Longitudinal Analysis

Kenya, Ethiopia, and Rwanda have been developing simultaneously along distinctly different paths. Comparing them is analytically productive precisely because they are at different stages of the same journey - and because the differences illuminate what drives ecosystem development rather than what merely accompanies it.

Kenya is the continent's deepest startup market by experience and institutional infrastructure, representing approximately 52 percent of all startups in Eastern Africa. In 2025, Kenyan startups raised $984 million - approximately one third of all African startup funding, the strongest single-market performance since 2022. The composition matters as much as the size: 60 percent was debt financing, driven by large infrastructure deals in off-grid energy from d.light, Sun King, M-KOPA, Burn, and PowerGen. Equity funding nearly doubled to $383 million, but the deal count among ventures raising $100,000 or more fell 23 percent - the weakest performance among Africa's Big Four. Substantive treatment of the post-correction Kenya capital landscape sits in Investor Landscape and The Financing Journey Through the Scaling Phases. Depth has not produced breadth. The $5–20 million growth stage remains severely underserved. A capital structure built on external equity produced a market that contracted sharply when external equity withdrew - and recovered in 2025 primarily through debt instruments and energy infrastructure that look quite different from the tech venture model Kenya is usually described through.

Rwanda represents the opposite construction: an ecosystem shaped by sustained government will rather than market dynamics. The State of Innovation Rwanda 2026 documents ecosystem value reaching $394 million in 2024 and total startup employment growing from approximately 700 jobs in 2018 to nearly 10,000 by 2024. Startups raised $38 million in 2024, nearly six times higher than in 2022.

Stephen Brien's April 2026 analysis gives the fiscal architecture behind that divergence with unusual precision. Rwanda sustained public investment at approximately 11 percent of GDP through successive external shocks across the 2010s and into the early 2020s. Kenya's debt service payments came to exceed its development expenditure - interest absorbing around 5 percent of GDP against capital spending of roughly 3.5 percent. The ecosystem consequences of that divergence are downstream of a governing orientation that was either built or was not. The substantive treatment of why governing orientations rather than programmatic interventions produce sustained ecosystem capability - anchored across Acemoglu-Robinson on extractive vs inclusive institutions, Whitfield-Therkildsen-Buur-Kjær on the politics of African industrial policy, and Mkandawire on African developmental states - sits in The Political Economy of Capital Allocation. Rwanda is also one of four African success cases analysed in Studwell's How Africa Works (Atlantic Monthly Press, 2026), which identifies the developmental political coalition as the additional ingredient that distinguishes African early-success cases from those whose policies alone have not produced sustained transformation. The substantive treatment of Studwell's framework as it applies to this publication sits in Approach and Methodology.

That configuration is also Rwanda's vulnerability: ventures remain heavily dependent on government contracts and government-facilitated foreign investment, and the political will underpinning the model has required a continuity of leadership that is not universally replicable. The State of Innovation Rwanda 2026 documents a pronounced missing middle - most firms remain in the sub-$10 million valuation range, with very few reaching the $10–20 million tier where sustainable scaling typically begins.

Ethiopia is the largest frontier market by population with the least developed formal venture ecosystem, attracting less than 1 percent of Africa's startup capital despite being the continent's second-largest country by population. The structural challenges are well understood: limited formal finance, uneven market access, and weak institutional infrastructure for venture development. The national startup and scaleup strategy, the situational assessment, and the Dealroom-powered data platform developed through Systemic Innovation's FCDO-funded engagement represent an attempt to build institutional infrastructure before the ecosystem reaches the scale at which its absence becomes costly - rather than after, as was the case in Kenya and Nigeria.

Drawing on longitudinal data from all three markets, and informed by the entrepreneurial ecosystem literature, a four-phase model of East African ecosystem development is emerging. It is a description of observed patterns, not a prescription - the path is shaped by local attributes, political economy, and institutional history in ways that no universal model can fully capture.

The four-phase model engages three foundational research programmes in entrepreneurial ecosystem development.Mack and Mayer's Urban Studies paper "The evolutionary dynamics of entrepreneurial ecosystems" - established the canonical four-stage evolutionary model (birth, growth, sustainment, decline) applied to North American technology ecosystems, naming the dynamic feedback loops through which each phase produces the conditions for the next. Brown and Mason's Small Business Economics paper "Looking inside the spiky bits" - established the typology distinction between embryonic and scale-up ecosystems, identifying that the analytical attributes that mark a productive ecosystem in early phases differ substantively from those that mark productivity in mature phases. Roundy, Brockman and Bradshaw's Journal of Business Venturing Insights paper “The resilience of entrepreneurial ecosystems" - established that entrepreneurial ecosystem resilience operates not as a property of individual ventures but as an emergent characteristic of the system's collective adaptive capacity. The substantive treatment of Stam's entrepreneurial ecosystem framework - the foundational scaffolding for both the Mack-Mayer evolutionary work and the broader ecosystem development literature - sits in Ecosystem Characteristics. The four-phase East African model below operates in dialogue with this scholarship while taking seriously the contextual distinctions that anchor Rationale: the conditions, mechanisms, and timelines that shape African ecosystem development are not contextual variants of the North American or European patterns documented in the foundational scholarship.

The pioneer phase

Lasting roughly five to eight years from first significant tech venture activity, establishes the foundations on which everything else is built: a small number of ventures constructing foundational infrastructure, a tight informal network of founders and early investors, minimal external capital, heavy reliance on diaspora founders and returnees, and almost no formal ecosystem support. Kenya experienced this roughly between 2007 and 2015. Rwanda approximately from 2010 to 2017. Ethiopia from approximately 2015 to 2021. The pioneer phase is underappreciated in ecosystem development literature because it produces few visible outputs - Brown and Mason's embryonic-ecosystem typology captures part of this but understates the institutional-infrastructure-building work that distinguishes durable pioneer phases from those that fail to transition. The substantive treatment of why diaspora founders and returnees are particularly consequential at this phase - anchored in Saxenian's The New Argonauts on cross-regional entrepreneurial networks - sits in Ecosystem Characteristics. It is in this phase that the most consequential ecosystem characteristics are established - the culture of experimentation, the norms around collaboration and knowledge sharing, and the foundational infrastructure on which activation depends.

The activation phase

Lasting roughly three to six years, brings rapid growth in venture numbers and external capital, the emergence of formal ecosystem support infrastructure, growing international investor interest, and government engagement with innovation as a policy priority. Kenya roughly from 2015 to 2020. Rwanda approximately from 2017 to 2022. Ethiopia is entering this phase now. The programme-rich, capability-thin dynamic typically begins here: external capital creates demand for programme infrastructure faster than endogenous capability can develop, producing an ecosystem that appears well-resourced but is structurally dependent on the continuation of external funding. Substantive treatment of this dynamic sits in (Stalled) Acceleration and The Political Economy of the Ecosystem.

The maturation-under-stress phase

 lasting roughly two to five years, exposes the structural weaknesses accumulated during activation: capital contraction forces significant venture failure, consolidation, and M&A activity; donor-dependent programme infrastructure thins as programmes close; the ecosystem faces its first real test of endogenous resilience. Kenya has been in this phase since approximately 2022. Rwanda entered it slightly later. The Roundy-Brockman-Bradshaw framework on entrepreneurial ecosystem resilience operates structurally here: the correction has provided a natural sorting mechanism that is painful and productive simultaneously, exposing whether the ecosystem's collective adaptive capacity exists to absorb shocks or whether resilience was an attribute of individual well-capitalised ventures rather than the system. M&A activity rose to a record 67 deals across the continent in 2025 - a 72 percent increase year on year - as ventures consolidate and capital concentrates in survivors with demonstrable unit economics.

The fourth phase - endogenous consolidation

Is emerging in Kenya and beginning to be visible across East Africa. AVCA's 2025 report documents domestic investor participation reaching a new high: African investors accounted for 45 percent of total venture fund commitments in 2025, up from an average of 23 percent between 2022 and 2024. African DFIs contributed 63 percent of DFI capital deployed. Seven African funds ranked among the top ten most active investors by deal count, led by Launch Africa Ventures with 14 deals. This is the most important longitudinal finding from the correction period: the resilience that sustained the ecosystem through the contraction came primarily from local capital and endogenous capability. African-headquartered investors did not flee when market conditions turned hostile. That distinction - between capital that stays and capital that withdraws - is the empirical core of the argument for building endogenous capability rather than dependence on external programme delivery. Substantive treatment of the post-correction investor landscape and the structural shift toward domestic capital sits in Investor Landscape.

The data across the three cases suggests four consistent tipping point triggers.

The first is a founding infrastructure event that changes the economics of digital venture building: M-PESA's maturation as a payments layer in Kenya around 2010–12, government broadband infrastructure investment in Rwanda, and the liberalisation of Ethiopia's telecommunications sector alongside Safaricom's market entry. Infrastructure events do not cause ecosystems but they determine what becomes commercially viable within them.

The second is a first-wave exit or valuation milestone that demonstrates to the broader investor community that the ecosystem can produce investable outcomes. Paystack's acquisition by Stripe in 2020 was a watershed exit for the African tech ecosystem - not because of the capital it generated directly but because of the credibility it conferred on the continent's startup market. Rwanda has not yet experienced an equivalent milestone of its own, which partly explains why Rwandan ventures remain more government-dependent than Kenyan ones.

The third is a governance crisis or market correction that forces the ecosystem to develop more durable norms and institutions. The 2022–25 correction has played this role for East Africa broadly. Events of this kind are painful and often accelerate the institutional development that would otherwise take longer. The ventures that survived the maturation-under-stress phase are, on the evidence available, better businesses than those that did not: Africa: The Big Deal documents that the 2025 recovery saw 215 startups raise $1 million or more and 69 exceed $10 million, with megadeals concentrated in ventures with clear unit economics and credible paths to profitability.

The fourth is the emergence of a critical mass of second-generation founders who invest in and mentor the next generation. In Kenya, the growing presence of Paystack, Andela, and Flutterwave alumni as angel investors and early-stage mentors represents this transition. The substantive treatment of why second-generation founder networks operate as the structural backbone of ecosystem capability - anchored across the Endeavor Network Effect research and Burt's structural holes framework - sits in What Working Support Infrastructure and Founders and Leadership Teams. Endeavor Insight and Endeavor Kenya's March 2026 mapping of the Kenyan entrepreneurship network confirms this dynamic directly: founders are paying forward through mentorship, angel investments, and second-generation company creation - and that multiplier effect is the real engine of Kenya's entrepreneurship network, more consequential for long-run ecosystem development than any accelerator cohort or challenge fund.

Phase-appropriate interventions matter more than the quality of any individual programme. The acceleration model the ecosystem relies on is poorly calibrated to the pioneer phase - it assumes a pool of activation-phase ventures that typically does not yet exist. Interventions that work well in activation - cohort-based acceleration, demo days, international investor matchmaking - are often irrelevant or actively counterproductive in the pioneer phase, where what is needed is foundational infrastructure, informal network development, and patient capital with long enough time horizons to allow the ecosystem to develop its own institutional logic. The substantive treatment of why programme-rich, capability-thin equilibria persist across activation phases sits in (Stalled) Acceleration and An Honest Appraisal.

The gap between Ethiopia and Kenya in ecosystem maturity is an opportunity, not merely a challenge. Rwanda has demonstrated that sequencing institutional development ahead of capital concentration can produce a more managed ecosystem trajectory than the capital-led, institution-follows model that characterised Kenya's development. Ethiopia has the opportunity to make similar choices. The Ethiopian Data Foundations study - substantive treatment in A Theory of Ecosystem Change - provides the four Systemic Growth Metrics framework for tracking whether those choices are generating the firm-level outcomes reform is intended to produce. Whether the institutional bandwidth and political will exist to use it is a question the donor funding contraction has made considerably harder to answer.