFPRI (2015), available at http://www.ifpri.org/publication/evaluation-impact-e-verification- counterfeit-agricultural-inputs-and-technology-adoption. 149 See DeBouef et al., “Counterfeiting in African Agriculture – Challenges and Solutions,” Bill and Melinda Gates Foundation (2014), available at https://www.agrilinks.org/library/counterfeiting-african-agriculture-inputs-challenges-and-solutions 150 DeBouef et al. (2018). 151 See QualiTrace (www.qualitracegh.com/about/). 152 See mPedigree (https://mpedigree.com) and Sproxil (https://www.sproxil.com/) for more details on such input verification business models and underlying technologies. 153 LORI (https://www.lorisystems.com/) and Kobo360 (https://www.kobo360.com/) models do already have relevance for African agriculture, with Kobo360 for example exploring partnerships with a number of agribusiness players in West Africa. 154 See iProcure (https://iprocu.re/), Logistimo (https://www.logistimo.com), and Virtual City (http://www.virtualcity.co.ke), and WeightCapture (http://www.weightcapture.com/). ENDNOTES 215 155 See iProcure (https://iprocu.re/). 156 For some illustrative data on the potential impacts of such solutions, see the self-reported impact reporting by Virtual City, available at https://www.virtualcity.co.ke/solution/agroforce-2/. 157 See, e.g., https://www.sap.com/products/what-is-erp.html. 158 For Africa-based, see Farmforce (https://farmforce.com/), EProd (http://www.eprod-solutions.com/), Metajua (http:// metajua.com/). 159 TaroWorks (https://taroworks.org) started a digital field force tracking and management tool incubated from Grameen Foundation’s CKW advisory services model in Uganda, but has evolved into a stand-alone digital field force management and ERP solution with features like order management, location mapping, and CRM. 160 See SourceTrace (http://www.sourcetrace.com/apps/), CropIn (http://www.sourcetrace.com/apps/), and Annona (https:// annona.co/). 161 See https://www.accenture.com/us-en/insight-accenture-digital-agriculture-solutions. 162 Vodafone’s Connected Farmer grew out of Vodafone’s role as part of the Connected Farmer Alliance. While the product is marketed as a standalone service in South Africa, in East Africa this offering is embedded in Digifarm as the B2B dimension of Digifarm’s technology stack. See https://www.vodacombusiness.co.za/business/solutions/internet-of-things/agriculture/ connected-farmer. 163 See https://www.olamgroup.com/sustainability/reimagine/olam-farmer-information-system.html. 164 Much of this discussion draws on IFC. “Handbook: Digital Financial Services for Agriculture” (2018), available at https:// www.ifc.org/wps/wcm/connect/region__ext_content/ifc_external_corporate_site/sub-saharan+africa/resources/dfs- agriculture; see also MCF RAFFL, “Inflection Point: Unlocking Growth in the Era of Farmer Finance” (2016). 165 For the older Inflection Point reports that have proved critical to framing the dialogue around financial services for agriculture, see https://www.raflearning.org/post/inflection-point-unlocking-growth-era-farmer-finance. For the most comprehensive recent report on this topic, see also IFC. “Handbook: Digital Financial Services for Agriculture” (2018), available at https://www.ifc.org/wps/wcm/connect/region__ext_content/ifc_external_corporate_site/sub-saharan+africa/ resources/dfs-agriculture. 166 For the category of traditional financial service providers who are digitalising their business models it is often difficult to define the boundary line between those enterprises that can be classified as D4Ag solutions and those that are simply financial service providers who happened to digitalise some of their approach. To ensure clarity of definition and scope, the report tries to focus only on those institutions that have truly distinct digital products – digital channel, digital branding, heavily digitalised operations – rather than the delivery of traditional financial products with some digitalisation of background processes (e.g., SMS notifications for customer management) or background analytics tools (e.g., new credit scoring algorithms that include digital data streams). 167 World Bank Global Findex, 2017. 168 Better Than Cash Alliance (BTCA). “The Role of Digital Payments in Sustainable Agriculture and Food Security” (2017), available at https://btca-prod.s3.amazonaws.com/documents/313/english_attachments/Agriculture_Report.pdf?1508858199. 169 GSMA (2016), available at https://www.gsmaintelligence.com/research/?file=29e480e55371305d7b37fe48efb10cd6&downlo ad. 170 See IFC. “Handbook” (2018), full citation in note 100; see also BTCA (2017). 171 See BTCA (2017); see also GSMA, “Market size and opportunity in digitising payments in agricultural value chains” (2016), (https://www.gsmaintelligence.com/research/?file=29e480e55371305d7b37fe48efb10cd6anddownload). 172 As part of its work under GES in 2012 through the end of the scheme in 2017, and in collaboration with the Nigerian Central Bank, Cellulant registered 17 million farmers in the country and channelled nearly $1 billion of input subsidies to 7 million of these farmers, achieving very high levels of linkage and uptake to agricultural input purchases. The programme was discontinued when the government subsidy scheme lapsed and Cellulant has since pivoted its model, but the example of GES is still a notable one. See https://cellulant.com/blog/agritech-in-africa-how-an-e-wallet-solution-powered-nigeria- governments-ges-scheme/. 173 Zoona’s (https://ilovezoona.com/) model has evolved significantly in recent years, but the organisation is at its core a third-party provider of mobile payments focused on building a reliable, cash-in/out network and facilitating B2C and B2B payments. In agriculture, Zoona’s model was at one stage a major channel for G2P payments to farmers and later B2P payments as lead firms that contract with thousands of farmers use Zoona to reduce individual payments; the agribusiness makes one payment to Zoona, which then make e-voucher or mobile payments to each of the contracted farmers that can be redeemed with input retailers or cash-in/out agents. 174 See note 28 for an overview of SNS (https://smartnkunganire.rw/). SNS was intentionally designed to first serve as a payments and supply chain management tool for Rwanda’s national agro-input subsidy programme with the objectives of improving the programme’s efficiency, productivity, and transparency. Now that the system is in place, however, the model is evolving to give each farmer in the SNS system an ‘IKOFI’ universal digital wallet that allows farmers to send and receive money (zero transaction fee), pay agro-dealers, receive payment for their harvest, pay into the national long-term savings scheme, and ultimately pay for health care and other services via a phone (USSD/SMS) while also generating a valuable financial track record that can serve as a gateway to other financial services. See https://ktpress.rw/2019/05/bank-of-kigali- launches-ikofito-boost-agriculture-financing/. 175 An astonishing 20% of SmartMoney’s rural customers make digital payments for goods and services in their daily lives and input payments are fully digitalised in most SmartMoney communities. 176 This discussion draws heavily on IFC. 2018. ‘Handbook’, full citation in note 100. 177 See, e.g., evidence on the impact of savings (regular and commitment savings accounts) on farmer investments, yields, and incomes in Brune, L. et al. 2015. ‘Facilitating Savings for Agriculture: Field Experimental Evidence from Malawi’, NBER Working Paper No. 20946 (https://www.nber.org/papers/w20946). 178 See World Bank. 2017. Global Findex 2017. Where data is available, unsurprisingly, savings access levels are even lower for smallholder farmers than for the population at large. CGAP’s smallholder farmer diaries show savings access levels of ~10% (5-20% range) for formal savings accounts and ~15% (5-25%) for informal savings clubs across countries like Uganda, Tanzania, Mozambique, Nigeria, and Côte d’Ivoire (see https://www.cgap.org/sites/default/files/small_holders_data_portal/). 216 ENDNOTES 179 See notes 172-174 for information on Cellulant, Zoona, and SNS e-wallet models. 180 See http://fizambia.com/?p=1464. 181 Currently 35% of the Agri-Wallet (https://agri-wallet.com/) farmers who use the wallet, save. As part of an ‘ecosystem’ with earmarked credit, Agri-wallet helps farmers to save and in turn enables them to access short term loans through Rabobank. See https://www.cta.int/en/digitalisation/all/article/agri-wallet-a-wallet-for-smallholder-farmers-sid00f60f624-f62a-4b58- bd27-bd2c838b724f. 182 See myAgro (https://www.myagro.org/); see also IFC. ‘Handbook’ (2018), full citation in note 100. 183 https://www.akdn.org/our-agencies/aga-khan-foundation/akf-digital-savings-groups-dsg. 184 https://akaboxi.com/ 185 In 2016, Dalberg and ISF estimate a $200 billion global smallholder financing demand and a $150 billion financing gap. Using the Sub-Saharan Africa smallholder household population as proxy relative to the global smallholder farmer population and a 5-10% credit access estimate for African smallholders, we estimate that the Africa share of the gap is roughly 25%, i.e., roughly $25-35bn (€ 30bn). See MCF RAFLL. 2016. ‘Inflection Point Report’. 186 One Acre Fund. 2016. ‘Scaling up agricultural credit in Africa’, available at https://oneacrefund.org/documents/104/ Scaling_Up_Agricultural_Credit_In_Africa_Farm_Finance.pdf. 187 See, e.g., Dalberg & MCF RAFLL. 2018. ‘Big data could mean big opportunity: why we should stay excited for data analytics in smallholder finance.’ (https://www.raflearning.org/sites/default/files/learning_brief_5_-_data_analytics-final. pdf?token=g6FuZCx4); see Dalberg & MCF RAFLL. 2016. ‘The business case for digitally enabled smallholder finance.’ (https://www.raflearning.org/post/learning-brief-1-business-case-for-digitally-enabled-smallholder-finance). 188 For more on digital smallholder loan innovation from banks like KCB, Advans, and Opportunity International, see MCF RAFLL. 2017. ‘Case for digitalising smallholder finance’. 189 Digital MFI Musoni’s Kilimo Booster, for example, offers a flexible digital loan with grace periods and repayment plans tailored to the individual farmers’ production circumstances coupled with a fully digital field registration, loan disbursement and repayment experience. Musoni found that in addition to offering loans to farmers on terms that set them up for successful repayment, the digital platform allowed them to easily “deliver additional services via mobile, without having to constantly make changes to the core banking system.” IFC (2018). 190 Akellobanker (http://www.akellobanker.com/how-it-works) offers easy access to tractor hire, improved seed, medical services and farm labour on credit, by leveraging data and mobile technology to offer structured re-payments compatible to the user’s needs. The platform integrates mobile money and use of USSD to facilitate instant access, disbursements and repayments. The technology uses the historical data collected to generate automated digital credit scores. 191 See note 105. 192 See https://apolloagriculture.com/. 193 For the note of caution on the potential viability of these models see IFC. 2018. ‘Handbook’ (full citation in note 100). Ongoing portfolio analyses by Dalberg, IDH, and the MCF Rural Agriculture Finance Learning Lab of organisations like Tulaa, Opportunity International, and Digifarm (publications forthcoming in 2019) suggest, however, that despite many unanswered business models’ questions and challenges, at their core these models can be a viable pathway to both scale and sustainability. 194 Ibid. See also AFI. 2017. ‘Digitally Delivered Credit: Consumer Protection Issues and Policy Responses to New Models of Digital Lending.’ (https://www.afi-global.org/publications/2633/Digitally-Delivered-Credit-Consumer-Protection-Issues-and- Policy-Responses-to-New-Models-of-Digital-Lending). 195 See overview of crowdfunding models in Nigeria that informs this discussion at ICT4DBlog. 2018. ‘Crowdfarming platform enabled investment into Nigerian Agriculture’ (https://ict4dblog.wordpress.com/2018/11/20/crowdfarming-platform- enabled-investment-in-nigerian-agriculture/). 196 See https://techpoint.africa/2018/03/12/farmcrowdy-office-tour/. 197 See http://www.farmable.me/. 198 ICT4DBlog (2018). 199 Ibid. 200 For more details on FarmCrowdy’s (https://www.farmcrowdy.com/) model, see Ibid. See also: https://www. coruscatesolution.com/create-farmcrowdy-app-clone/. 201 See Growsel (https://www.growsel.org/), Thrive Agric (https://www.thriveagric.com/), Livestock Wealth (https://www. livestockwealth.com/), and Bayseddo (https://www.bayseddo.com/). 202 This discussion draws on the best recent overview of the insurance opportunity for smallholder farmers, see Initiative for Smallholder Finance (ISF). 2018. ‘Protecting growing prosperity: Agricultural insurance in the developing world.’ (https:// www.raflearning.org/sites/default/files/sep_2018_isf_syngneta_insurance_report_final.pdf?token=1i4u5GwD). 203 Ibid. 204 IFC. 2018. ‘Handbook’ (full citation in note 100). 205 For 3% estimate, see ISF. 2018. ‘Protecting prosperity’. Earlier estimates have been 6%. IFC. 2018. ‘Handbook’ (see note 100). 206 Ibid. 207 For an overview of these models, see ISF. 2018. ‘Protecting growing prosperity’. For details on each, see Pula (https://www. pula-advisors.com/), ACRE Africa (https://acreafrica.com/), Oko (https://www.oko.finance/), and World Cover (https:// www.worldcovr.com/). 208 SumAfrica is a Netherlands G4AW-supported programme, now on a commercial basis, that involves a consortium of a local insurer in Uganda (Ugandan Agro Insurance Consortium (AIC)) and the Dutch company EARS, which develops and provides the satellite-based drought index. See https://g4aw.spaceoffice.nl/en/projects/g4aw-projects/62/scaling-up-micro- insurance-in-africa-sum-africa-.html; see also, https://www.propertycasualty360.com/2019/02/12/sum-africa-project-offers- unique-insurance-service-to-farmers-in-uganda/?slreturn=20190427213616. ENDNOTES 217 209 See http://www.winners-project.org/. 210 IFC. 2018. ‘Handbook’ (see note 100). 211 See in-depth discussion in Dalberg & MCF RAFLL publications on these topics in note 187. 212 See an overview of such models in a recent study by Dalberg and MCF RAFLL, available at https://www.raflearning.org/ post/learning-brief-5-big-data-could-mean-big-opportunity-why-we-should-stay-excited-for-data. 213 For details on each of these players, see FarmDrive (https://farmdrive.co.ke/), Harvesting (https://harvesting.co/), YAPU (https://www.yapu.solutions/), and SatSure (https://www.satsure.co/). 214 See https://www.rabobank.com/en/about-rabobank/in-society/rabobank-foundation/index.html. 215 CTA. 18/04/2018. ‘Input loans boost farmer take-up rates for satellite-based advisory service’ (https://www.cta.int/en/ digitalisation/all/article/input-loans-boost-farmer-take-up-rates-for-satellite-based-advisory-service-sid0c75ed4b1-173d-4c1d- a899-be03866bd3f3). 216 CTA (forthcoming). Study on perceived change on credit-worthiness by financial or lending institutions of smallholder farmers availing comprehensive and up-to-date farm data sets including spatial data. 217 See https://ensibuuko.com/. 218 A major source for this section is USAID. 2018. ‘Data Driven Agriculture’, available at (https://www.usaid.gov/sites/ default/files/documents/15396/Data_Driven_Agriculture_Farmer_Profile.pdf). 219 The platform builds capacity throughout the CGIAR to generate and manage big data, assisting CGIAR and its partners’ efforts to comply with open access/open data principles to unlock important research and datasets. See https://bigdata.cgiar. org/about-the-platform/. 220 See https://theodi.org/topic/agriculture-and-food/. 221 http://www.data4sdgs.org/. 222 GODAN, launched in 2013, is a sector coalition that is working toward the aim of making agricultural and nutritional data more available, accessible, usable, and unrestricted worldwide. GODAN is the leading sector association on Data4Ag issues and has seen particularly accelerated growth in the past few years, from ~350-400 members in 2017 to 920+ in April of 2019. See https://www.godan.info/. 223 Powered by weather data from aWhere and many other data sources, the WB Ag Observatory is both an internal function/service for the World Bank Group and an outwardly facing tool and capacity-building entity for governments throughout Africa. The observatory has as its mission the focus on harnessing big data, artificial intelligence, and machine learning for productive and resilient agriculture worldwide through better agriculture sector decision-making. See the WBG Ag Observatory overview presentation available at: https://olc.worldbank.org/system/files/Harnessing%20Big%20 Data%2C%20Artificial%20Intelligence%20and%20Machine%20Learning%20for%20productive%20and%20resilient%20 agriculture.pdf. 224 See http://kaop.co.ke/. 225 http://fews.net/about-us. 226 http://geoglam.org/index.php/en/global-regional-systems-en/crop-monitor-for-amis. 227 See http://dataviz.vam.wfp.org/. 228 For more information about the CropWatch Mozambique tool, see https://www.itu.int/en/ITU-D/Regional-Presence/ AsiaPacific/SiteAssets/Pages/E-agriculture-Solutions-Forum-2018/CropWatch%20for%20ESF.pdf. 229 See https://gro-intelligence.com/about. 230 TCS (https://www.tcs.com) has developed an agricultural analytics engine called agEYE™, along with a web-based application that provides historic, current, and future data on crops. The application offers crop health, soil moisture, weather forecast, disease severity forecast, and disease identification at a village level to farmers and other stakeholders in the agri-value chain, including macro agri-decisionmakers. These parameters are derived from near real-time remote sensing data and weather data from third-party service. The service is primarily deployed in India, but has also seen some adoption in South Africa pilots. 231 See https://6grain.com/. 232 See https://www.mckinsey.com/solutions/acre. 233 See https://bigdata.cgiar.org/inspire/inspire-challenge-2018/cubica-the-new-farmer-advisory-app/. 234 https://www.satsure.co/. 235 https://satelligence.com/. 236 See https://www.microsoft.com/en-us/ai/ai-for-earth?activetab=pivot1%3aprimaryr6. 237 See https://earthengine.google.com/. 238 For the MercyCorp AgriFin definition of ‘Super Platforms’, see CGAP. 2018. ‘Super Platforms: Connecting Farmers to Markets in Africa’, available at https://www.cgap.org/blog/super-platforms-connecting-farmers-markets-africa). This CGAP blog post, and the underlying Dalberg & MercyCorps AFA ‘Digital marketplace benchmarking report’ it referred to, frame the ‘super platform’ concept more narrowly than this report. The digital marketplaces in question have all key features we have highlighted for super platforms, but all are commercial enterprises with e-commerce, or e-commerce combined with payments, at their core. While e-commerce, or rather buyer-seller digital marketplaces, needs to be a key component of super platforms, we believe that there are many more variants of such models including government- and donor-led platforms with digital marketplace components (e.g., SNS Rwanda, FtMA Rwanda) and bank-led models (e.g., KCB/ MobiGrow). 239 For holistic Service Delivery Models (SDM), see the forthcoming case studies on SDM models from MasterCard Foundation, IDH, and Dalberg at https://www.raflearning.org/post/the-business-case-smallholder-finance-introducing-the-sdm-case- study-series); for integrated digital marketplaces, see MercyCorp AFA & Dalberg. 2018. ‘Digital marketplace benchmarking report.’). 218 ENDNOTES 240 See emerging insights coming out of Dalberg, IDH, and MCF RAFLL studies of integrated market linkage models with Super Platform features (e.g., Digifarm, Tulaa) (https://www.raflearning.org/post/the-business-case-smallholder-finance- introducing-the-sdm-case-study-series). 241 While SNS was built and is being managed by the Bank of Kigali (BoK)/TecHouse, the system is governed jointly by BoK and the government of Rwanda via the Rwanda Agriculture Board (RAB). SNS already covers elements of advisory services (i.e., SMS-based advice and alerts to 1.4 million farmers) and financial access (i.e., B2P, G2P, C2C payment functionality, universal e-wallet, BoK savings accounts), market linkages (agro-dealer linkage as part of the subsidy programme), and supply chain management. The next steps in the system’s evolution include insurance product distribution, the provision of credit products via BoK, and an off-take market linkage virtual digital marketplace. See https://smartnkunganire.rw. 242 As part of the recently launched and BMGF-funded Digital Green advisory data ecosystem consortium in Ethiopia, ATA will be looking at opportunities to integrate or link major national assets including national digital advisory infrastructure (e.g., 80-28 hotline), digital payments and e-wallet for agriculture (e.g., potential partnership with Ethiotelecom), and perhaps market linkage initiatives. 243 See https://www.enam.gov.in/. 244 See https://ftma.org/; see also note 105 for details. 245 KCB, East Africa’s largest commercial bank, entered into a €27 million partnership with MasterCard Foundation in mid- 2018 to promote financial inclusion for at least 2 million smallholder farmers in Kenya and Rwanda. In addition, KCB group committed at the time to extending at least ~€180 million to farmers in the two countries in affordable loans over a five-year period. Today, the digital MobiGrow product already reaches 380,000 famers, with a plan to reach 1.5-2 million more in the next few years. ( https://ke.kcbgroup.com/business/agri/MobiGrow). 246 GSMA. 2016. ‘Market size and opportunity in digitising payments in agricultural value chains.’ (https://www.gsma.com/ mobilefordevelopment/resources/market-size-and-opportunity-in-digitising-payments-in-agricultural-value-chains/). 247 Digifarm has already registered 950,000 farmers by early 2019, though the number of clients using market linkages and receiving credit is still relatively low at this early stage of the product’s build-out. The platform is continuing to grow and evolve in terms of its reach and functionality. 248 See https://www.ecofarmer.co.zw/value-chain-services. 249 MFN (https://www.mastercard.us/en-us/about-mastercard/corp-responsibility/social-sustainability/the-mastercard-labs-for- financial-inclusion.html ) is a platform that digitises marketplaces, payments, workflows and farmer financial histories within the agriculture sector. MFN increases farmer linkages to markets and formal financial services relevant to their needs and aspirations. The platform brings together various agri-sector stakeholders, such as farmers, farmer producer organisations, buyers, financial institutions and value-added services providers, amplifying the collective positive impact on farming communities. 250 See, e.g., https://www.awhere.com/muiis-project-in-uganda-transitions-to-a-business-that-helps-farmers/. 251 See in-depth profile in Olam. 2019. Olam Insights, May 2019 (https://www.olamgroup.com/content/dam/olamgroup/ investor-relations/ir-library/olam-insights/olam-insights-pdfs/Olam_Insight2019_Issue1.pdf). 252 The Figure illustrating Taobao’s model and the related text below draw on several sources. See, e.g.,: World Bank. 2019. ‘E-commerce for poverty alleviation in China’ (http://blogs.worldbank.org/eastasiapacific/e-commerce-poverty-alleviation- rural-china-grassroots-development-public-private-partnerships); Su, Q. & Yan, D. 2018. ‘Rural Taobao yields benefits for farmers by analyzing big data’, China Daily Asia (http://epaper.chinadailyasia.com/asia-weekly/article-13996.html); Xinxua. 2017. ‘China’s prominent techfin shares rural poverty alleviation lessons with FAO.’ (http://www.xinhuanet. com//english/2017-07/15/c_136446325.htm); Ding, D. et al. 2017. ‘From Ant Financial to Alibaba’s Rural Taobao Strategy - How Fintech Is Transforming Social Inclusion.’ https://www.researchgate.net/publication/328274301_From_ Ant_Financial_to_Alibaba’s_Rural_Taobao_Strategy_-_How_Fintech_Is_Transforming_Social_Inclusion; 2017. Chen, J. 2017. ‘Ant Financial: Our rural china practice’. (https://www.slideshare.net/ExternalEvents/ant-financial-our-rural-finance- practice); Alibaba. 2016. ‘Rural Taobao Overview.’ https://www.alibabagroup.com/en/ir/pdf/160614/09.pdf . 253 Ibid. Chapter 3 254 FAO launched the e-agriculture community of practice in 2002, which we believe was one of the first formal conversations around D4Ag, following soon after the start of GSMA’s mAgri programme in late 2001. While our database does not go back far enough, GSMA’s no-longer functional ‘mAgri’ tracker documented fewer than 10 D4Ag solutions in Africa prior to 2005 and 66 D4Ag solutions by 2010 (not directly comparable to our database number). 255 This analysis is primarily based on self-reported survey data and desk research and is not exhaustive of all current and prior D4Ag activity in Africa. As a result, the figures across all years are likely understated – particularly the data for the earliest years, given that a significant (though uncounted) number of solutions have gone out of business. Using the number of solutions captured in the GSMA mAgri data tracker in 2012 to adjust for this survivorship bias yields a CAGR of 35% in terms of the number of solutions, rather than 45% calculated based on our database. While the figures are not exact, they help illustrate the likely growth trajectory of the sector over the last 7+ years. 256 We estimate that the database currently captures only 90-95% of the relevant solutions in the space given the difficulty of tracking very new start-ups. Approximately seventy D4Ag enterprises in our D4Ag database are now defunct, but there is a strong survivorship bias in the data. Comparison to earlier estimates by GSMA and others suggests there are likely another 50-100 defunct solutions that have not been reflected in our data. Most of these defunct organisations were part of the advisory services use case and were launched before 2015. 257 This includes solutions that were launched in the first few months of 2019 prior to the finalisation of this report. 258 Our database captured ~360 unique companies that offered these 390 solutions. Roughly 15 enterprises offered more than one solution, ranging from two up to 12 solutions (e.g., both Viamo’s ‘3-2-1’ services and Orange’s mAgri services comprise over 10 solutions in partnership with other organisations across the Sub-Saharan Africa region). 259 While data are spotty for these kinds of projects, directional estimates provided in interviews by major agriculture sector funders in Africa – such as BMGF, WB/IFC, USAID, GIZ, DFID and the EU – or implementers like Mercy Corps suggest ENDNOTES 219 that the number of donor-funded D4Ag projects or projects with D4Ag components is growing rapidly. For instance, a review of World Bank agriculture projects a few years ago concluded that ~80% had some sort of digital component (e.g., use of SMS for M&E) (WB interview (2019). 260 More so than other use cases, the financial access category presents quite a few definitional challenges in terms of where the border should be drawn between D4Ag financial access solutions and financial service providers and products that (i) are not sufficiently agricultural (i.e., are not tailored to the needs of smallholder farmers even if they happen to be used by smallholder farmers and (ii) are not sufficiently digitalised (e.g., traditional banks that have started to introduce digital channels for client communications). Our database, for instance, excludes digital payment solutions that are not specifically crafted for smallholder farmers (e.g., M-Pesa). Likewise, the database excludes banks and MFIs who have started to digitalise some of their operations but have not launched fully digital products, i.e., those that are not branded as being digital, or still require significant in-person interaction. 261 The macro agri-intelligence number appears artificially low in this analysis. There are many D4Ag solutions (60+) that have macro agro-intelligence components, but where macro agri-intelligence is just a secondary or ancillary revenue stream and not the primary focus of the enterprise and hence is not shown here. 262 The 44% figure over the past three years is the self-reported growth in farmer registrations among the Dalberg-CTA D4Ag survey respondents; the 55% CAGR over the past eight years is based on a roughly estimated 1 million farmers registered for D4Ag solutions in Africa in 2010–2011 based on desk research and the GSMA mAgri tracker. 263 Dalberg-CTA database analysis triangulated with interviews and desk research (see Methodology appendix). 264 There are an estimated 73 million smallholder farmer households (63 smallholder households plus 10 million pastoralists households) and 250 million total smallholder farmers (190 million smallholder farmers plus 60 million pastoralists) in Sub- Saharan Africa (See Lowder, S.K. et al. 2016. ‘The Number, Size, and Distribution of Farms, Smallholder Farms, and Family Farms Worldwide’. World Development (www.sciencedirect.com/science/article/pii/S0305750X15002703) and the Methodology appendix in this report). These values yield a penetration in the range of 13–45% for registered smallholder farmers depending on the denominator used (i.e., share of all farmers or share of smallholder farmer households). 265 Our database captured a very small number of agents as registered farmers due to reporting errors – but this number should be negligible and a rounding error, likely on the order of a maximum of a few thousand users. 266 In areas where many D4Ag solutions have expanded rapidly (e.g., Kenya), duplicate registrations could account for as much as 30–40% of the total registration count based on our comparisons of total estimated country level registrations vs. the share of farmers reporting the use (at any point) of D4Ag services. In locations with few D4Ag solutions, duplicated registrations likely account for fewer than 10% of the total registration count. Given the fact that Kenya is exceptional in its levels of D4Ag solution penetration and use, we assume that a maximum of 20% of farmer registrations were duplicates across the region, yielding an estimate of approximately 26 million unique farmers registered for D4Ag solutions in Sub-Saharan Africa. However, given our inability to estimate this number with confidence, we use the total number of registrations – 33 million – for the remainder of the report. 267 Please see Annex for a detailed methodology behind calculations for MNO, agribusiness, and FSP reach. 268 These include Econet, MTN, Orange, Airtel, and Safaricom/Vodafone. In addition to their own mAgri deployments, Orange and Vodafone have also launched mAgri solutions in partnership with other players (i.e., Orange has partnered with Brastorne Partners and Viamo; and Vodafone has partnered with Esoko in Ghana). 269 Interviews and desk research; see, e.g., Askew, K. 2018. ‘From ‘revolutionary’ tech to empowering farmers: How Olam leverages its African footprint to improve cocoa sustainability’. Food Navigator (www.foodnavigator.com/Article/2018/09/27/ How-Olam-leverages-its-African-footprint-to-improve-cocoa-sustainability). 270 We base this figure on publicly available data, survey data from the CTA-Dalberg survey, and expert assessments from interviews; as the information is not publicly revealed in some cases; there is a wide range of uncertainty around this number. 271 These numbers are particularly challenging to come by as most agribusinesses do not directly report the number of farmers reached by their digital offerings, and many have just announced plans to introduce D4Ag services to their farmers. 272 The definition of ‘engaged user’ includes users who were defined by the surveyed D4Ag enterprise as being ‘active’. The definition of active is subjective, but exceeds the use of the solution once per month during the crop season for advisory, market linkage, and supply chain management solutions. For financial services, this definition was less applicable – a farmer may only use the solution once but still be active or a customer in good standing – for instance, in the case of a digital savings account, digital credit product or agri-index insurance. 273 GSMA has conducted case studies of M-Kilimo in Kenya (2011), Airtel M’chikumbe 212 in Malawi (2017), Orange Senekela in Mali (2015), and Tigo Kilimo in Tanzania (2015). GSMA website (www.gsma.com/mobilefordevelopment/ resources/mfarmer-case-studies). 274 Note that these are not unique users, as some farmers may be served by more than one D4Ag enterprise. However, this ‘double counting’ is likely small at the moment and concentrated in a few countries with high D4Ag activity, such as Kenya and Nigeria. 275 These findings are based on our comprehensive review of D4Ag solutions in the region. It is possible, however, that the two countries without D4Ag enterprises (e.g. Seychelles, Sao Tome and Principe, Mauritania, Equatorial Guinee, Eritrea, Djibouti) do have some presence of D4Ag solutions that we were not able to uncover during our study. 276 Taking a broader view, of the 390 solutions in our database, excluding non-revenue seeking MNO and agribusinesses solutions, commercial enterprises stood behind 74% of all solutions and an unknown share of the 15% of solutions that were backed by NGOs did have some earned revenues, so the number of revenue-seeking solutions in the broader D4Ag sector is likely over 80%. 277 We defined companies with a profitable and stable business model as those that claimed that their costs were less than 90% of their operating budgets and revenues were more than 90%. Please note that many enterprises claimed that their costs and revenues were both less than 90% of their operating budgets, in which case we could not determine their profitability and did not include them in this count. 278 For instance, AGRA examined the economics of 15 African D4Ag enterprises in depth and found that only a third had sustainable economics in the absence of substantial ongoing donor support. AGRA. 2016. ‘Digital Harvest’. Enterprises in the AGRA sample were, however, more established than the average D4Ag solution in our survey. 279 Looking at this data in the broader start-up context, studies of new business starts in the US and Europe and oft-cited benchmarks from the tech VC industry suggest that 2-3 years are required, on average, for companies to reach profitability, 220 ENDNOTES with most companies starting to break even at some point in the second year and reaching steady profitability in the third. See, e.g., US Small Business Administration data on business starts in the US (www.sba.gov); see also Mansfield, M. 2019. ‘Startup statistics: The numbers you need to know’. 280 Our interviews with VC experts for Africa suggest that a 20-30% share of profitable and/or sustainable enterprises is to be expected in a highly social sector of this type in Africa. 281 This rough projection applies the 26% share of profitable enterprises to the 289 commercial D4Ag solutions in the database, and then assumes a 50-75% failure rate for non-profitable solutions and 20-30% failure rate for profitable solutions over the next 3 years, not counting new business entry which is likely to be substantial and will feature some firm that break even early. 282 See note 284 for details based on self-reported revenue/user/year for solutions in our database which draws on both survey data and interviews with leading D4Ag solution providers. The ranges of self-reported per-farmer revenues are wide because each covers a broad variety of underlying business models. For instance, D4Ag market linkage players that use virtual buyer- seller marketplaces tend to earn a very small fee for matching supply with demand, often no more than a few Euros of value for the transaction. Whereas digitally-enabled value chain integrator types of market linkage models, such as those from Twiga, iProcure or Tulaa, can earn 10-20x this amount for their value chain intermediation services. For financial services, please note that these numbers do not include interest income as the focus of these benchmarks in our data was on D4Ag credit, insurance, and payment intermediaries rather than traditional FSPs who have digitised their value proposition. 283 The number of addressable farmers is likely different for each use case. For advisory services, we believe that it is possible for multiple members of a smallholder farmer household to be a user. As such, we have defined the addressable market as the total number of estimated smallholder farmers and pastoralists in Africa, i.e., ~250 million. For the other use cases, it is likely that these services are used at the level of the household (e.g., only one market linkage application or insurance or credit per smallholder farm), and as such we use 73 million as the total number of addressable farmers for these use cases. This is without a doubt a radical simplification of a complex reality, but likely does provide a directional sense for the market’s size. 284 The range for total addressable market is quite high given the wide range of ARPU within and across use cases. We estimate that ARPU for advisory services is in the order of €0.90-8.90 per farmer/year, financial access services is €0.40–€6.70 per farmer/year, market linkage between €2.70–€50 per farmer/year, and supply chain management to be €0.60–€8.90 per farmer/year. For financial access it is typical for payments, credit, and insurance products to be combined, so actual revenue for a fully integrated financial access player would be in the €3-14 range. These figures are based on self-reported figures from survey respondents and figures shared with us during interviews with implementers. 285 We assumed access to a mobile phone as an important constraint to smallholder farmer ability to use digital solutions. Access to a mobile phone could mean multiple things, however, so we looked at this figure in multiple ways. The most restrictive way to look at this figure is to assume that only individual unique mobile subscribers have access to a mobile phone. For this, we assumed a minimum of 39% of smallholder farmers (using GSMA’s 2018 estimate of unique subscribers across Africa and applying a 1.3:1 urban to rural access ratio to account for the fact that rural penetration is lower than urban penetration). A second way of looking at this figure is to look at phone ownership data for individual smallholder farmer households. We estimate this to be 50-60% based on phone ownership to unique subscriber ratios from select countries (e.g., Nigeria) and smallholder household surveys. A third way of understanding household access to a mobile phone is to use smallholder farmer household ownership of a phone. Based on CGAP smallholder farmer level data from a handful of countries across Africa (and ratios of household phone ownership to unique subs and any phone ownership), we estimate this figure to be ~70% across the Sub-Saharan Africa region today. It is possible that smallholder farmers could theoretically access phones that are not in their household to use D4Ag solutions, but we did not include that here as we do not have reliable estimates, and it is likely that farmers need reliable, regular access to a phone to use solutions, which is much harder if the phone is not within the household. Therefore, for the purposes of our TAM calculation, we used a range of 39-70% to represent the likely minimum and maximum levels of connectivity among smallholder farmers. Another potential connectivity constraint is rural signal coverage, which we estimate at 70%+ in Sub-Sahara Africa today, so comparable to the household penetration of phones figure. 286 This estimate is calculated from known (self-reported) revenues of ~ €107 million from 76 enterprises in our database. To this we added estimated revenues for enterprises whose revenues were not already known. Where we knew the user base, but not revenues, we used average revenue per user estimates by primary use case. For solutions where we did not know the user figures, we applied the average user figure for deployments (removing big outliers) to estimate the number of users, along with the same average revenue per user (ARPU) estimates. The numbers are a conservative estimate. For instance, revenues of D4Ag data intermediaries (e.g., data analytics players, drone companies with agriculture projects) are not estimated with the exception of those that reveal this information publicly and where it is possible to identify the agriculture specific revenue streams. In the case of financial access solutions, we focus on digital intermediation revenues and product fees (e.g., farmer credit scoring revenues) but not interest income on farmer loans. 287 Penetration of addressable market derived by dividing the mid-point value for sector earned revenues (~€127 million) by the average TAM in conservative (€1.6 billion) and less conservative (€2.9 billion) scenarios, which yields a penetration of 4-8%, or 6% on average. 288 There are insufficient data to make impact comparisons for the other D4Ag use cases. 289 Chatterjee, S. 2017. ‘Promise Or Peril? Africa’s 830 Million Young People By 2050’. UNDP (www.africa.undp.org/content/ rba/en/home/blog/2017/8/12/Promise-Or-Peril-Africa-s-830-Million-Young-People-By-2050.html). 290 ICT Update. 2016. Youth E-agriculture Entrepreneurship (www.cgspace.cgiar.org/bitstream/handle/10568/89782/ICT083E_ PDF.pdf?sequence=2&isAllowed=y). 291 Sudarkasa, M. and Odetayo, W. 2019. ‘Accelerating youth agri-incubators in Africa – Lessons from Nigeria’. CTA (www. cta.int/en/blog/all/article/accelerating-youth-agri-incubators-in-africa-lessons-from-nigeria-sid0a3ff95df-679d-4509-8fb4- 2236985a138a). 292 CGIAR. ‘Youth Involvement in Agribusiness: Examples from Africa’. Undated blog post (www.ccafs.cgiar.org/blog/youth- involvement-agribusiness-examples-africa). 293 Mulligan, G. 2019. ‘5 Agri-tech Start-ups Join Senegal Startup Accelerator‘. Disrupt Africa (www.disrupt-africa. com/2019/01/5-agritech-startups-join-senegal-startup-accelerator). 294 Halperin, M. ‘Information and communication innovations in East Africa’. CGIAR, undated blog post (www.ccafs.cgiar.org/ news/information-and-communication-innovations-east-africa); Kamau, K. 2018. ‘Partnerships to Increase Open Weather Data’s Impact’. ICT Update (www.ictupdate.cta.int/2018/05/24/partnerships-to-increase-open-weather-datas-impact). In particular, changes in rainfall will have severe implications for the 90% of agriculture in Africa that is rain-fed. ENDNOTES 221 295 Tricarico, D. and Darabian, N. 2016. ‘Weather forecasting and monitoring: Mobile solutions for climate resilience’. GSMA (www.gsma.com/mobilefordevelopment/wp-content/uploads/2016/02/Weather-forecasting-and-monitoring-mobile-solutions- for-climate-resilience.pdf). 296 CGIAR. 2013. ‘Climate-smart Villages – a community approach to sustainable agricultural development’ (www.cgspace. cgiar.org/bitstream/handle/10568/33322/CCAFSClimate-SmartVillages2013.pdf). 297 Creech, H. et al. 2014. ‘ICTs for Climate Change adaptation in Africa’. World Bank (www.documents.worldbank.org/ curated/en/651751468003576392/pdf/882250WP0Box380limateChange0summary.pdf). 298 Esoko website (www.esoko.com/mobile-phones-help-northern-ghanas-farming-families-beat-climate-change/). 299 Index Insurance Forum. 2013. ‘ACRE/Syngenta Foundation for Sustainable Agriculture - Kenya, Rwanda, Tanzania’ (https://www.indexinsuranceforum.org/project/acresyngenta-foundation-sustainable-agriculture-kenya-rwanda-tanzania). 300 Ibid. Advisory services and financial access are not the only use cases that can support climate resilience. For example, market linkage solutions could provide farmers access to new types of fertilisers with more or less nitrogen as soil contents alter in response to changing climate. See CGIAR. 2013. ‘Climate-smart Villages – a community approach to sustainable agricultural development’ (www.cgspace.cgiar.org/bitstream/handle/10568/33322/CCAFSClimate-SmartVillages2013.pdf). 301 CTA. 2018. ‘20,000 Ethiopian Smallholders Targeted with Climate Smart Technology’ (www.cta.int/en/climate/all/ article/20-000-ethiopian-smallholders-targeted-with-climate-smart-technology-sid0c0262306-ffa7-403f-a621-9cb426e0e6f6). 302 Manning, L. 2018. ‘Ignitia Raises $1.1m Series A to Expand Tropical Weather Forecasting Service to Smallscale Farmers in Nigeria’. AgFunder News (www.agfundernews.com/ignitia-raises-1-1m-series-a-to-expand-tropical-weather-forecasting-service- to-smallscale-farmers-in-nigeria.html). 303 Tricarico, D. and Darabian, N. 2016. ‘Weather forecasting and monitoring: Mobile solutions for climate resilience’. GSMA (www.gsma.com/mobilefordevelopment/wp-content/uploads/2016/02/Weather-forecasting-and-monitoring-mobile-solutions- for-climate-resilience.pdf). 304 ILO. 2016. ‘Addressing Gender Gaps in Africa’s Labour Market’ (www.ilo.org/africa/media-centre/pr/WCMS_458102/ lang--en/index.htm). 305 GSMA. 2019. ‘Connected Women – The Global Gender Gap Report 2019’ (www.gsma.com/mobilefordevelopment/wp- content/uploads/2019/03/GSMA-Connected-Women-The-Mobile-Gender-Gap-Report-2019.pdf). 306 Ibid. 307 Technoserve. ‘Brewing Prosperity in East Africa: Coffee Initiative Final Report’. Undated report (https://www.technoserve. org/files/downloads/Coffee-Initiative-Final-Report.pdf). 308 MyAgro website (www.myagro.org). 309 Meinzen-Dick et al. 2011. ‘Engendering agricultural research, development, and extension’. IFPRI (www.ebrary.ifpri.org/ utils/getfile/collection/p15738coll2/id/126799/filename/126995.pdf). 310 Seeds and fertiliser packets given to the women are provided by Bayer/Monsanto, which is funding the project. Chapter 4 311 As explained in more detail later in Chapter 4, the high end of the range in this estimate is based on a 44% annual growth rate for registred farmers and revenues (based on the last 3 year trend). The low end of the range, shows a growth rate that is half this rate (22%), which is what would happen if the number of farmers added over next 3 years (in absolute terms) was identifical to the number added inthe past 3 years (i.e., same pace of farmer acquisition due to challenges in moving into new and more difficult markets).” 312 For examples of other sector experts who share this perspective, see, e.g., Syngenta Foundation for Sustainable Agriculture (SFSA). 2019. ‘How can digital solutions help to feed a growing world?’ (https://www.syngentafoundation.org/agriservices/ whatwedo/digitalsolutions); see also Schrader, L. et al. 2018. ‘Super Platforms: Connecting Farmers to Markets in Africa’ (https://www.cgap.org/blog/super-platforms-connecting-farmers-markets-africa). 313 Chapter 2 and 3 delved into some of the challenges of ‘pure-play’ information and advisory service D4Ag business models. For more on these issues for early generation advisory solutions like Esoko, see e.g., Miller-Wise, H. 2017. ‘Why we broke up the company: a former CEO of mAgri pioneer Esoko speaks.’ Next Billion (https://nextbillion.net/why-we-broke-up-the- company-a-former-ceo-of-m-agri-pioneer-esoko-explains/). See also the related and highly informative sector actor discussion thread in https://www.ictworks.org/esoko-agricultural-market-prices-failures/#.XPPKFogzY2x. 314 See, e.g., Syngenta Foundation’s perspective that “a broad, holistic approach needs to be taken to drive greater value and impact for smallholder farmers” while at the same time improving the commercial viability of D4Ag models. SFSA (2019). 315 See impact discussion of bundled solutions in Chapter 3. 316 Analysis of the CTA-Dalberg database suggests that at least 25% and perhaps as many as 35% of D4Ag advisory and market linkage solutions, for instance, feature agents as part of the model directly or indirectly. 317 SFSA 2019. 318 For more formal value chains, D4Ag supply chain management solutions, for instance, are already inherently agent-based models as they help agribusiness interact with the agents and farmers in their value chains. 319 Our interviews suggest, for instance, that the incremental costs of integrating agents are likely to sit in the range of an incremental €0.3-1.8 per farmer per month (assuming fixed agent compensation of €90-270 per month, a realistic range for much of Africa, and agent to farmer ratios of 1:150 to 1:300). This sum of €7-21 incremental costs per farmer per year is a meaningful amount for most African smallholders, but is not prohibitive if the increased costs can pay for themselves through improved D4Ag solution impacts. The threshold for this to break-even is not high, using average Sub-Sahara Africa farm monthly household income ranges of €50-250, these numbers would mean that a farmer would need to see only a 0.5- 3% improvement in incomes (or reduction in costs) to justify the economics of the agent model. 320 See the Taobao case study at the end of Chapter 2. 321 See, e.g., Better Than Cash Alliance. 2017. ‘The Role of Digital Payments in Sustainable Agriculture and Food Security’. 222 ENDNOTES 322 GSMA estimated the total available for farmer B2P payment digitalisation at €57 billion ($64 billion) in Sub-Saharan Africa in 2020 and assumed 50 basis points (0.5%) revenues for digital payment providers, €280 million. GSMA. 2016. ‘Market size and opportunity in digitising payments in agricultural value chains.’ This number does not map neatly to Dalberg’s total addressable market sizing for D4Ag due to methodological differences, but we likewise estimate the agriculture payment digitalisation opportunity in Africa to be in the hundreds of millions (>€500 millions) in 2019, though much of this revenue potential is only available for digital payment backbone providers rather than D4Ag intermediaries linking farmers to payments (i.e., only taking a share of the 50 basis points of revenue for their services). 323 Ibid. 324 Better Than Cash Alliance. 2017. ‘How Digitizing Agricultural Input Payments in Rural Kenya Is Tackling Poverty: The Case of One Acre Fund’. 325 It is also important to note the underlying location-based technologies that enable many of these advanced technologies to function, and how they have improved in recent years. 326 Twenty-eight per cent of respondents answered IoT, 32% answered blockchain, and 20% answered machine learning. 327 This discussion draws heavily on USAID’s recent report on agriculture data for smallholder farmers. See USAID. 2018. ‘Digital Farmer Profiles: Reimagining Smallholder Agriculture’ (www.usaid.gov/sites/default/files/documents/15396/Data_ Driven_Agriculture_Farmer_Profile.pdf). 328 Ibid. 329 Ibid. 330 Climate Focus and USAID. 2018. ‘Policy Brief – ICT Solutions for Inclusive Agricultural Value Chains’ (www.agrilinks.org/ sites/default/files/brief_5_-_ict_solutions_for_agricultural_value_chains.pdf). 331 USAID, 2018. ‘Digital Farmer Profiles: Reimagining Smallholder Agriculture’ (www.usaid.gov/sites/default/files/ documents/15396/Data_Driven_Agriculture_Farmer_Profile.pdf). 332 Ibid. 333 CTA is undertaking a number of projects in Africa (including in Tanzania, Ghana, Swaziland and Benin) to survey, identify and demark plots of land. The Omidyar Network is similarly working in Ghana to help build capacity within the government to use drones for land tenure adjudication. A number of D4Ag enterprises have also deployed innovative solutions for these purposes on the ground. Meridia in Ghana (see https://www.meridia.land/) is one of the notable examples in this space. 334 Ibid. 335 AGRA. 2017. ‘Africa Agriculture Status Report’ (www.agra.org/wp-content/uploads/2018/05/Final-AASR-2017-Aug-282. pdf). 336 Ibid. 337 Orange. 2017. ‘IoT in Africa, let’s go!’ (www.orange.com/en/news/2017/Decembre/IoT-in-Africa-let-s-go). 338 Juergen V. 2018. ‘Big Data Shows Big Promise Feeding the World’. IFPRI (www.ifpri.org/blog/big-data-shows-big-promise- feeding-world). 339 Neethirajan, S. and Jayas, D.S. 2007. ‘Sensors for Grain Storage’. ASABE. 340 Sol Chip website (www.sol-chip.com/applications_livestock.asp). 341 Ujuzikilimo website (https://www.ujuzikilimo.com/). 342 CTA. 2018. ‘Is the Internet of Things the future of farming?’ (www.cta.int/en/article/is-the-internet-of-things-the-future-of- farming-sid0f442c544-a1bc-4445-af71-f78e25403a36). 343 CGIAR Platform for Big Data website (www.bigdata.cgiar.org/about-the-platform/). 344 Tollefson, J. 2018. ‘Big-data project aims to transform farming in world’s poorest countries’. Nature International journal of science (www.nature.com/articles/d41586-018-06800-8). 345 CGIAR Platform for Big Data website (www.bigdata.cgiar.org/about-the-platform/). 346 Protopop, I. and Shanoyan, A. 2016. ‘Big Data and Smallholder Farmers: Big Data Applications in the Agri-Food Supply Chain in Developing Countries’. The International Food and Agribusiness Management Review (www.ifama.org/resources/ Documents/v19ia/920150139.pdf). 347 McDade, M., 2018. ‘10 ways CGIAR is opening up data for agricultural innovation’. CGIAR (www.bigdata.cgiar.org/10- ways-cgiar-is-opening-up-data-for-agricultural-innovation/). 348 Parker, S. 2018. ‘Machine-learning app to fight invasive crop pest in Africa’. Mongabay (www.news.mongabay. com/2018/10/machine-learning-app-to-fight-invasive-crop-pest-in-africa/). 349 For more details on Plantix (https://plantix.net), see, e.g., Schiller, B. 2017. ‘Machine Learning Helps Small Farmers Identify Plant Pests And Diseases’, Fast Company (https://www.fastcompany.com/40468146/machine-learning-helps-small- farmers-identify-plant-pests-and-diseases). 350 Ibid.; USAID, 2018. ‘Digital Farmer Profiles: Reimagining Smallholder Agriculture’ (www.usaid.gov/sites/default/files/ documents/15396/Data_Driven_Agriculture_Farmer_Profile.pdf). 351 Climate Focus and USAID. 2018. ‘Policy Brief – ICT Solutions for Inclusive Agricultural Value Chains’ (www.agrilinks.org/ sites/default/files/brief_5_-_ict_solutions_for_agricultural_value_chains.pdf). 352 Haider, I. 2018. ‘How blockchain can help smallholder farmers’. FAO (http://www.fao.org/e-agriculture/blog/how- blockchain-can-help-smallholder-farmers). 353 Climate Focus and USAID. 2018. ‘Policy Brief – ICT Solutions for Inclusive Agricultural Value Chains’ (www.agrilinks.org/ sites/default/files/brief_5_-_ict_solutions_for_agricultural_value_chains.pdf). 354 Bolt, J. 2019. ‘Agri-wallet, a wallet for smallholder farmers’. CTA (www.cta.int/en/digitalisation/all/article/agri-wallet-a- wallet-for-smallholder-farmers-sid00f60f624-f62a-4b58-bd27-bd2c838b724f). ENDNOTES 223 355 Kriticos, S. 2019. ‘Keeping it clean: Can blockchain change the nature of land registry in developing countries?’ World Bank (www.blogs.worldbank.org/developmenttalk/keeping-it-clean-can-blockchain-change-nature-land-registry-developing- countries). 356 Climate Focus and USAID. 2018. ‘Policy Brief – ICT Solutions for Inclusive Agricultural Value Chains’ (www.agrilinks.org/ sites/default/files/brief_5_-_ict_solutions_for_agricultural_value_chains.pdf). 357 Ibid. 358 As a directional order of magnitude, the ~390 D4Ag players focused on African smallholder agriculture are already absorbing a significant share of the ~€280 million of all grant, debt, and equity funding annually that was going into African D4Ag ecosystems (see analysis later in this chapter). The precise share of this amount that went to individual enterprises is unknown, as we do not know the volume of donors’ public good D4Ag investments not specific to any individual enterprise. We also do not know what share of required D4Ag enterprise funding these investments satisfied, but we can project that if such funding was to grow at the pace of new solution starts or at the historical pace of farmer registrations (44-45% across both variables), the funding need would quickly approach €1 billion within 3 years or at the latest within 5 years if future sector growth was half of what had seen historically. 359 As a rough thought experiment, we have sized the amount of D4Ag public good investment into Ethiopia over the past few years including the EthioSIS soil information system investments, the livestock surveillance and pest and disease surveillance systems being put in place with the Bill and Melinda Gates Foundation’s help, the government (ATA’s) investment into farmer registration and the setup of the 80-28 national advisory infrastructure, and upcoming investments into digital finance, market linkage, and advisory platforms for the sector. The total amount of such investments across donors and the government totalled well over €50 million. Scaling this amount from Ethiopia to the entire Sub-Saharan Africa population suggests that the required level of D4Ag infrastructure investment is at least €500 million (scaled based on population size) and potentially as €1.2-2 billion (scaled assuming a fixed cost of €20-50 million per country). This number also does not include some critically important underlying infrastructure that is relevant for the agriculture sector, but is not agriculture-sector specific. The most notable example is weather (hydromet) infrastructure, where anecdotal interviews with Africa weather infrastructure experts suggest that the funding gap for Africa is in the high hundreds of millions or low billions of Euros. For instance, the World Bank has estimated a few years ago that hydro-meteorological service modernisation for just 15 countries in Africa will require $600 (€534 million), perhaps €1.5+ billion if extrapolated to the continent overall. Of this amount, in 2018, the World Bank was tracking ~$900 million (€800 million) of active or pipeline hydromet investment for Sub-Saharan Africa, a gap of many hundreds of millions of Euros. https://www.gfdrr.org/ en/africa-hydromet-program/about. 360 Assuming, as a thought experiment that half to a third of donor funds in 2019 were allocated to D4Ag infrastructure (€53- 80 million annually), this would mean a funding stream of €250-400 over 5 years compared to the estimated D4Ag funding need of 1.2-2 billion (excluding hydromet services), a resulting gap of perhaps 1-1.5 billon. These numbers are, of course, pure conjecture, but do point to the relative magnitudes of the sums involved. 361 Among others, this includes the African Development Bank, the Bill and Melinda Gates Foundation, the Department for International Development (DFID), the Dutch Foreign Ministry, Mastercard Foundation, Syngenta Foundation, USAID, and the World Bank. 362 The estimate is based on a proprietary databased of equity and debt transactions for D4Ag solutions that are either headquartered in Africa or headquartered in other geographies but have Sub-Saharan African smallholder farmers as their primary focus. 363 Disrupt Africa. 2019. ‘African Tech Startups Funding Report 2018’ (www.disrupt-africa.com/funding-report/). 364 Finistere Ventures. 2018. ‘2018 Agtech Investment Review’. Pitchbook (www.files.pitchbook.com/website/files/pdf/Finistere_ Ventures_2018_Agtech_Investment_Review_xeO.pdf). 365 IBM. 2019. ‘Yara and IBM join forces to transform the future of farming’ (www.newsroom.ibm.com/2019-04-26-Yara-and- IBM-join-forces-to-transform-the-future-of-farming). 366 McLelland, J. 2015. ‘Top 5 tech innovations in agriculture’. Raconteur (www.raconteur.net/sustainability/top-5-tech- innovations-in-agriculture); ReliefWeb. 2015. ‘Detailed soil information for Africa now available’ (www.reliefweb.int/report/ world/detailed-soil-information-africa-now-available). 367 FAO. ‘Google and FAO partner to make remote sensing data more efficient and accessible’. Undated blog post (www.fao. org/news/story/en/item/350761/icode/); UN. 2016. ‘FAO, Google collaborate on satellite data tools to manage natural resources’ (www.un.org/sustainabledevelopment/blog/2016/04/fao-google-collaborate-on-satellite-data-tools-to-manage- natural-resources/); Sunga, I. 2017. ‘These 5 innovations will transform the lives of smallholder farmers’. WEF (www. weforum.org/agenda/2017/01/these-5-innovations-will-transform-the-lives-of-smallholder-farmers/). 368 Bosch Africa website (www.bosch.africa/news-and-stories/real-change-real-people/). 369 Bosch. 2018. ‘Agriculture in the future: In dialogue with the Global Head Digital Farming of Bayer AG’ (www.bosch.com/ stories/thought-leader-tobias-menne/). 370 TCS. 2018. ‘Corporate Sustainability Report 2017-18’ (www.tcs.com/content/dam/tcs/pdf/discover-tcs/investor-relations/ corporate-sustainability/GRI-Sustainability-Report-2017-2018.pdf). 371 G4AW website (www.g4aw.spaceoffice.nl/en/projects/international/data-and-services/mobile/mkrishi/). 372 Hsu, J.W. 2018. ‘Rwanda first in Africa to join Alibaba-led EWTP’. Alizila (www.alizila.com/rwanda-first-in-africa-to-join- alibaba-led-ewtp/). 373 GSMA. 2018. ‘The Mobile Economy Sub-Saharan Africa 2018’ https://www.gsmaintelligence.com/research/?file=b9a6e620 2ee1d5f787cfebb95d3639c5&download. 374 GSMA traditionally suggests a 1.3 ratio of urban to rural unique subscribers, which given the 60% urban/40% rural mix of Sub-Saharan Africa in 2018, implies a 39% rural unique subscriber penetration. 375 Anderson, J. and Sobol, D. 2018. ‘CGAP National Surveys of Smallholder Households’. CGAP (www.cgap.org/sites/ default/files/publications/Executive-Summary-CGAP-National-Surveys-of-Smallholder-Households-Nov-2018_1.pdf). 376 GSMA estimates, for instance, that 66% of mobile subscriptions will be via a smartphone in Sub-Saharan Africa by 2025, up from 36% in 2018, which with linear growth in adoption would suggest 85%+ smartphone penetration by 2030. See https://www.gsmaintelligence.com/research/?file=b9a6e6202ee1d5f787cfebb95d3639c5&download. 224 ENDNOTES 377 GSMA. 2018. ‘The Mobile Economy Sub-Saharan Africa 2018’ (www.gsmaintelligence.com/research/?file=809c442550e5 487f3b1d025fdc70e23b&download); Bayen, M. 2018. ‘Africa: a look at the 442 active tech hubs of the continent’. GSMA (https://www.gsma.com/mobilefordevelopment/blog-2/africa-a-look-at-the-442-active-tech-hubs-of-the-continent/). Chapter 5 378 Hall, M. 2017. ‘Close Skills Gaps to Prepare Africa’s Workforce for Tomorrow’s Jobs’. WEF (www.weforum.org/ press/2017/05/close-skills-gaps-to-prepare-africa-s-workforce-for-tomorrow-s-jobs/). 379 GIZ website (www.giz.de/en/worldwide/57293.html). 380 Regulatory environment was measured based on strength of laws, regulations and policies that promote the provision and use of ICT services using the World Bank’s 2017 Enabling the Business of Agriculture (EBA) ICT indicator. Ability to adopt and use mobile internet was measured based on the infrastructure, affordability, consumer readiness, content and services needed to use mobile internet using the 2017 Mobile Connectivity Index (MCI) by GSMA. 381 Malabo Montpellier Panel. 2019. ‘Smart Value Chains: Policy Innovations for the Digitalisation of African Agriculture.’ https://www.mamopanel.org/resources/reports-and-briefings/. Chapter 6 There are no endnotes for this chapter. Annex 1 – Country Case Studies 382 Findings in this section come primarily from stakeholder interviews, the majority of which were conducted in a two-week research trip to Addis Ababa, Ethiopia. 383 Rounded to the nearest million and based on the number of users of Ethiopia-headquartered solution providers. Because all of these firms operate only in Ethiopia and no foreign providers are present, this also describes the number of users in the country. This figure does not adjust for users who may use more than one solution. 384 Data is based on 2 of 4 solutions for which gender breakdown of users was available. 385 World Bank. 2018. ‘Ethiopia Agricultural Growth Program’. Programme documentation (www.documents.worldbank.org/ curated/en/339541521833858063/pdf/Ethiopia-ET-Agricultural-Growth-Program.pdf). 386 Ethiopian Agricultural Transformation Agency (ATA) website (www.ata.gov.et/programs/highlighted-deliverables/8028- farmer-hotline/). 387 Because mobile money is controlled by the state, Ethiopia has been able to scale the number of users registered with the service more rapidly. However, adoption (that is, the proportion of registered users actually using these services) remains low. 388 Rounded to the nearest hundred thousand and based on the number of users of Ghana-headquartered solution providers, rather than on the number of any users in the country. This figure does not adjust for users who may use more than one solution. 389 Data is based on solutions for which gender breakdown of users was available. 390 Ministry of Food and Agriculture, Ghana. 2017. ‘Agriculture Sector Progress Report’ (www.mofa.gov.gh/site/wp-content/ uploads/2018/09/MoFA%202017%20AGRICULTURAL%20PROGRESS%20REPORT_Final.PPMED.MoFA.pdf). 391 FAO website (http://www.fao.org/ghana/fao-in-ghana/ghana-at-a-glance/en/). 392 Ibid. 393 Ministry of Food and Agriculture, Ghana. 2019. ‘Planting for Food and Jobs (PFJ) Campaign for 2019 Launched’ (www. mofa.gov.gh/site/wp-content/uploads/2019/01/Planting%20for%20Food%20and%20Jobs%20(PFJ)%20Campaign%20 for%202019%20Launched.pdf). 394 Rounded to the nearest hundred thousand and based on the number of users of Nigeria-headquartered solution providers, rather than number of users of any D4Ag solutions in country. This figure does not adjust for users who may use more than one solution. 395 Data is based on solutions for which gender breakdown of users was available. 396 World Bank. 2015. ‘More, and More Productive, Jobs for Nigeria: A Profile of Work and Workers’ (www.documents. worldbank.org/curated/en/650371467987906739/pdf/103937-WP-P146872-PUBLIC-Nigeria-Jobs-Report.pdf). 397 Mgbenka, R.N. et al. 2016. ‘A review of smallholder farming in Nigeria: need for transformation’. International Journal of Agricultural Extension and Rural Development Studies. 398 Olurounbi, R. 2018. ‘Nigeria Seen as Biggest Rice Buyer in 2019, Behind China’. Bloomberg (www.bloomberg.com/news/ articles/2018-11-14/usda-sees-nigeria-rice-imports-increasing-to-3-4m-tons-in-2019). 399 Private equity investors like CardinalStone and Sahel Capital are mobilising investments into agricultural processing projects like Crest Agro to leverage digital tools for the aggregation and collection of information from farmers supplying the processing plants. Sterling Bank, in collaboration with AFEX and Binkabi, has committed up to 10 billion Naira (~€24 million) to a blockchain-supported agricultural commodity trading platform. Around ten venture capital firms and incubators are exploring the potential of AgTech investments, with companies like Venture Garden Group looking to open a new €17.8 million fund in 2019 partially focusing on AgTech. Some of Nigeria’s largest agribusinesses have also started to direct their attention towards digitalisation. Dangote is developing digital solutions to optimise the company’s internal processes, collect and manage detailed data about their suppliers, and streamline their domestic supply chains of rice and sugar. Companies like Flourmill and Indorama are set to do the same. ENDNOTES 225 400 Farmers need to reach a certain level of development (in terms of irrigation systems, last mile transportation, soil quality, input availability, etc.) before digital solutions can provide a viable way to increase their yields and incomes. 401 WEF. 2018. ‘The Global Competitiveness Report 2018’ (http://www3.weforum.org/docs/GCR2018/05FullReport/ TheGlobalCompetitivenessReport2018.pdf). 402 Rounded to the nearest hundred thousand and based on the number of users of Senegal-headquartered solution providers, rather than number of users of any D4Ag solutions in country. This figure does not adjust for users who may use more than one solution. 403 Data is based on solutions for which gender breakdown of users was available. 404 World Bank website (www.data.worldbank.org/indicator/nv.agr.totl.zs). 405 Source: USAID. 406 Stads, G. and Sene, L. 2011. ‘Private-Sector Agricultural Research and Innovation in Senegal’. IFPRI, Rutgers University, and McGill University (www.asti.cgiar.org/pdf/private-sector/Senegal-PS-Report.pdf). 407 Source: USAID: https://www.usaid.gov/senegal/agriculture-and-food-security 408 Rounded to the nearest hundred thousand and based on the number of users of Kenya-headquartered solution providers, rather than any users in country. This figure does not adjust for users who may use more than one solution. 409 Data is based on solutions for which gender breakdown of users was available. 410 USAID website (https://www.usaid.gov/kenya/agriculture-and-food-security). 411 World Bank Group. 2019. ‘Doing Business 2019 – Training for Reform: Economy Profile – Kenya’ (www.doingbusiness. org/content/dam/doingBusiness/country/k/kenya/KEN.pdf).# 412 Georgetown University Initiative on Innovation, Development and Evaluation website (www.gui2de.georgetown.edu/ projects/DigiFarm). 413 Gichamba, A. et al. 2017. ‘An Assessment of e-Extension Platforms in Kenya’. International Journal of Innovative Studies in Sciences and Engineering Technology (www.ijisset.org/wp-content/uploads/2017/08/IJISSET-030713.pdf). 414 AgriFin website (www.agrifinfacility.org/about-us). 415 Privacy International website (www.privacyinternational.org/state-privacy/1005/state-privacy-kenya#dataprotection). 416 Findings in this section come primarily from stakeholder interviews, the majority of which were conducted in two-week research trip to Kigali, Rwanda. 417 Rounded to the nearest hundred thousand and based on the number of users of Nigeria-headquartered solution providers, rather than number of users of any D4Ag solutions in country. This figure does not adjust for users who may use more than one solution. 418 World Bank. 2017. ‘Agriculture, forestry, and fishing, value added (% of GDP)’ (www.data.worldbank.org/indicator/ NV.AGR.TOTL.ZS). 419 USAID and The World Economic Forum. 420 USAID and The World Economic Forum. 421 AfDB website (www.afdb.org/en/countries/east-africa/rwanda/rwanda-economic-outlook/). 422 It is important to note that these are centralised access points in each district, but do not necessarily provide everyone in each district with easy access to network connections. 423 Fripp, C. 2013. ‘ICT – all that is between Rwanda and value-added agriculture’. ICT4Ag (www.ict4ag.org/en/media- corner/press-review/479-ict-all-that-is-between-rwanda-and-value-added-agriculture.html). 424 ICT4Ag website (www.ict4ag.org/en/agenda/sessions.html); Ibid. 425 In 2015-16 Rwanda developed its own ICT4Rag strategy. See Ministry of Agriculture and Animal Resources. ‘National ICT4Rag Strategy (2016-2020)’. Undated report (www.minagri.gov.rw/fileadmin/user_upload/documents/policies_and_ strategy/ICT4RAg_STRATEGIC_PLAN_2016-2020_final__final__3_.pdf). 426 One study concludes that there is a “positive association between land use consolidation and crop yields, but only among farm households with landholdings greater than one hectare, which is well above the average farm size in Rwanda.” See Nilsson, P. 2018. ‘The Role of Land Use Consolidation in Improving Crop Yields among Farm Households in Rwanda’. The Journal of Development Studies (www.tandfonline.com/doi/pdf/10.1080/00220388.2018.1520217?needAccess=true). 427 Smart Nkunganire System website (www.smartnkunganire.rw/). 428 Africa Legal Network. 2015. ‘Investment Guide – Rwanda’ (www.africalegalnetwork.com/wp-content/uploads/2015/12/ Rwanda-Investment-Guide-2015.pdf); Nkurunziza, M. 2018. ‘Govt targets investments worth $90m in agriculture’. New Times (www.newtimes.co.rw/business/govt-targets-investments-worth-90m-agriculture). 429 Lab website (www.klab.rw/public/about). 430 AgriHack Talent Initiative website (http://hackathon.ict4ag.org/tag/rwanda/). 431 Ntirenganya, E. 2018. ‘$100m innovation fund in offing’. New Times (www.newtimes.co.rw/section/read/230566). 432 Ibid. 433 There are a number of challenges specific to the Sahel. Chief among these include: (i) governance and security challenges (e.g., rising insecurity, violent conflicts, cross-border threats), which are compounded by weak state institutions and inadequate provision of public services; (ii) development and humanitarian challenges (namely food insecurity, forced displacement, and vulnerability to external shocks). As an illustration, ~6.9 million people across the Sahel are currently experiencing a food crisis. These challenges are exacerbating security issues in the region; (iii) socio-economic challenges resulting from unemployment, inequality and lack of job opportunities. To face these challenges, the G5 Sahel group was created in December 2014 with its membership comprised of Burkina Faso, Chad, Mali, Mauritania and Niger. 434 Rounded to the nearest hundred thousand and based on the number of users of Nigeria-headquartered solution providers, rather than number of users of any D4Ag solutions in country. This figure does not adjust for users who may use more than one solution. 226 ENDNOTES 435 OECD.Stat website (www.stats.oecd.org/viewhtml.aspx?datasetcode=DACSECTOR&lang=en). 436 PIP G5 Sahel website (https://www.conference-g5sahel.org/copie-de-apropos-g5). 437 Mali, with mobile phone penetration above the Sub-Saharan Africa average, and Burkina Faso have strong digital foundations off of which to build attractive and widely-used D4Ag products. The resulting popularity of mobile money, even among rural populations, led MTN, Orange, and Airtel to launch cross-border money transfers between these two nations and several non-G5 neighbours in 2016. Chad and Niger, by contrast, exhibit mobile penetration rates towards the lowest end of the Sub-Sahara Africa spectrum. These countries have historically taxed MNOs at extraordinary rates (e.g., 50% in Chad as of 2014), discouraging investment and resulting in the second lowest and lowest scores in GSMA’s Connectivity Index. Niger, however, has taken steps to increase its connectivity, granting Airtel a new license, reviewing its MNO tax system, funding its part of the Trans-Saharan Backbone network, and attracting Orange Bank’s service to apply for licensing. As a result, Niger currently boasts the highest compound annual growth rate (CAGR) of unique mobile phone subscribers in Africa (6%). Mauritania, in contrast to the other four G5 nations, has a majority urban population, so its relatively high mobile penetration rate conceals exceedingly low coverage and poor service in rural areas – this contributes to dismal mobile money penetration despite decent phone penetration. 438 World Bank. 2017. ‘The Global Findex Database’; defined as “Used a mobile phone or the internet to access an account (% age 15+)” and “Used a mobile phone or the internet to access an account, rural (% age 15+).” 439 These figures are based on interviews and questionnaires. 440 IICD. 2010. ‘Increasing agricultural production through ICT’ (https://iicd.org/documents/increasing-agricultural- production-through-ict-lessons-learned-from-a-farmers-federation-in-burkina-faso/). 441 SNV website (http://www.snv.org/project/satellites-pastoralism-and-climate-change-stamp). 442 An aggregator is an application, software, or organisation that gathers multiple sources of data or information, processes them (possibly), and redistributes them. Annex 2 There are no endnotes for this annex. Annex 3 – Detailed Methodology 443 See Lowder, S.K. et al. 2016. ‘The Number, Size, and Distribution of Farms, Smallholder Farms, and Family Farms Worldwide’. World Development (www.sciencedirect.com/science/article/pii/S0305750X15002703); see also, a more detailed report estimating all farms in Sub-Sahara Africa at 77 million of which 82% are <2 hectares in Lowder, S.K. et al. 2016 ‘Transformation in the size and distribution of farmland operated by household and other farms in Sub-Saharan Africa’, 2016 AAAE Fifth International Conference, available at https://ageconsearch.umn.edu/record/246969/?ln=en. 444 For the best source on pastoralist numbers in Africa, see United Nations Economic Commission for Africa (UNECA). 2017. ‘New Fringe Pastoralism: Conflict and Insecurity and Development in the Horn of Africa and the Sahel.’ (https://www. uneca.org/sites/default/files/PublicationFiles/new_fringe_pastoralism_eng1.pdf ). See also, Cervigni, R., & Morris, M., eds. 2016. ‘Confronting Drought in Africa’s Drylands: Opportunities for Enhancing Resilience’, Africa Development Forum Series (https://openknowledge.worldbank.org/bitstream/handle/10986/23576/9781464808173.pdf?sequence=4). For an estimate of pastoralist household size, see, e.g., ElHadi, Y. et al. 2012. ‘Factors influencing transient poverty among agro-pastoralists’, African Crop Science Journal. CTA 228 BIBLIOGRAPHY BIBLIOGRAPHY Acre Africa. 2015. ‘Implementation of Agricultural AfDB.. 2016. ‘Feed Africa – Strategy for agricultural Insurance’. PowerPoint Slides. (http://www.acreafrica.com/ transformation in Africa 2016-2025’. (https://www.afdb.org/ wp-content/uploads/2016/11/IMPLEMENTATION-OF- fileadmin/uploads/afdb/Documents/Generic-Documents/ AGRICULTURAL-INSURANCE-RWANDA.pdf) Feed_Africa-_Strategy_for_Agricultural_Transformation_in_ Africa_2016-2025.pdf ) Adamon, N. et al. 2017. ‘Africa’s Agricultural Transformation: Identifying Priority Areas and Overcoming Challenges’. Africa African Financial and Economic Data. (https://africadata.com/ Economic Brief. African Development Bank. (https://www.afdb. blog/2016/10/04/rwanda-system-boosts-agribusiness-loans/) org/fileadmin/uploads/afdb/Documents/Publications/AEB_ Volume_8_Issue_3.pdf) African Union Commission. 2006. ‘African Youth Charter’. UN. (http://www.un.org/en/africa/osaa/pdf/au/african_youth_ Addom, B. and Enghild, A. 2018. ‘Attracting private sector charter_2006.pdf) investment to digitalise African farming’. Technical Centre for Agricultural and Rural Cooperation. CTA blog post. (https:// Afrika, Jean-Guy. 2017. ‘Connectivity in Africa: reaping the www.cta.int/en/blog/article/attracting-private-sector-investment- benefits of digital globalization’. World Trade Organisation. to-digitalise-african-farming-sid053a6cbed-475f-46fe-bc5d- (https://www.wto.org/english/tratop_e/devel_e/a4t_e/s2Jean- a5feb137495c) Guy_Afrika.pdf) AFI. 2017. ‘Digitally Delivered Credit: Consumer Protection AGRA. 2017. ‘Doubling productivity and incomes of smallholder Issues and Policy Responses to New Models of Digital Lending.’ farmers in Africa’. (https://agra.org/doubling-productivity-and- (https://www.afi-global. org/publications/2633/Digitally- incomes-of-smallholder-farmers-in-africa/) Delivered-Credit-Consumer-Protection-Issues-and-Policy- AGRA. 2016. ‘Digital Harvest’. https://www.raflearning.org/ Responses-to-New-Models-of-Digital-Lending) sites/default/files/20161024_digital_harvest_final_report.pdf Africa News. ‘Digital in 2018: Africa’s internet users increase by and https://www.raflearning.org/sites/default/files/20160929_ 20%’. (http://www.africanews.com/2018/02/06/digital-in-2018- digital_harvest_case_studies_final.pdf africa-s-internet-users-increase-by-20-percent/) AGRA. 2018. ‘Africa Agriculture Status Report: Catalyzing African Center for Economic Transformation. 2017. ‘African Government Capacity to Drive Agriculture Transformation’. Transformation Report- Agriculture Powering Africa’s Economic (https://agra.org/wp-content/uploads/2018/10/AASR-2018. Transformation’. (http://acetforafrica.org/acet/wp-content/ pdf) uploads/publications/2017/10/ATR17-full-report.pdf ) AGRA. 2016. ‘ICT4Ag business models: How to sustain and African Development Bank Group (AfDB). 2018. ‘African grow the Digital Harvest’. Rural and Agricultural Finance Learning Development Bank says staple crops processing zones will (RAFL). (https://www.raflearning.org/post/ict4ag-business- transform African agriculture’. (https://www.afdb.org/en/ models-how-sustain-and-grow-digital-harvest) news-and-events/african-development-bank-says-staple-crops- AgroNature Nigeria. 2018. ‘FG, IFAD Flags-off Construction processing-zones-will-transform-african-agriculture-18485/) of 22 Rural roads for farmers in 7 States’. (https:// AfDB. 2017. ‘Accelerating youth e-agriculture entrepreneurship agronaturenigeria.com/fg-ifad-flags-off-construction-of-22-rural- Partner Workshop’. (https://www.afdb.org/fileadmin/uploads/ roads-for-farmers-in-7-states/) afdb/Documents/Events/ayaf2017/Partner_meeting_Nigeria_ AgroNigeria. 2018. ‘Tag: 2017 Wet Season Agricultural TOR_23_and_24_April.pdf) Performance Survey (APS) report’. (https://www.agronigeria. com.ng/tag/2017-wet-season-agricultural-performance-survey- aps-report/) BIBLIOGRAPHY 229 Aker, J.C., et al. 2012. ‘Zap it to me: The impacts of a mobile Boateng, George. 2017. ‘Agricultural transformation in Africa: cash transfer program’. Unpublished manuscript. (http://sites. The myths, key issues, and the new pathway’. African Center for tufts.edu/jennyaker/files/2010/02/Zap_it_to_Me_26feb2012. Economic Transformation. Blog post. (http://acetforafrica.org/ pdf) media/blog/agricultural-transformation-in-africa-the-myths-key- issues-and-the-new-pathway/) Alibaba. 2016. ‘Rural Taobao Overview.’ https://www. alibabagroup.com/en/ir/pdf/160614/09.pdf Bobbi Gray. 2018. ‘Digital Farmer Profiles: Reimagining Smallholder Agriculture’. Grameen Foundation. (https:// Alliance for a Green Revolution in Africa. 2017. ‘Africa grameenfoundation.org/resource/digital-farmer-profiles- Agriculture Status Report’. reimagining-smallholder-agriculture) Ansu, Yaw. 2014. ‘2014 African Transformation Report: Growth Boettiger, S., et al. 2017. ‘Successful agricultural transformations: with Depth’. African Center for Economic Transformation. Six core elements of planning and delivery’. Mckinsey. (https:// PowerPoint Presentation. (https://www.cgdev.org/sites/default/ www.mckinsey.com/industries/chemicals/our-insights/successful- files/ACET%20ATR%20presentation%20to%20CGD%20 agricultural-transformations-six-core-elements-of-planning-and- 042914.pdf) delivery ) Apps4Ag Database. Technical Centre for Agricultural and Rural Brand, Nikki. 2017. ‘Bridging Ghana’s Digital Divide’. Agrilinks. Cooperation. (https://www.agrilinks.org/post/advancing-digital-technology- Ashour et al. 2015. “An Evaluation of the Impact of agriculture-ghana) E-verification on Counterfeit Agricultural Inputs and Technology Briter Bridges. ‘Technology for agriculture in Africa: a fourth Adoption in Uganda”. IFPRI. http://www.ifpri.org/publication/ revolution without a third?’. (https://briterbridges.com/ evaluation-impact-e-verification-counterfeit-agricultural-inputs- agriculture-in-africa-a-fourth-revolution-without-a-third and-technology-adoption Brune, L. et al. 2015. ‘Facilitating Savings for Agriculture: Field Bafana, Busani. 2014. ‘Denting youth unemployment through Experimental Evidence from Malawi’, NBER Working Paper No. agriculture’. UN-AfricaRenewal Magazine. UN. (https://www. 20946. (https://www.nber.org/papers/w20946) un.org/africarenewal/magazine/special-edition-agriculture-2014/ denting-youth-unemployment-through-agriculture) Business Wire. 2018. ‘Alibaba Cloud Launches ET Agricultural Brain at the Shanghai Computing Conference: Barry Callebaut. 20/06/2016. ‘Barry Callebaut collaborates Pig farming enterprise demonstrates how AI technology is with SAP to offer an innovative app to boost sustainability data revolutionizing industry’. (https://www.businesswire.com/ management’. (https://www.barry-callebaut.com/en/group/ news/home/20180606006487/en/Alibaba-Cloud-Launches- media/news-stories/barry-callebaut-collaborates-sap-offer- Agricultural-Brain-Shanghai-Computing) innovative-app-boost-sustainability) Business Wire. 2014. ‘Three-Year Study Concludes that Belmaachi, Lina. 14/02/2018. ‘African Agtech Market Map: CocoaLink ‘Significantly Improved the Livelihoods’ of 99 Technologies Changing the Future of Agriculture in Africa’. Ghanaian Cocoa Farmers’. (https://www.businesswire. AgFunder News. (https://agfundernews.com/african-agtech- com/news/home/20140318005928/en/Three-Year-Study- market-map.html) Concludes-CocoaLink-%E2%80%9CSignificantly-Improved- Benin, Samuel. 2016. ‘Agricultural productivity in Africa: Livelihoods%E2%80%9D) Trends, patterns, and determinants’. International Food CGAP. 2019. ‘Fintechs and Financial Inclusion: Lessons Policy Research Institute. (http://www.ifpri.org/publication/ Learned’. https://www.cgap.org/sites/default/files/ agricultural-productivity-africa-trends-patterns-and-determinants) publications/2019_05_Case_Study_Fintech_and_Financial_ Better Than Cash Alliance (BTCA). 2017. “The Role of Digital Inclusion.pdf Payments in Sustainable Agriculture and Food Security” (2017). CGAP. 2018. ‘Super Platforms: Connecting Farmers to Markets https://btca-prod.s3.amazonaws.com/documents/313/english_ in Africa’. https://www.cgap.org/blog/super-platforms- attachments/Agriculture_Report.pdf?1508858199 connecting-farmers-markets-africa) Bill and Melinda Gates Foundation. 2017. BMGF ICT4Ag Strategy. (internal document) 230 BIBLIOGRAPHY Chamberlin, Jordan. 2007.’Defining smallholder agriculture in Ding, D. et al. 2017. ‘From Ant Financial to Alibaba’s Ghana’. International Food Policy Research Institute. (http:// Rural Taobao Strategy - How Fintech Is Transforming www.ifpri.org/publication/defining-smallholder-agriculture- Social Inclusion.’ https://www.researchgate.net/ ghana) publication/328274301_From_Ant_Financial_to_Alibaba’s_ Rural_Taobao_Strategy_-_How_Fintech_Is_Transforming_ Chemonics International Inc. 2017. ‘Maximizing agricultural Social_Inclusion revenue and key enterprises in targeted sites II’. (https://www. chemonics.com/wp-content/uploads/2017/10/MARKETS_II_ Diop, Makhtar. 2016. ‘Foresight Africa 2016: Banking on Final_Report_revised-submission_8Dec2017.pdf) agriculture for Africa’s future’. Brookings blog post. (https://www. brookings.edu/blog/africa-in-focus/2016/01/22/foresight-africa- Chen, J. 2017. ‘Ant Financial: Our rural China practice’. 2016-banking-on-agriculture-for-africas-future) (https://www.slideshare.net/ExternalEvents/ant-financial-our- rural-finance-practice) Doing Business. 2019. ‘Training for Reform: Economy Profile, Kenya’. (http://www.doingbusiness.org/content/dam/ Chowdhury, A.H., et al. 2011. ‘Contribution of farmer to-farmer doingBusiness/country/k/kenya/KEN.pdf) video to capital assets building: Evidence from Bangladesh’. Journal of Sustainable Agriculture 35(4), 408–435. Access Agriculture. Edia, Hannah. 2017. ‘Agric Tech in Nigeria: How Technology (https://www.accessagriculture.org/sites/default/files/1._ Powers Agriculture’. Farmcrowdy blog post. (http://blog. summary_Chowdhury_et_al_2011_ENGLISH.pdf) farmcrowdy.com/agric-tech-company-in-nigeria-how-technology- powers-agriculture/) Cornell University. Agricultural Experiment Station. ‘Digital Agriculture’. (https://cuaes.cals.cornell.edu/digital-agriculture/) Ehui, S. 2018. ‘Foresight Africa viewpoint – Why technology will disrupt and transform Africa’s agriculture sector in a good Cotton Made in Africa. (https://www.cottonmadeinafrica.org/ way’. Brookings. (https://www.brookings.edu/blog/africa-in- en/about-us/the-initiative) focus/2018/01/26/foresight-africa-viewpoint-why-technology- CTA Catalogue. Technical Centre for Agricultural and Rural will-disrupt-and-transform-africas-agriculture-sector-in-a-good- Cooperation. way/) Dalberg & MCF MCF Rural Agriculture Finance Learning Lab Ehui, Simeon. 2018. ‘How can digital technology help transform (RAFLL). 2018. ‘Big data could mean big opportunity: why we Africa’s food system?’. World Bank. Blog post. (http://blogs. should stay excited for data analytics in smallholder finance.’ worldbank.org/ic4d/how-can-digital-technology-help-transform- (https://www.raflearning.org/sites/default/files/learning_ africa-s-food-system) brief_5_-_data_analytics-final.pdf?token=g6FuZCx4) Ehui, Simeon. 2018. ‘Why technology will disrupt and transform Dalberg & MCF Rural Agriculture Finance Learning Lab Africa’s agriculture sector—in a good way’. World Bank. Blog (RAFLL). 2017. ‘The business case for digitally-enabled post. (https://blogs.worldbank.org/voices/why-technology-will- smallholder finance’. Rural and Agricultural Finance Learning disrupt-and-transform-africa-agriculture-sector-good-way) (RAFL). (https://www.raflearning.org/sites/default/files/ Ekekwe, N. 2018. ‘How Digital Technology Is Changing business_case_for_digitally_enabled_smallholder_finance. Farming in Africa?’. Harvard Business Review. (https://hbr. pdf?token=TH74R3Qn) org/2017/05/how-digital-technology-is-changing-farming-in- DeBouef et al. 2014. “Counterfeiting in African Agriculture – africa) Challenges and Solutions,” Bill and Melinda Gates Foundation. Enable impact. ‘Hello Tractor’. (http://www.enableimpact.com/ https://www.agrilinks.org/library/counterfeiting-african- invest/hello-tractor/) agriculture-inputs-challenges-and-soalutions Esoko. 2013. ‘Mobile and Development Intelligence’. GSMA. Digital Financial Services. 2017.’DFS Lab Announces (https://www.gsma.com/mobilefordevelopment/programme/ Investment in African Fintech Startups’. PR Newswire. (https:// magri/esoko/) www.prnewswire.com/news-releases/dfs-lab-announces- investment-in-african-fintech-startups-300476075.html) Esoko. 2017. ‘Insight on Baseline Study of Smallholder Farmers in Ghana’. (https://esoko.com/insight-baseline-study- Digital Green. ‘We stand with the world’s smallholder farmers’. smallholder-farmers-ghana/) (http://www.digitalgreen.org/global-impact/) BIBLIOGRAPHY 231 Ewan, K. ‘WeFarm: Access and share vital information, FAO. 2013. ‘ICT Uses for Inclusive Agricultural Value Chains’. even without access to internet’. Changemakers. (https://www. http://www.fao.org/3/a-aq078e.pdf changemakers.com/globalgoals2015/entries/wefarm) FAO. 2010. ‘Agricultural value chain development: Threat Farmdrive. (https://farmdrive.co.ke/about-us) or opportunity for women’s employment?’. Gender and Rural Employment Policy Brief. (http://www.fao.org/docrep/013/i2008e/ Farmerline. 2018. ‘Farmerline launches new CocoaLink Service: i2008e04.pdf) A free mobile app that puts the power of mobile technology in farmers’ hands’. (https://farmerline.co/2018/05/24/farmerline- FAO. 2006. ‘Rapid growth of selected Asian economies. Lessons launches-new-cocoalink-service-a-free-mobile-app-that-puts-the- and implications for agriculture and food security’. Synthesis report. power-of-mobile-technology-in-farmers-hands/) (www.fao.org/docrep/009/ag087e/AG087E00.htm) Federal Ministry of Communications Government of Nigeria. FAO and International Telecommunication Union (ITU). 2016. 2017. ‘Nigeria ICT Roadmap 2017-2020’. (http://www. ‘E-agriculture strategy guide-Piloted in Asia-Pacific countries’. commtech.gov.ng/Doc/Nigeria_ICT_Roadmap_2017-2020.pdf) (http://www.fao.org/3/a-i5564e.pdf) Finsmes. 2018.’Wefarm Secures $5M New Financing Round’. Ford, Neil. 18/08/2017. ‘Dangote turns to Nigerian agriculture’. (http://www.finsmes.com/2018/03/wefarm-secures-5m-new- African Business. (https://africanbusinessmagazine.com/sectors/ financing-round.html) agriculture/dangote-turns-nigerian-agriculture/) Fleshman, Michael. 2009. ‘Laying Africa’s roads to Gates Foundation. 2014. ‘Using Mobile Data for Development’. prosperity’. UN-AfricaRenewal Magazine. UN. (https://www. (https://docs.gatesfoundation.org/Documents/Using%20 un.org/africarenewal/magazine/january-2009/laying- Mobile%20Data%20for%20Development.pdf) africa%E2%80%99s-roads-prosperity) Geodata for Agriculture and Water. 2016. ‘SAGCOT Integrated Food and Agriculture Organization of the United Nations Knowledge and Information for Agriculture’. (https://nso-g4aw. (FAO). 2018. ‘Sustainable Agricultural Mechanization: A akvoapp.org/en/project/5313/) Framework for Africa’. http://www.fao.org/3/CA1136EN/ ca1136en.pdf Geodata for Agriculture and Water. 2018. ‘Sat4Farming - Information services for sustainable cocoa farm management FAO. 2018. ‘Gender and ICTs: Mainstreaming gender in the in Ghana’. (https://g4aw.spaceoffice.nl/en/projects/g4aw- use of information and communication technologies (ICTs) for projects/91/sat4farming---information-services-for-sustainable- agriculture and rural development’. (http://www.fao.org/3/ cocoa-farm-management-in-ghana.html) i8670en/I8670EN.pdf) German Development Institute (GDI). 2017. ‘Unlocking the FAO. 2017. ‘Ghana E-Agriculture Programme Ministry of Power of Irrigation for Sub-Saharan Africa’. https://www.die- Food and Agriculture, Republic of Ghana’. e-agriculture News. gdi.de/uploads/media/BP__7.2017.pdf (http://www.fao.org/e-agriculture/news/ghana-e-agriculture- programme-ministry-food-and-agriculture-republic-ghana) GFAR, GODAN, CTA. 2018. ‘Digital and Data-Driven Agriculture: Harnessing the Power of Data for Smallholders’. FAO. 2017. ‘Regional Overview of Food Security and Nutrition https://f1000research.com/documents/7-525 in Africa 2017. The food security and nutrition–conflict nexus: building resilience for food security, nutrition and peace’. Gichamba, A., et al. 2017. ‘An Assessment of e-Extension (https://reliefweb.int/sites/reliefweb.int/files/resources/a-i7967e. Platforms in Kenya’. International Journal of Innovative Studies in pdf) Sciences and Engineering Technology (IJISSET), 3, 36-40. (http:// ijisset.org/wp-content/uploads/2017/08/IJISSET-030713.pdf) FAO. 2017. ‘Information and Communication Technology (ICT) in Agriculture’. A Report to the G20 Agricultural Deputies. Giokos, Eleni and McDougall, Mary. 2018. ‘This drone startup (http://www.fao.org/3/a-i7961e.pdf) helps farmers beat disease and drought’. CNN Business. (https:// money.cnn.com/2018/02/15/technology/aerobotics-farm-app- FAO. 2017. ‘The number of people suffering from chronic drones/index.html) undernourishment in sub-Saharan Africa has increased’. News Article. (http://www.fao.org/news/story/en/item/1062874/ Githakwa, Hellen. 2017. ‘Strengthening Kenya’s icode/) ‘Digital Harvest’. Africa Lead. (https://www.africaleadftf. org/2017/07/31/strengthening-kenyas-digital-harvest/) 232 BIBLIOGRAPHY GIZ (Deutsche Gesellschaft für Internationale Zusammenarbeit). Grameen Foundation. 2018. ‘Grameen Foundation and Partners 2019. ‘Harnessing the chances of digitalisation for rural bringing satellite technology to benefit smallholder cocoa farmers development.’ http://www.ruralfinanceandinvestment.org/sites/ in Ghana’. (https://grameenfoundation.org/press-releases/ default/files/180807_Harnessing-A4.pdf grameen-foundation-and-partners-bringing-satellite-technology- benefit-smallholder) GIZ, 2017. ‘Use of ICT for Agriculture in GIZ Projects’. https://www.snrd-africa.net/wp-content/uploads/2017/04/ GSMA. 2019. ‘The Mobile Economy.’ (https://www. study-useofictforagricultureingizprojectsgizsnrdafri gsmaintelligence.com/research/?file=b9a6e6202ee1d5f787cfebb9 ca-160609104357-1.pdf 5d3639c5&download) GIZ. 2013. ‘The African Cashew initiative (ACi) Promoting GSMA. 2019. ‘The Mobile Gender Gap Report’. (https://www. Competitiveness of African Cashew Farmers’. Europa. (https:// gsma.com/mobilefordevelopment/resources/mobile-gender-gap- europa.eu/capacity4dev/file/16810/download?token=4zE0-9px report-2019/) cannot open) GSMA. 2018. ‘Opportunities in agricultural value chain GOGLA/Dalberg. 2018. ‘Off-grid Solar Market Trend Report’. digitization’. (https://www.gsma.com/mobilefordevelopment/wp- (https://www.lightingglobal.org/2018-global-off-grid-solar- content/uploads/2018/01/Opportunities-in-agricultural-value- market-trends-report/) chain-digitisation-Learnings-from-Ghana.pdf) Government of Ethiopia. 2009. ‘The National Information GSMA. 2018. ‘The Mobile Economy Sub-Saharan Africa’. and Communication Technology (ICT) Policy and Strategy’. (https://www.gsmaintelligence.com/research/?file=809c442550e (http://www.ethiopia.gov.et/policies-and-strategies/-/ 5487f3b1d025fdc70e23b&download) asset_publisher/f4J0Jqrj9180/content/the-national- information-and-communication-technology-ict-policy-and- GSMA. 2018. ‘The Mobile Economy West Africa’. (https:// strategy?inheritRedirect=false ) www.gsmaintelligence.com/research/?file=e568fe9e710ec776d82 c04e9f6760adb&download) Government of Nigeria. ‘ICT infrastructure an essential foundation for implementing the WSIS process in Nigeria’. GSMA. 2018. ‘The Mobile Economy’. (https://www.gsma.com/ (https://www.ncc.gov.ng/archived/speeches/EVC’S%20 mobileeconomy/wp-content/uploads/2018/02/The-Mobile- PAPER%20FOR%20e-NIGERIA%20-%20ICT%20 Economy-Global-2018.pdf) INFRASTRUCTURE%20AND%20IMPLEMENTATION%20 GSMA. 2018. ‘The role of digital in improving traceability OF%20THE%20WSIS%20PROCESS%20IN%20NIGERIA.. and certification in the agricultural last mile’. ( https://www. pdf) gsma.com/mobilefordevelopment/blog-2/the-role-of-digital-in- Government of Senegal. 2016. ‘Emergent Senegal Plan Priority improving-traceability-and-certification-in-the-agricultural-last- Action Plan 2014-2018’. (http://www.ongpacte.org/en/plan- mile/) senegal-emergent-plan-dactions-prioritaires-2014-2018/) GSMA. 2018. ‘Twiga Foods: Improved market access for Government of Senegal. 2014. ‘Emerging Senegal Plan: 2014- farmers and a reliable supply for vendors’. (https://www.gsma. 2018 Priority Action Plan’. (https://www.pseau.org/outils/ com/mobilefordevelopment/wp-content/uploads/2018/05/ biblio/resume.php?d=5454) Twiga-Foods-Improved-market-access-for-farmers-and-a-reliable- supply-for-vendors.pdf) Government of Zambia. 2018. ‘7th National Development Plan 2017-2021 – Zambia’. (http://www.globalcrrf.org/crrf_ GSMA. 2018. ‘The Mobile Gender Gap Report’. (https://www. document/7th-national-development-plan-2017-2021-zambia/) gsma.com/mobilefordevelopment/connected-women/the-mobile- gender-gap-report-2018/) Goyal, A. 2008. ‘Information technology and rural market performance in central India (Doctoral GSMA. 2017. ‘Prerequisites to digitising the agricultural last dissertation)’. (https://pdfs.semanticscholar.org/fbfc/ mile’. (https://www.gsma.com/mobilefordevelopment/resources/ bb7ee8b5e449c5d71b65d803d4aef4201bac.pdf) prerequisites-to-digitise-the-agricultural-last-mile/) Graham, Karen. 2017. ‘Africa’s agriculture needs a big dose of GSMA. 2017. ‘Digital inclusion and mobile sector taxation in digital transformation’. Digital Journal. (http://www.digitaljournal. Niger’. (https://www.gsma.com/mobilefordevelopment/wp- com/tech-and-science/technology/africa-s-agriculture-needs-a- content/uploads/2017/01/Digital-inclusion-and-mobile-sector- big-dose-of-digital-transformation/article/504124) taxation-in-Niger_English_report.pdf) BIBLIOGRAPHY 233 GSMA. 2017. ‘Farmers’ Club -A mobile agriculture iCow. 2012. ‘iCow Impact Study Results’. (http://www.icow. service by Vodafone Ghana’. (https://www.gsma.com/ co.ke/blog/icow-impact-study-results) mobilefordevelopment/wp-content/uploads/2017/06/farmers- club-mobile-agriculture-service-vodafone-ghana.pdf) ICT Authority. ‘National ICT Masterplan’. (http://icta.go.ke/ national-ict-masterplan/) GSMA. 2017. ‘M’chikumbe 212-A mobile agriculture service by Airtel Malawi’. (https://www.gsma.com/mobilefordevelopment/ ICT4DBlog. 2018. ‘Crowdfarming platform enabled investment wp-content/uploads/2017/07/M%E2%80%99chikumbe-212-A- into Nigerian Agriculture’. (https://ict4dblog.wordpress. mobile-agriculture-service-by-Airtel-Malawi.pdf) com/2018/11/20/crowdfarming-platform-enabled-investment- in-nigerian-agriculture/) GSMA. 2017. ‘Mapping the mobile money gender gap: Insights from Côte d’Ivoire and Mali’. (https://www.gsma.com/ IFC. 2018. ‘Digital Financial Services for Agriculture mobilefordevelopment/wp-content/uploads/2017/07/CW_ Handbook’, available at https://www.ifc.org/wps/wcm/ Cote_Mali_gendergap_Phase2_V2_WEBOK.pdf) connect/4ca05121-fe39-42ae-891f-76203c7b91f0/Digital+ Financial+Services+for+Agriculture_IFC%2BMCF_2018. GSMA. 2017. ‘State of the Industry Report on Mobile Money’. pdf?MOD=AJPERES (https://www.gsma.com/mobilefordevelopment/wp-content/ uploads/2018/05/GSMA_2017_State_of_the_Industry_Report_ Inclusive Finance Platform. ‘Creating Equal Opportunities’. on_Mobile_Money_Full_Report.pdf) (http://www.inclusivefinanceplatform.nl/documents/npm- 20geodata-20and-20ict-20solutions-20for-20inclusive-20finance- GSMA. 2017. ‘The Mobile Economy Sub-Saharan Africa’. 20and-20food-20security.pdf) (https://www.gsmaintelligence.com/research/?file=7bf3592e6d7 50144e58d9dcfac6adfab&download) Initiative for Smallholder Finance (ISF). 2018. ‘Protecting growing prosperity: Agricultural insurance in the developing GSMA. 2017. ‘Mapping the mobile money gender gap: insights world.’ (https://www.raflearning.org/sites/default/ from Cote d’Ivoire and Mali’. (https://www.gsma.com/ files/sep_2018_isf_syngneta_insurance_report_final. mobilefordevelopment/resources/mapping-the-mobile-money- pdf?token=1i4u5GwD) gender-gap-our-insights-from-cote-divoire-and-mali/) International Food Policy Research Institute. ‘Topic: Food GSMA. 2016. “Market size and opportunity in digitising security’. (http://www.ifpri.org/topic/food-security) payments in agricultural value chains”. (https://www. gsmaintelligence.com/research/?file=29e480e55371305d7b37fe4 Kirui, O.K., et al. 2013. ‘Impact of mobile phone-based money 8efb10cd6anddownload) transfer services in agriculture: evidence from Kenya’. Quarterly Journal of International Agriculture, 52 (892-2016-65177), 141. GSMA. 2015. ‘Market size and market opportunity for (https://ageconsearch.umn.edu/bitstream/173644/2/3_Kirui. agricultural value-added-services’. (https://www.gsma.com/ pdf) mobilefordevelopment/wp-content/uploads/2015/02/Market- size-and-market-opportunity-for-agricultural-value-added-services- Liangzhi Y. et al. 2010. ‘What is the irrigation potential of Agri-VAS.pdf) Africa’, IFPRI. GSMA. 2015. ‘Tigo Kilimo: Impact Evaluation’. (https:// Liao, Rita. 2018. ‘Exclusive: Alibaba gets into farming – without www.gsma.com/mobilefordevelopment/wp-content/ getting its hands dirty’. Tech in Asia. (https://www.techinasia. uploads/2015/09/GSMA_Tigo_Kilimo_IE.pdf) com/alibaba-ai-et-brain-agriculture) Hershey. (https://www.thehersheycompany.com/en_ca/ Lighting Global. 2018. ‘Off-Grid Solar Market Trends responsibility/good-business/creating-goodness/cocoa- Report’. (https://www.lightingglobal.org/wp-content/ sustainability/cocoa-link.html) uploads/2018/02/2018_Off_Grid_Solar_Market_Trends_ Report_Summary.pdf) Hildebrandt, N., et al. 2015. ‘Price Information, Inter-Village Networks, and Bargaining Spillovers: Experimental Evidence Lohento, Ken and Ajilore, Oluwabunmi. 2015. ‘ICT and from Ghana’. Preliminary draft. Retrieved from http://sites. Youth in Agriculture’. (https://www.researchgate.net/ bu.edu/neudc/files/2014/10/paper_345.pdf publication/308968253_ICT_and_Youth_in_Agriculture) IBA Ventures. ‘ Nfrnds’. (https://iba.ventures/ventures/nfrnds/) 234 BIBLIOGRAPHY Lowder, S.K., et al. 2016. ‘The number, size, and distribution Mercy Corps AFA & Dalberg. 2018. ‘Benchmarking of farms, smallholder farms, and family farms worldwide’. World E-Commerce Models for African Smallholders’. https://www. Development, 87, 16-29. (https://www.sciencedirect.com/science/ findevgateway.org/sites/default/files/publication_files/afa_ article/pii/S0305750X15002703) ecommerce_benchmark_slideshare_9.17_fnl.pdf Macauley, H. and Ramadjita, T. 2015. ‘Cereal crops: Mercy Crops. 2018. ‘Growing up hungry: Quick facts about Rice, maize, millet, sorghum, wheat’. Feeding Africa. African malnutrition’. (https://www.mercycorps.org/articles/growing- Development Bank. (https://www.afdb.org/fileadmin/uploads/ hungry-quick-facts-about-malnutrition afdb/Documents/Events/DakAgri2015/Cereal_Crops-_Rice__ Maize__Millet__Sorghum__Wheat.pdf) Mercy Crops. 2018. ‘Mercy Corps’ Social Ventures: FarmDrive’. (https://www.mercycorps.org/sites/default/files/MC_SVF_ Macharia, Joel. 2014. ‘Internet access is no longer a luxury’. UN- One%20pager%20FarmDrive%20July2018.pdf) AfricaRenewal Magazine. UN. (https://www.un.org/africarenewal/ magazine/april-2014/internet-access-no-longer-luxury) MercyCorps AFA & Dalberg. 2019. ‘FtMA Digitalization Lessons Learned’. http://mercycorpsagrifin.org/wp-content/ Magowe, N. 2017. ‘Creating Shared Value Price’. uploads/2019/01/AFA-FtMA_Digitization-and-lessons-learned_ Changemakers. (https://network.changemakers.com/challenge/ FIN.pdf creatingsharedvalue/refining/magri) Microsoft Research Podcast. 2018. ‘AI for the Developing World MAgro. ‘Impact’. (https://www.myagro.org/impact/) with Dr. Ranveer Chandra’. Blog post. (https://www.microsoft. com/en-us/research/blog/ai-developing-world-dr-ranveer- Maina, Saruni. 2018. ‘Safaricom’s DigiFarm Aims to Put More chandra/ ) Coins in Farmers’ Pockets Through Technology’. Techweez. (https://techweez.com/2018/07/23/safaricom-digifarm-more- Mpala, Daniel. 2018. ‘ SA’s Aerobotics launches five new coins-farmers/) precision agriculture products’. Ventureburn. (http://ventureburn. com/2018/10/aerobotics-new-precision-agriculture-products/) Malabo Monpellier Panel. 2019. ‘Smart Value Chains: Policy Innovations For the Digitalisation of African Agriculture.’ Mtyszler Bitbucket (https://mtyszler.bitbucket.io/ https://www.mamopanel.org/resources/reports-and-briefings/ cocoalivelihoods_cocoa/) Malabo Montpellier Panel. 2018. ‘Mechanized – Transforming Muro, M. et al. 2017. ‘Digitalization and the American Africa’s Value Chains’. https://www.mamopanel.org/media/ workforce’. Brookings. (https://www.brookings.edu/research/ uploads/files/MaMo2018_Mechanized_Transforming_Africas_ digitalisation-and-the-american-workforce/) Agriculture_Value_Chains.pdf Mzuyanda, C. 2014. ‘Assessing the impact of primary MasterCard Foundation Rural Agriculture Finance Lab agricultural co-operative membership on smallholder farm (RAFLL). 2017. ‘How can digital tools enable smallholder performance (crops) in Mnquma Local Municipality of the finance’. https://www.slideshare.net/MaliaBachesta/raf-ll-wapl- Eastern Cape Province’. (http://libdspace.ufh.ac.za/bitstream/ session-5 handle/20.500.11837/315/M%20Agric%20%28Agric%20 Eco%29%20Dissertation%20CHRISTIAN%2C%20M%20-%20 Mattern, Max. 2018. ‘How Ghana Became One of Africa’s Top 2015.pdf?sequence=1&isAllowed=y) Mobile Money Markets’. CGAP Blog post. (https://www.cgap.org/ blog/how-ghana-became-one-africas-top-mobile-money-markets) National Planning Commission. 2016. ‘Federal Democratic Republic of Ethiopia: Growth and Transformation Plan II McArthur, J.M and McCord, G.C. 2014. ‘Fertilizing Growth: (GTP II)’. Europa. (https://europa.eu/capacity4dev/resilience_ Agricultural Inputs and their effects in Economic Development’. ethiopia/document/growth-and-transformation-plan-ii-gtp- Global Economy and Development (Working Paper 77). Brookings. ii-201516-201920) (https://www.brookings.edu/wp-content/uploads/2014/09/ fertilizing-growth-final-v3.pdf) NEPAD Planning and Coordinating Agency. 2015. ‘The CAADP Results Framework [2015-2025]’. McNally, P. 2014. ‘Case Study: Reuters Market Light’. GSMA. (https://www.gsma.com/mobilefordevelopment/programme/ Norton Rose Fulbright. (https://www.nortonrosefulbright.com/ magri/reuters-market-light/) 404?item=%2fknowledge%2fpublications%2f129789%2fthe-qa- alloysius-attah-of-farmerline&user=extranet%5cAnonymous&site =NRFWeb) BIBLIOGRAPHY 235 Nsingize, G. 2013. ‘The contribution of agricultural cooperatives Opportunity International. ‘Financing Smallholder Farmers to to small holder farmers’ household income’. A Case of COAMV Increase Incomes and Transform Lives in Rural Communities’. Cooperative, Burera District, Rwanda. (http://edepot.wur. (https://opportunity.org/content/News/Publications/ nl/279035) Knowledge%20Exchange/Financing-Smallholder-Farmers- Opportunity-International.pdf) O’Reilly, M. 2014. ‘When do we award larger grants?’. Indigo trust. (https://indigotrust.org.uk/tag/icow/) Precision Agriculture for Development. ‘Randomized Controlled Trials’. (http://precisionag.org/our-findings/rcts) Obonyo, Raphael. 2018. ‘Towards a food-secure Africa’. UN- AfricaRenewal Magazine. UN. (https://www.un.org/africarenewal/ Prinsloo et al. 2017. “The role of the Namibian Livestock magazine/december-2017-march-2018/towards-food-secure- Traceability Systems in containing the recent foot-and- africa) mouth disease outbreak.” https://ieeexplore.ieee.org/ document/8016172 OECD. 2019. ‘Digital Transformation of the Agriculture and Food System.’ https://issuu.com/oecd.publishing/docs/the_ Punch. 2017. ‘Nigeria needs more investment in rural digital_transformation_of_the_a infrastructure to boost agriculture – Adesina’. (https://punchng. com/nigeria-needs-more-investment-in-rural-infrastructure-to- Olurounbi, R. 2018. ‘Nigeria Seen as Biggest Rice Buyer in boost-agriculture-adesina/) 2019, Behind China’. Bloomberg. (https://www.bloomberg. com/news/articles/2018-11-14/usda-sees-nigeria-rice-imports- PWC. 2017. ‘Africa Agribusiness Survey (2016/2017)’. (https:// increasing-to-3-4m-tons-in-2019) www.pwc.co.za/en/assets/pdf/agri-businesses-insights-survey- may-2016.pdf) Ombati, Egra. 28/11/2016.’Winning Startups from Botswana, Senegal and Uganda earn investments at Pitch AgriHack!’. Rainforest Alliance. 2018. ‘Digital and Satellite Technology IHUB. (https://ihub.co.ke/blogs/29024/winning-startups-from- Program Launches to Support Ghana’s Smallholder Cocoa botswana-senegal-and-uganda-earn-investm) Farmers’. Press Release. (https://www.rainforest-alliance.org/ press-releases/digital-satellite-technology-program-launch-ghana) Omidyar. ‘Segmenting Enterprises to Better Understand Their Financial Needs’. (https://www.omidyar.com/sites/default/files/ Rapsomanikis, George. 2015. ‘The economic lives of smallholder Enterprise_Segmentation_Report.pdf) farmers: An analysis based on household data from nine countries’. Food and Agriculture Organization of the United Omidyar. 2018. ‘Bringing Insurance to Smallholder Farmers: Nations. Why We Invested in Pula’. Blog post. (https://www.omidyar. com/blog/bringing-insurance-smallhold-farmers-why-we- Rural and Agricultural Finance Learning (RAFL) and ISF invested-pula) Advisors. 07/04/2016. ‘Inflection Point: Unlocking growth in the era of farmer finance’. (https://www.raflearning.org/post/ One Acre Fund. 2016. ‘Scaling up agricultural credit in inflection-point-unlocking-growth-era-farmer-finance) Africa’. https://oneacrefund.org/documents/104/Scaling_Up_ Agricultural_Credit_In_Africa_Farm_Finance.pdf Shepherd Ben. 2017. ‘How Can Africa Better Participate in Global Value Chains? A Focus on Trade Facilitation’. One Acre Fund. 2012. ‘Catalyzing Smallholder Agricultural International Centre for Trade and Sustainable Development. Finance’. (https://oneacrefund.org/documents/101/Dalberg_ (https://www.ictsd.org/bridges-news/bridges-africa/news/how- Skoll_Citi_Catalyzing_Smallholder_Agricultural_Finance_Farm_ can-africa-better-participate-in-global-value-chains-a-focus-on) Finance.pdf) Siba, E. 26/01/2018. ‘Foresight Africa viewpoint – Nutrition One Acre Fund. ‘Bridging the Microfinance Gap for security: The last mile of Africa’s food security agenda’. Smallholder Farmers’. Blog post. (https://oneacrefund.org/blog/ Brookings blog post. (https://www.brookings.edu/blog/africa-in- bridging-microfinance-gap-smallholder-farmers/) focus/2018/01/26/foresight-africa-viewpoint-nutrition-security- One Acre Fund. ‘What Climate Change Means for Agriculture the-last-mile-of-africas-food-security-agenda/) in Africa’. Blog post. (https://www.oneacrefund.org/blog/what- Siba, Eyerusalem and Signé, Landry. 30/10/2017. ‘Four things climate-change-means-agriculture-africa). you should know about food security in Africa’. The Washington Post. (https://www.washingtonpost.com/news/monkey-cage/ wp/2017/10/30/four-things-you-should-know-about-food- security-in-africa/?utm_term=.b245eb26bfe0) 236 BIBLIOGRAPHY Signé, Landry. 16/10/2017. ‘The quest for food security The Guardian Reporter. 2018. ‘AGRIinsight platform to boost and agricultural transformation in Africa: Is the CAADP the agriculture in Africa’. IPP Media. (https://www.ippmedia.com/ answer?’. Brookings. (https://www.brookings.edu/blog/africa-in- en/news/agriinsight-platform-boost-agriculture-africa) focus/2017/10/16/the-quest-for-food-security-and-agricultural- transformation-in-africa-is-the-caadp-the-answer). The Guardian. 2017. ‘Only modernised commercial farms will fill Africa’s plate, economists warn’. (https://www.theguardian. Solon, O. 21/06/2018. ‘MFarm empowers Kenya’s farmers with com/global-development/2017/oct/10/younger-farmers-and- price transparency and market access’. Wired. (https://www. more-tractors-africa-needs-to-scale-up-says-report ) wired.co.uk/article/mfarm) The Guardian. 2016. ‘On Ghana’s cocoa farms, Fairtrade is not Statista. (https://www.statista.com/statistics/225614/net- yet working for women’. (https://www.theguardian.com/global- revenue-of-alibaba/) development/2016/mar/11/ghana-cocoa-farms-fairtrade-not-yet- working-for-women) Su, Q. & Yan, D. 2018. ‘Rural Taobao yields benefits for farmers by analyzing big data’, China Daily Asia. (http://epaper. The Mastercard Foundation. 2018. ‘Inflection point: Unlocking chinadailyasia.com/asia-weekly/article-13996.html) growth in the era of farmer finance’. (https://mastercardfdn. org/wp-content/uploads/2018/06/Inflection-Point_April-20160- Syngenta Foundation for Sustainable Agriculture (SFSA). 2019. accessible.pdf) ‘How can digital solutions help to feed a growing world? An introduction with case studies.’ Theunissen, Ian. 2015. ‘E-agriculture: How ICT is taking (https://www.syngentafoundation.org/file/12811/download). farming into the future’. Biztech Africa. (https://www. biztechafrica.com/article/e-agriculture-how-ict-taking-farming- Sustainable Food Lab. 2017. ‘Enabling Smallholder Farmers to future/10841/) Improve Their Incomes’. Improving Incomes Challenge White Paper. (http://sustainablefoodlab.org/wp-content/uploads/2017/09/ Uduji, J.I., et al. 2018. ‘The impact of e‐wallet on informal farm BFP-Improving-Incomes-WEB.pdf) entrepreneurship development in rural Nigeria’. The Electronic Journal of Information Systems in Developing Countries, e12066. Technical Centre for Agricultural and Rural Cooperation. (https://onlinelibrary.wiley.com/doi/pdf/10.1002/isd2.12066) 2018. ‘Input loans boost farmer take-up rates for satellite-based advisory service’. (https://www.cta.int/en/digitalisation/all/ UN Office of the Special Adviser on Africa. ‘Comprehensive article/input-loans-boost-farmer-take-up-rates-for-satellite-based- Africa Agriculture Development Programme (CAADP)’. Peace, advisory-service-sid0c75ed4b1-173d-4c1d-a899-be03866bd3f3) Security and Development. (http://www.un.org/en/africa/osaa/ peace/caadp.shtml) Technical Centre for Agricultural and Rural Cooperation. 2018. ‘Insurance: Incentive for Climate Resilience’. (http://www. United Nations Economic Commission for Africa. ‘ Towards cta.int/en/article/insurance-incentive-for-climate-resilience- improved access to broadband in Africa’. (https://www.uneca. sid079fde192-3613-406f-acbf-4fca6e3c4e62) org/publications/towards-improved-access-broadband-africa) Technical Centre for Agricultural and Rural Cooperation. United Nations Economic Commission for Africa. ‘National 2017. ‘A farmer centric mobile app’. (https://www.cta.int/en/ Assembly of Zambia - ICT Master Plan’. (https://www.uneca. article/a-farmer-centric-mobile-app-sid04085fa31-6f26-40be- org/publications/national-assembly-zambia-ict-master-plan) a46c-986b049be6e4) United States Department of Agriculture Economic Research Technoserve. 2015. ‘A fresh approach for smallholder coffee Service. ‘International Agricultural Productivity’. (https:// farmers’. (https://www.technoserve.org/blog/a-fresh-approach- www.ers.usda.gov/data-products/international-agricultural- for-smallholder-coffee-farmers ) productivity) The Conversation. 2018. ‘Why Rwanda’s development model USAID. 2018. ‘Policy Brief #5: ICT Solutions for Inclusive wouldn’t work elsewhere in Africa’. (http://theconversation. Agricultural Value Chains’. (https://www.agrilinks.org/sites/ com/why-rwandas-development-model-wouldnt-work-elsewhere- default/files/brief_5_-_ict_solutions_for_agricultural_value_ in-africa-89699) chains.pdf); The Economist. 2018. ‘Africa has plenty of land. Why is it so USAID. 2018. ‘Digitizing the Agriculture Value Chain’. hard to make a living from it?’. (https://www.economist.com/ (https://www.usaid.gov/sites/default/files/documents/15396/ middle-east-and-africa/2018/04/28/africa-has-plenty-of-land- Why_Where_and_How_Digital_Tools_Impact_the_Value_ why-is-it-so-hard-to-make-a-living-from-it) Chain.pdf) BIBLIOGRAPHY 237 USAID. 2018. ‘Using Digital Tools to Support Ghana’s Virtual City. ‘Agroforce’. (https://www.virtualcity.co.ke/ Smallholder Farmers: A Case Study of Advance II’. (https:// solution/agroforce-2/) www.usaid.gov/sites/default/files/documents/15396/ ADVANCE_II_Case_Study_2.pdf) Vota, Wayan. 28/12/2016. ‘Farmers Will Benefit Most from Mobile Financial Services’. ICT works. (https://www.ictworks. USAID. 2018. ‘Senegal: In-depth Assessment of Extension and org/farmers-will-benefit-most-from-mobile-financial-services/#. Advisory Services’, Developing Local Extension Capacity (DLEC) XG62kVVKipp) Project. (http://rescar-aoc.org/wp-content/uploads/2018/04/ DLEC-Senegal-In-Depth-Assessment-Extension-Final.pdf) Waruingi, Joseph and Muriithi, Erick. 2018. ‘Digital harvest’. Alliance for a Green Revolution in Africa. USAID. 2018. ‘Data Driven Agriculture: The Future of Smallholder Farmer Data Management’. (https://www.usaid. Wei, He. 07/08/2018. ‘AI tech boosting smart agriculture’. gov/digitalag/documents/data-driven-agriculture) China Daily. (http://www.chinadaily.com.cn/a/201808/07/ WS5b68f3f3a3100d951b8c8f62.html) USAID. 2018. ‘Digital Farmer Profiles: Re-Imagining Smallholder Agriculture’. Welteji, D. 2018. ‘A critical review of rural development policy of Ethiopia: access, utilization and coverage’. Agriculture (https://www.usaid.gov/sites/default/files/documents/15396/ & Food Security, 7(1), 55. (https://agricultureandfoodsecurity. Data_Driven_Agriculture_Farmer_Profile.pdf) biomedcentral.com/articles/10.1186/s40066-018-0208-y) USAID. 2018. ‘Digital Tools in Agricultural Programming’. Wiebe, K.D. et al. 2017. ‘The effects of climate change on (https://www.ictworks.org/wp-content/uploads/2018/04/ agriculture and food security in Africa’. International Food Programming-Toolkit-Digital-Tools-for-Agriculture.pdf) Policy Research Institute. (http://www.ifpri.org/publication/ effects-climate-change-agriculture-and-food-security-africa) USAID. 2018. ‘Where and How Digital Tools Impact the Value Chain’. (https://www.usaid.gov/sites/default/files/ World Bank Group. 2019. ‘Africa Smallholder Agriculture documents/15396/Why_Where_and_How_Digital_Tools_ Digital Disruption Conference Proceedings’ (full report Impact_the_Value_Chain.pdf). forthcoming). USAID. 2017. ‘Finding the best fit. One Acre Fund’s integration World Bank Group. 2019. ‘Future of Food: Harnessing Digital of digital tools in Kenya’. (https://www.usaid.gov/digitalag/one- Technologies to Improve Food System Outcomes’.(https:// acre-fund-case-study). openknowledge.worldbank.org/handle/10986/31565) USAID. 2017.’The future of microfinance for Kenya’s World Bank Group. 2019. ‘E-commerce for poverty alleviation smallholder farmers’. Musoni Kenya and Kilimo Booster. (https:// in China’. (http://blogs.worldbank.org/eastasiapacific/e- www.usaid.gov/digitalag/musoni-case-study) commerce-poverty-alleviation-rural-china-grassroots- development-public-private-partnerships) USAID. 2014. ‘Smallholders and Inclusive Growth in Agricultural Value Chains’. (https://www.agrilinks.org/sites/ World Bank Group. 2018. ‘Working with Smallholders’. A default/files/resource/files/FIELD%20Report%20No%2018%20 Handbook for Firms Building Sustainable Supply Chains. (http:// Smallholders%20in%20Value%20Chains.pdf) documents.worldbank.org/curated/en/284771480330980968/ pdf/110543-Handbook-Working-with-Smallholders.pdf). USAID. 2011.’iCOW Area of Focus: Livestock information’. (http://measict.weebly.com/uploads/3/2/4/3/3243215/icow- World Bank Group. 2018. ‘The Role of Digital Identification in aginfo-mobilephone-kenya.pdf) Agriculture: Emerging Applications’. (https://openknowledge. worldbank.org/handle/10986/31216) UTZ Certified. 2018. ‘Ghanaian farmers to benefit through new SAT4Farming program’. (https://utz.org/corporate-news/ World Bank Group. 2018. ‘Women, Agriculture and Work in ghanaian-farmers-benefit-new-sat4farming-program/) Africa’. (www.worldbank.org/en/programs/africa-myths-and- facts/publication/women-agriculture-and-work-in-africa). Verdier-Chouchane, Audrey and Karagueuzian, Charlotte, 2016. ‘Moving towards a green productive agriculture in Africa: The World Bank Group. 2017. ‘The Global Findex Database- role of ICTs’. Africa Economic Brief. African Development Bank. Measuring Financial Inclusion and the Fintech Revolution’. (https://www.afdb.org/fileadmin/uploads/afdb/Documents/ (https://globalfindex.worldbank.org/) Publications/AEB_Vol_7_i_7_-_Moving_towards_a_green_ productive_agriculture_in_Africa_The_role_of_ICTs.pdf) 238 BIBLIOGRAPHY World Bank Group. 2016. ‘Will Digital Technologies Transform World Cocoa Foundation. 2012. ‘Ghana Cocoa Board Agriculture in Developing Countries?’. http://documents. and World Cocoa Foundation Sign Agreement to worldbank.org/curated/en/481581468194054206/pdf/ Continue Successful CocoaLink Programme’. (http://www. WPS7669.pdf worldcocoafoundation.org/wp-content/uploads/09-2012- CocoaLink_COCOBOD.pdf) World Bank Group. 2017. ‘ICT in Agriculture e-Sourcebook’. (https://openknowledge.worldbank.org/bitstream/ World Cocoa Foundation. 2012. ‘Cocoa Market Update’. handle/10986/27526/9781464810022.pdf) Published Reports and Resources. (http://www.worldcocoafoundation. org/wp-content/uploads/Cocoa-Market-Update-as-of-3.20.2012. World Bank Group. 2014. ‘ICTs for agriculture in Africa’. pdf) eTransform AFRICA. (http://siteresources.worldbank.org/EXTIN- FORMATIONANDCOMMUNICATIONANDTECHNOLO- World Economic Forum. ‘Putting the New Vision GIES/Resources/282822-1346223280837/Agriculture.pdf) for Agriculture into Action: A Transformation Is Happening’. (http://www3.weforum.org/docs/WEF_FB_ World Bank Group. 2013. ‘Turn Down the Heat: Climate NewVisionAgriculture_HappeningTransformation_Report_2012. Extremes, Regional Impacts, and the Case for Resilience’. pdf) (http://www.worldbank.org/content/dam/Worldbank/ document/Full_Report_Vol_2_Turn_Down_The_Heat_%20 World Economic Forum. 03/05/2017. ‘The Future of Jobs and Climate_Extremes_Regional_Impacts_Case_for_Resilience_ Skills in Africa: Preparing the Region for the Fourth Industrial Print%20version_FINAL.pdf Revolution’. (https://www.weforum.org/reports/the-future- of-jobs-and-skills-in-africa-preparing-the-region-for-the-fourth- World Bank Group. 2012. ‘ICT in agriculture: connecting industrial-revolution) smallholders to knowledge, networks, and institutions’. (http:// documents.worldbank.org/curated/en/455701468340165132/ World Economic Forum. 2019. ‘Innovation with a Purpose: ICT-in-agriculture-connecting-smallholders-to-knowledge- Improving Traceability in Food Value Chains through networks-and-institutions) Technology Innovations’. (http://www3.weforum.org/docs/ WEF_Traceability_in_food_value_chains_Digital.pdf) World Bank Group. 2009. ‘Agriculture and climate risk enterprise (ACRE) Kilimo Salama—Kenya, Rwanda, Tanzania’. World Economic Forum. 2018. ‘Innovation with a Purpose: (http://documents.worldbank.org/curated/ The role of technology innovation in accelerating food systems en/858031490693709582/pdf/113740-BRI-PartnerProfiles- transformation’. (https://www.weforum.org/reports/innovation- ACRE-PUBLIC.pdf) with-a-purpose-the-role-of-technology-innovation-in-accelerating- food-systems-transformation) World Bank Group. 2008. ‘What are the links between agricultural production and food security?’, World Development World Health Organization. 2017. ‘WHO’s Africa Nutrition Report 2008. (http://siteresources.worldbank.org/INTWDRS/ Report highlights an increase in malnutrition in Africa’. Afro- Resources/477365-1327599046334/8394679-1327606607122/ WHO News. (https://www.afro.who.int/news/whos-africa- WDR08_07_Focus_C.pdf) nutrition-report-highlights-increase-malnutrition-africa) World Bank Group. 2003. ‘E-Choupal: ITC’s rural networking Xinxua. 2017. ‘China’s prominent techfin shares rural poverty project. Empowerment Case Studies. (http://siteresources.worldbank. alleviation lessons with FAO.’ (http://www.xinhuanet.com// org/INTEMPOWERMENT/Resources/14647_E-choupal-web. english/2017-07/15/c_136446325.htm) pdf) Yara International. 2015. ‘The Yara N-Sensor: Complete World Cocoa Foundation. 2013. ‘New CocoaLink Mobile Solution to Precision Farming’. (https://brandlibrary.yara.com/ Technology Program to Improve Livelihoods of 100,000 download.php?docid=doc6khw1fmpfqdz78id08&variant=master Cocoa Farmers in Côte d’Ivoire’. PR Newswire. (https://www. &type=original) prnewswire.com/news-releases/new-cocoalink-mobile-technology- program-to-improve-livelihoods-of-100000-cocoa-farmers-in-cote- divoire-209342621.html) Allan Rickman The Technical Centre for Agricultural and Rural Cooperation (CTA) PO Box 380 6700 AJ Wageningen The Netherlands www.cta.int