As the AgriFin Accelerate (AFA) program concludes and we gear up for AgriFin 3.0, we engaged an independent reviewer to evaluate the program and gauge AFA’s influence at the client, institutional and ecosystem level. This is an impartial opinion of the program.

>> Download this Ecosystem Evaluation Brief here << 


The AgriFin Accelerate (AFA) program is a USD 24.7 million program of Mercy Corps, supported by the Mastercard Foundation that ran from 2015-2021. AgriFin Accelerate was designed to narrow the financial “inclusion gap” for smallholder farmers (SHFs) who regularly lack access to affordable, accessible, demand-driven financial products and services that drive higher productivity and income for farm families. 

The AgriFin Accelerate (AFA) Final Ecosystem Review and Program Evaluation was designed by Causal Design to answer a series of evaluation questions that gauge AFA’s influence at the client, institutional and ecosystem level as the program concludes. It combines an extensive document review with key informant interviews through a case study design to yield nuanced findings that were mapped to the evaluation questions.

The main findings, discussion and recommendations of this evaluation relate to the ecosystem and institutional level findings, and thus broadly how and to what extent AFA was able to support a digital financial services (DFS) and digital information services (DIS) ecosystem in three countries (Kenya, Tanzania, and Zambia), how this influenced partners and to what end, and as best as possible how the ecosystem and institutions were responsive to SHF needs and priorities. This evaluation is unique in its wide-angle view of AFA and its focus on the ecosystem and key actors within it, and thus uniquely complements the array of research, evaluation and learning products produced and made available by Mercy Corps AgriFin.


The report asserts that AFA was essential in brokering critical business partnerships, generating actionable knowledge, and ultimately laying the foundation for a series of commercial DFS and DIS services that are being increasingly adopted by smallholder farmers (SHFs) in these three countries. Fundamental to this success was AFA’s active, ground-level engagement with partners that resulted in new and emerging business relationships, and effective bundled services, very much as the project intended.

The evaluation report identifies several key findings that address multiple evaluation questions, while also providing nuanced insights into the ecosystem and partnerships as a whole include:

  • Theory of Change assumptions that need revision:
    • including how SHFs are defined, and thus targeted (with implications for investment “leakage” and unintended beneficiaries);
    • the potential role of government, and crucially what an ecosystem consists of, with implication for how AFA and future iterations focuses its efforts;
  • AFA’s relevance to SHFs: Relevance is bolstered by AFA’s mainstreaming of human-centered design (HCD) research, and its commitment to effective partnerships that ensure alignment with objectives and relevant services;
    • Some partnership challenges, mostly revolving around project management complications, and differing technical capacities or incompatible platforms for sharing data, code, etc.
  • The importance of learning and adapting, both by AFA and partners, and its ecosystem effects;
    • AFA demonstrated a steadfast commitment to developing learning products, and sharing this information transparently, which had a clear, marked effect on partners, partner capacity, and in nascent programming, especially evident in Zambia.
  • The significance of field agents as components of the ecosystem, and their potential catalytic role in deepening the client experience.


There are myriad recommendations woven throughout the report which will require reflection and thought, by the AFA team, advisory council, and partners. These recommendations were not put forth lightly or naively, but instead reflect the challenge of a program like AFA attempting to affect an ecosystem—a process which can be accomplished in different ways, but which also includes much embedded uncertainty. The report suggests the following (which would potentially guide how to weigh other recommendations as well):

  • A structured review of the AFA Theory of Change, and the results framework itself: This could be conducted while simultaneously reviewing the wealth of research and evaluation products that AFA has produced, to gauge which assumptions have held up over the life of the project and which have not.

This process may force a rethinking of how AFA, in future iterations, can best affect an ecosystem.Its learning agenda, learning-focused events, and mentoring of partner institutions, were universally approved by partners of all kinds. Between AFA’s advantageous position as a font of DFS/DIS, agriculture, and rural economy expertise, and the challenges inherent in developing discrete products and services, this report suggests that AFA may be best positioned as a sectoral leader, defining the rules and norms that govern the emerging agri-tech space.

  • Mercy Corps AFA may also prove a global leader in defining and establishing what a “conducive environment” requires for autonomous DFS/DIS development, whether in terms of regulations, enforcement, infrastructure and education. It can accomplish this moreover, while maintaining its engagement, if differently, at each level of analysis, whether ecosystem, institutional/partner, and client levels.
  • More granularly, AFA can continue to examine with rigor how to close the gap between registration among clients, and regular, advanced use of DFS/DIS tools and services. HCD research has been crucial in contextualizing farmer experiences, but the registration-active user gap remains a challenge, suggesting a knowledge gap that AFA is well positioned to address.
  • This report suggests using a broader set of analytical tools, in addition to extant HCD research, to yet more deeply understand how clients perceive, interact with, or are somehow prevented from more deliberately engaging with services and products ostensibly designed for their benefit. This same approach will be necessary to better unpack the gendered dimensions of DFS/DIS uptake as well.

Each of the highlighted themes is further unpacked into sub-components, and substantiated with evidence, underscoring their significance to the evaluation, findings, and recommendations in the Full Ecosystem Evaluation Report.

Our Director of Impact and Assessment, Collins Marita dove deeper into the report with a panel of technical experts during our 5th Annual Learning Event in April 2021. Watch the session.

>> Download this Ecosystem Evaluation Brief here <<