Future Research Agenda
Knowledge has limits
Knowledge has limits. Honest acknowledgment of those limits is as important as confident assertion of what is known. The FCDO RISA Fund-funded EADC programme produced the most comprehensive empirical record of high-growth firm dynamics in East Africa that now exists. It does not exist for West Africa. It does not exist in depth for Southern Africa beyond South Africa itself. It does not exist for Francophone markets. The architecture of what is known is heavily East African - and the prescriptions that emerge from an East African evidence base may not translate cleanly to Lagos, Accra, Dakar, or Johannesburg.
This publication acknowledges these gaps directly and names them as the founding research priorities of the African Scaleup Lab. They are not gaps waiting for a funder to notice them. They are the specific knowledge deficits that motivated the Lab's design - and that the Lab's ten-year mandate is structured to close.
The academic framing of this agenda has been developed in collaboration with various partners, building on the Ethiopian Data Foundations study and the Africa Entrepreneurial Ecosystem Index. A detailed research pathways note - covering ten interconnected priorities across firm dynamics, ecosystem diagnostics, policy evaluation, and scaling evidence has been developed as the analytical foundation for the Lab's academic partnership programme. What follows is the strategic summary for practitioners and policy audiences.
West African ecosystem dynamics
Nigeria's ecosystem is Africa's largest by funding volume but is almost entirely uncharted at the level of high-growth firm operational dynamics. The EADC model - state of ecosystem reports, scale diagnostics, venture case studies - needs a West African equivalent. The operational conditions that produce those ventures, and the structural barriers that prevent more of them, are empirically undocumented. The field's most significant market is also its largest analytical blind spot.
The francophone West African context - operating under OHADA legal infrastructure, different banking relationships, and a distinct regulatory tradition - requires not just geographic extension of the East African framework but a structurally different analytical approach. That is the Lab's first major geographic research priority.
Failure intelligence
The absence of documented failure cases is the most consequential gap in the African scaling knowledge base. What failed, why, at what stage, under what conditions and what the founders did next. Disrupt Africa's 2024 report documents 200 funded startups down from 633 at the peak without providing causal analysis of the ventures that disappeared between those numbers. That analysis is the most important data the ecosystem does not have. A systematic failure intelligence programme - modelled on the Fuckup Nights movement but anchored in rigorous case documentation - would be one of the highest-impact knowledge investments the ecosystem could make. That none exists, three years after the most consequential correction the ecosystem has experienced, is itself evidence of the Capability Trap: the infrastructure for honest accountability is precisely what the current incentive structure cannot produce.
Enabling Condition 2 - the structured African failure intelligence commission - addresses the immediate funding question. The Lab holds and builds the cumulative dataset, providing the longitudinal architecture that a single foundation grant cannot sustain on its own. The academic research agenda extends this into firm stalling and decline dynamics more broadly: identifying the triggers, thresholds, and conditions that cause ventures to plateau or contract, and developing the early-warning indicators that would allow investors and policymakers to intervene before firms exit the growth trajectory rather than after. This is among the most underresearched dimensions of African firm dynamics - most research focuses on high-growth firms, ignoring the far more common pattern of stalling at early or mid-growth stages.
The missing middle: firms at 5–50 employees
The scaling ecosystem's analytical attention concentrates at the extremes: micro-enterprises at one end, tech startups at the other. The segment in between - firms with 5 to 50 employees, too large for microfinance and too small or undigitalised for venture capital - remains analytically invisible despite being central to job creation and productivity growth. This is the segment where structural transformation actually happens, and where the Capability Trap operates most acutely: the firms most capable of productive growth face the highest effective frictions.
The Scale Diagnostics work conducted across 42 ventures in Kenya, Ethiopia, and Rwanda provides the most granular primary data currently available on what African scaling ventures look like from the inside. Extending this methodology across a larger sample, integrating firm-level data with administrative records - registry, tax, labour data - and combining quantitative analysis with structured qualitative interviews would produce the first systematic evidence base on the operational dynamics of the missing middle across multiple African markets. This is among the highest-priority research investments available to BII, IGC, and the governments whose SME strategies currently operate without this evidence.
Platform scaling in low-trust, high-friction markets
The platform economy analysis identified earlier are a significant analytical gap: how do two-sided and multi-sided marketplaces build liquidity in markets characterised by low institutional trust, high transaction friction, and limited digital payment infrastructure? A systematic comparative study across platform archetypes - covering the informal-formal interface, the agent network layer, and the trust-building mechanisms that substitute for institutional enforcement - would provide the analytical foundation for a more targeted approach to platform support.
The correction period produced a natural experiment in which platform models were stress-tested at scale: the B2B e-commerce thesis failed comprehensively; the vertically integrated physical-digital model survived. The causal analysis of why has not been conducted. It is among the most commercially valuable research the Lab could commission - and one of the clearest cases where the evidence generated would directly improve capital allocation decisions by investors currently operating on intuition rather than documented pattern.
Gender-capital intersection mechanisms
The gender funding gap exists, is worsening, and is documented with unusual precision. The capital allocation is moving in the opposite direction of the performance evidence.
The five structural mechanisms identified are a framework, not a settled explanation. They need empirical testing - not just logical plausibility - to determine which are primary drivers and which are secondary amplifiers, and where intervention would generate the greatest structural effect per unit of resource deployed.
The Growth Firms Alliance's WSME segmentation framework provides the most rigorous foundation for this research. The Lab's gender-capital research priority is to test the five mechanisms empirically against that framework across East and West African markets producing the causal analysis that the descriptive evidence base now demands.
Platform scaling in low-trust, high-friction markets
As AI tools proliferate and costs fall, the question of who captures the productivity benefits - and who does not - will become the most consequential distributional question in the African scaling ecosystem within the next five years. Headline infrastructure figures do not describe the distributional pattern within African markets - which founders access AI capability, which do not, and what determines the difference.
Building that measurement framework now, before the adoption curve accelerates, is the difference between understanding the distribution after the fact and being able to influence it in real time. The Lab's AI research priority is not to produce another assessment of Africa's AI readiness that literature is already substantial. It is to produce the first longitudinal study of how AI capability access distributes within African scaling ventures: which sectors, which markets, which founder profiles, and which capital structures are associated with genuine AI-enabled competitive advantage rather than AI-labelled marketing.
AI adoption and capability distribution
As AI tools proliferate and costs fall, the question of who captures the productivity benefits and who does not will become the most consequential distributional question in the African scaling ecosystem within the next five years. Africa accounts for less than 1 percent of global data centre capacity, per the Africa Data Centres Association. Only 5 percent of African AI innovators have reliable access to advanced compute, according to UNDP analysis. These headline figures describe the infrastructure deficit. They do not describe the distributional pattern within African markets which founders access AI capability, which do not, and what determines the difference.
Building that measurement framework now, before the adoption curve accelerates, is the difference between understanding the distribution after the fact and being able to influence it in real time. The Lab's AI research priority is not to produce another assessment of Africa's AI readiness that literature is already substantial. It is to produce the first longitudinal study of how AI capability access distributes within African scaling ventures: which sectors, which markets, which founder profiles, and which capital structures are associated with genuine AI-enabled competitive advantage rather than AI-labelled marketing.
The Lab's founding research programme
These six priorities - West Africa, failure intelligence and firm stalling, the missing middle at 5–50 employees, platform scaling, gender-capital mechanisms, and AI distribution are not independent research questions. They are the interconnected gaps in the East African evidence base that prevent the analytical frameworks in this publication from being applied confidently beyond the three markets in which they were developed.
The detailed academic research agenda - covering ten specific pathways including cross-country firm dynamics replication, administrative data architecture, policy-to-firm outcome linkage, and comparative ecosystem indexing across Africa and Asia - has been developed with the IGC and BII networks and is available from Systemic Innovation on request. The Lab's founding research programme draws on that agenda selectively: the six priorities above are those where the evidence gap is most consequential for practitioners and policymakers, and where the Lab's applied research model - working directly with ventures, governments, and capital providers rather than through academic journals - generates the most distinctive marginal contribution.
The research agenda is not a wish list. It is a founding document. The difference is that a wish list requires someone else to act. A founding document is the institution acting.