Acceleration Goals

The 15 goals below were developed from the 2022 research to guide the evolution of African accelerators. Three years of ecosystem observation have validated their analytical direction while sharpening what each demands in practice. They are presented here not as a static 2022 proposal but as a living framework - some goals have been vindicated by subsequent experience (the rigorous evidence base now supports them more strongly than it did in 2022), some have been partially vindicated (direction right, specifics updated), some have been revised by experience (new evidence has changed what the goal looks like in practice), and AI has materially changed what several of them now make possible.

The four-cluster structure - strategic, design, delivery, and management - remains sound.

A note on the analytical foundation. The substantive evidence base on accelerator effectiveness is treated in (Stalled) Acceleration - the academic literature (Hochberg, Cohen/Bingham/Hallen, Gonzalez-Uribe & Leatherbee, Yu), the GALI ten-year synthesis, the WDI East Africa study, and the development-economics evidence on what capability-building actually produces (Bloom et al's Indian textile RCT, McKenzie/Iacovone/Maloney's Colombian group-consulting follow-up). The 15 goals here are the operational form of what that evidence base supports. They are not free-floating prescription; they are the design implications of the rigorous evidence base, calibrated to African operating conditions.

A second analytical clarification matters. The 15 goals divide structurally between two categories. Programme-design goals (5, 6, 7, 8, 10, 11, 12, 13) are within ESO operational control - implementing them is a question of programme architecture and staff capability. Structural goals (1, 2, 3, 4, 9, 14, 15) require donor-funding architecture reform to deliver - implementing them depends on changes ESOs cannot unilaterally make. The implementation strategy for each category is different. Conflating them produces frustration on both sides: ESOs blamed for failing to deliver structural goals their funding architecture prevents, and donors absolved of responsibility for design failures their architecture has produced.

1. Develop Africanised universal principles for acceleration, then localise

Status: vindicated.

The post-2022 correction made this point empirically. Ventures accelerated on growth-first, profitability-later frameworks were among the most exposed when capital contracted. Silicon Valley acceleration principles assume conditions - abundant follow-on capital, rapid market expansion, tolerance for cash burn - that African operating environments do not reliably provide. The Argidius/Porticus ten-year review is explicit: there is no one-size-fits-all recipe for effective acceleration, and the Silicon Valley blueprint specifically fails in ecosystems with thin early-stage capital. Principles first, then models, then localisation. Delivery models must be locally led - relying on the resources, experience, and expertise of in-market teams, even when linked to global entities.

2. Move from general to specialised acceleration

Status: vindicated.

Compared to an edtech venture, a fintech will have a very different trajectory, needs, and set of hurdles. Specialised acceleration yields materially better results because programme support can be designed around the unique challenges of a specific industry. The correction period reinforced this: the ventures that demonstrated resilience were those with deep sector expertise and specialised networks, not those optimised for the generic investor pitch.

"Having accelerator programmes focused on a particular sector or industry creates the capacity to build some level of specialisation tailored to the unique needs of a particular sector." - interviewee

Sector and stage specialisation is the prerequisite for the selection rigour that makes acceleration effective. Villgro Africa (healthcare), the Kenya Climate Innovation Center, and the Kigali Agribusiness Incubation Center demonstrate that depth of sector knowledge produces markedly better venture-support quality than breadth of participant numbers.

3. Provide a range of acceleration and scale up support models to increase reach

Status: partially vindicated

The more localised, bespoke, and hands-on the acceleration experience, the greater the impact - but this approach is expensive and selective by nature. A functioning ecosystem needs the full range. Bespoke scale-up support - top-end, best-in-class, reserved for ventures with demonstrated high-growth potential. Open Capital Advisors' fee-for-service model and Endeavor's founder-network model represent the two most developed African examples. Both are expensive and cannot be the only offer.

• Specialised accelerators - focused on specific verticals, with ventures participating at different stages of growth. The geographic concentration documented in Spatial Scaling Dynamics is reproduced in the support infrastructure: specialised accelerators must be deliberately designed for geographic reach - not merely tolerate remote participation as an afterthought.

• AI-assisted scale-up learning - the online scale-up academy model proposed in 2022 is now more achievable and more urgent. AI-powered learning platforms can deliver algorithmically personalised curricula based on assessed individual business needs, enabling ventures to pull the knowledge they need when they need it rather than receiving a fixed curriculum on a fixed schedule. The cost barriers that made this model aspirational in 2022 have fallen substantially, and the WDI East Africa study finding - that entrepreneurs are willing to pay US$150–200 for generic support but significantly more for targeted services with guaranteed outcomes - suggests a viable commercial architecture is now within reach.

• Scaling knowledge centre - a one-stop-shop resource, funded through development or philanthropic capital, serving ventures least likely to secure investment and formal support. SNDBX in Nairobi - which brings together a permanent residency of more than 30 professional disciplines under one roof, connecting entrepreneurs directly to vetted specialists in finance, legal, HR, tax, marketing, and governance - represents one practical architecture worth adapting at greater scale.

Highly localised incubators and hubs - these continue to serve local entrepreneurs in the earliest stages of business development. Incubation has its role. But these entities are not appropriate vehicles for supporting scaling ventures and should be firmly dissuaded from labelling their support as scaling offers. The label matters: it misleads founders about what they will receive and inflates accelerator participation statistics without changing outcomes.

4. Ensure accelerators are fully integrated into the wider scaling ecosystem

Status: vindicated.

Entrepreneurs should be able to move fluidly through the support ecosystem as they grow, with accelerators connecting effectively to locally active investors, pre-incubation services, and the broader infrastructure of scaling support. The GALI synthesis is specific on this: the quality of the network an accelerator is embedded in is a critically important part of the offer - not a bonus.

In ecosystems with limited early-stage funding, GALI's guidance is direct: participation in an accelerator alone is unlikely to alleviate capital constraints, and programmes must either provide funding directly or demonstrate concretely how they connect ventures to investors. Donor-funded accelerators that operate as islands - dispensing support without connection to the broader ecosystem - not only underperform but actively distort it by consuming resources that could flow to more integrated models.

5. Move away from one-size-fits-all approaches

Status: vindicated

"They don't really think deeply about individuals, they think of them as batches: cohort one, cohort two." - interviewee

AI-powered personalisation now makes this goal practically achievable at scale. Accelerators that continue to deliver fixed curricula to mixed cohorts are not choosing simplicity - they are choosing obsolescence. The Argidius SCALE framework - five evidence-based characteristics that distinguish impactful BDS from ineffective provision: Select the right enterprises, Charge enterprises (it improves performance), Address problems, Learn by evaluating, and Lead by example - makes the case for selection and problem-focused support. The most effective programmes identify what each enterprise specifically needs and address those needs directly. The GALI ten-year synthesis adds the architectural evidence: accelerators that use a flipped curriculum - giving entrepreneurs more time to work independently or in peer interaction - outperform those using classroom-based instruction. The support infrastructure should be designed around the venture's problem, not the programme's convenience.

6. Co-create programmes

Status: vindicated

Best practices in programme delivery include co-creating programmes with stakeholders and developing incentives aligned with venture needs. Co-creation with local communities ensures content is tailored to the realities of their markets, not the realities of whoever designed the programme. The WDI East Africa study finds that donor-imposed programme design - where international funders with limited local context determine what local entrepreneurs need - is one of the most consistent drivers of the generic, low-value content that produces training fatigue. The DGGF systemic review identifies stakeholder collaboration as a leverage point with outsized impact: facilitating cross-sector co-investment models and joint programme development unlocks local resources and amplifies impact in ways that donor-designed delivery cannot. AI-assisted needs assessment can now generate much richer diagnostic data about what a specific venture needs than a standardised intake questionnaire - removing one of the practical barriers to genuine co-design.

7. Sequence support and provide the right support at the right time

Status: vindicated.

Scale-ups are better served when they pull the knowledge and training they need at the time they need it, rather than having a fixed curriculum pushed at them. Batching is inefficient unless the entire cohort genuinely needs the same support simultaneously - which is almost never true of a mixed-stage cohort.

"The content is not what makes any of these programmes special. The value is more about the advice of when you need to do it, and it's the mentorship and the guidance. That's really the IP of a startup support programme." - interviewee

The WDI East Africa study confirms the timing failure in stark terms: generic cohort programmes overemphasise pitch deck preparation at the expense of substantive support for product-market fit, operational efficiency, leadership, and team building - delivering the wrong content at the wrong stage to ventures with fundamentally different needs. The Sustain Impact report documents the same problem at the organisational level: programme funding structures lock ESOs into predetermined delivery schedules that prevent real-time adaptation to venture needs. Sendemo's 2024 fieldwork estimates that fewer than 20 SSOs across the continent - roughly 2 percent of the approximately 1,200 operating - have a demonstrated track record of supporting high-growth ventures, concentrated in nine markets (South Africa, Nigeria, Egypt, Kenya, Tunisia, Morocco, Ghana, Senegal, Côte d'Ivoire). The ecosystem is producing SSO volume without producing SSO capability - which is what a structural misalignment between funding logic and outcome logic predicts.

8. Prioritise programme deliverables over process (outcomes, not outputs)

Status: vindicated, evidence base substantially strengthened.

Create clear deliverables - not processes. Set clear targets for the business's stage and then offer an array of tools to get there. The Y Combinator and TechStars models work because they are designed for founders raising similar amounts of capital at similar stages who need similar support. That model does not work in the African context, where variance across cohorts is far greater and the gap between programme completion and investability is wider. The WDI East Africa study documents that donors and ESOs continue to define success by graduating companies' revenue growth, jobs created, and capital raised - metrics that are difficult to attribute to ESO intervention and that Sankalp Africa 2025 session attendees identified as problematic given high startup failure rates. Alternative success measures - quality mentorship, strong network connections, community engagement, founder satisfaction, and long-term startup survival - are harder to count but more honest about what acceleration actually produces. This goal connects directly to Goal 15 on measurement: without outcome data, there is no way to distinguish programmes that produce outputs from programmes that produce results.

9.  Provide post-programme support as a core function, not an afterthought

Status: vindicated.

Accelerators were designed in thriving Western startup ecosystems - a three-month slice preceded and followed by abundant funding and support. The African ecosystem does not have the support infrastructure before, or competitive funding after, the programme to sustain this model. The ventures that most need post-programme support are precisely those without the networks and resources to access it independently. Post-programme engagement - systematic follow-up, alumni network activation, ongoing investor introductions - should be a core programme function with dedicated resource, not an optional extension that disappears when the grant ends. The WDI East Africa study identifies an emerging model worth scaling: alumni mentoring cohorts in lieu of fees - a structure that simultaneously addresses post-programme connection, builds the mentorship supply the ecosystem needs, and begins establishing the exchange-of-value relationships that can eventually replace free provision.

The same diagnosis is being arrived at independently from the climate-adaptation field. The In-Between Spaceslearning paper (Wasafiri/Climate KIC, April 2026), produced through a community of practice convened by the Systems Innovation Learning Partnership and the UNIDO-GEF Climate Adaptation Innovation & Learning initiative with around 25 intermediary practitioners, identifies the polarity between project delivery and long-term system learning as one of six core tensions defining intermediary work. Adaptation outcomes, the paper observes, "unfold over much longer horizons" than the 2–3 year project cycles through which the work is funded, yet "relationships, trust and insight often extend well beyond the life of a grant… even when funding does not." The relational and learning labour that sustains ecosystems across funding cycles is, in the paper's words, "difficult to fund, measure or currently legitimise" — a structural mismatch between how intermediary value is created and how it is commissioned. The convergence with the WDI East Africa findings and the GALI synthesis is itself diagnostic: the post-programme failure is not a sector-specific design oversight but a structural feature of how donor-funded intermediary architecture has been commissioned across both fields.

The design implication is sharper than usually acknowledged. When the literature identifies peer networking as the highest-value element of acceleration, the peer relationship that matters is founder-to-founder, not programme-to-programme. Networking among programme partners has its uses - coordination, learning, avoiding duplication - but it sits upstream and answers to entirely different incentives: institutional positioning, funder visibility, pipeline exchange. None of those translate into venture-level value. The relationship that compounds is the one between founders themselves, sustained beyond the cohort window and facilitated by technology that makes staying connected low-cost. Platforms designed for this purpose - such as Scaleup Global - make founder-to-founder connection persistent rather than episodic, and turn alumni networks from a reporting line into operational infrastructure.

10. Provide high quality guidance, coaching and mentorship by experienced locals

Status: vindicated.

Connections to skilled role models with whom founders can relate is especially important for those without much prior experience. Endeavor's model - linking founders to other founders in its local portfolio, supplemented by mentoring from local business leaders - is a template worth adapting. The GALI ten-year synthesis identifies peer networking and strategic introductions as the highest-value elements of acceleration from the perspective of top-performing ventures - not curriculum, not workshops, not capital. These are precisely the elements that disappear when programmes are designed around output legibility rather than founder need. The correction period created a larger cohort of experienced founders who have navigated a full venture cycle, including the difficult parts. Designing mechanisms that route their knowledge to earlier-stage ventures - structured mentorship programmes, alumni networks, peer matching - is one of the highest-return investments the ecosystem can make. GALI adds an important finding: enterprises that participate in multiple programmes on average derive benefit from both their first and subsequent acceleration experience, provided each programme is clearly designed to solve different problems at different stages. The goal is sequenced depth, not cohort recycling.

11. Build the skills required to scale a business, not start one

Status: vindicated.

Hard skills development in governance, legal standards, accounting, and monitoring and evaluation is necessary but relatively straightforward to deliver. What is much harder is retraining thinking - developing the cognitive, leadership, and delegation skills required at the scale-up stage that are fundamentally different from the skills that got a venture to its first ten employees. The WDI East Africa study identifies leadership and governance support as one of the most consistently underprovided elements of ESO programming: programmes capable of addressing these needs are rare, and their absence means ventures reaching scale often lack the board-level infrastructure that sustained growth requires. The board-level architecture itself is treated substantively in The Scaling Decision Log as one of the five inflection-point decisions; the implication for acceleration goal-setting is that programmes should be helping founders prepare for that decision, not delivering pitch coaching when governance restructure is the binding constraint. Mindset shifts matter. Accelerators can provide a safe space that encourages seeing failures as learning opportunities and fosters a culture of collaboration over competition. Both remain scarce in African ecosystem contexts, where social norms around face-saving and hierarchy can make the candour that genuine improvement requires uncomfortable to deliver and to receive.

12. Integrate mental health support as a programme standard

Status: revised by experience - empirical foundation now substantially stronger than in 2022.

Psychosocial support for founders should be a standard programme component, not an optional extra. The empirical foundation has grown considerably. A review indicates: entrepreneurs face systematically elevated mental health risks driven by financial uncertainty, role conflict, social isolation, and the cumulative pressure of personal accountability for organisational outcomes. The African context - where founders carry the weight of community expectations, family obligations, and the structural volatility of the operating environment - compounds these pressures considerably. By celebrating high levels of resilience, the ecosystem detracts attention from the corollary: the negative effects on mental health that resilience narratives obscure. The correction period intensified this. Founders who watched companies they had built for years fail, or who were forced into large-scale redundancies, experienced acute psychological stress the ecosystem has barely acknowledged. Programmes that claim to support founders comprehensively but ignore this dimension are not providing comprehensive support. The empirical case for treating founder mental health as core programme architecture rather than optional extension is now substantially clearer than it was in 2022.

13. Focus on the commercial value of non-financial support

Status: vindicated.

Market access, network connections, regulatory navigation, and governance improvement are frequently more consequential than the equity-free grant that anchors many programme value propositions. The bias toward headline grant sizes in programme marketing obscures this. The GALI ten-year synthesis provides the empirical grounding: in emerging markets, ventures see more benefit from acceleration in terms of revenue growth than capital raised. The value is in operational and commercial support, not the pitch-day signal - and programmes designed as if the opposite were true are optimising for the wrong outcome. Shifting the mindset - among both programme designers and founders - toward the full value of non-financial support is a prerequisite for founders engaging seriously with what programmes actually have to offer.

14. Improve financial sustainability and reduce donor dependency

Status: vindicated

Accelerators' dependence on donor funding has been one of the most consequential structural weaknesses of the ecosystem support landscape. The correction period exposed this directly: donor budgets contracted alongside VC funding, and many hubs were left without runway to adapt. The 54 Collective collapse demonstrated the systemic risk of large-scale donor-backed institutions whose operating viability depends entirely on a single funding relationship.

Accelerators with diverse, commercially grounded models proved more shock-resilient. Sustainability means revenue diversification, not the elimination of donor money. The routes available - corporate partnerships, fee-for-service revenue, equity or revenue participation, cross-subsidisation, and endowment development - are well documented; what is missing is the will to use them. The Argidius SCALE evidence is specific: charging improves job creation and return on investment, helps providers select the right candidates, and increases programme engagement. Free provision is not neutral - it actively undermines programme quality by removing the selection signal that distinguishes ventures with genuine growth ambition from those attending for the grant. The GALI ten-year synthesis confirms the structural reality: the equity-return model that sustains Y Combinator and TechStars works for approximately one in twenty accelerators even in the US and Europe. Donor or government support will remain part of the mix in African ecosystems for the foreseeable future - but it needs to be structured around organisational development, not output compliance. The Sustain Impact report proposes three complementary donor engagement modes that together can break the vicious cycle: Programme Funding, Organisational Development - including an Argidius-style ten percent OD buffer on top of every programme grant - and Ecosystem Strengthening over multi-year horizons. The ODA contraction projected through 2026 makes the urgency of this reform sharper rather than softer. The ESOs that adapt their funding architecture before the contraction lands will be operating from a position of strength; those that wait will adapt under crisis.

15. Improve data collection, monitoring and evaluation, and honest reporting

Status: vindicated

One reason accelerators remain underfunded relative to their potential is that they cannot prove quantifiable impact. They are not collecting, tracking, and analysing data on a continuous basis. If an accelerator can articulate value and impact through improved quality measurement and evaluation, it will attract more funding - and more importantly, will learn faster what is working and what is not. Without the capacity to test new business models or have honest conversations with donors about failures, ESOs cannot adapt. Shared monitoring frameworks and centralised data platforms - rather than each organisation maintaining fragmented, donor-specific reporting formats - would boost trust, transparency, and evidence-driven decision-making. AI tools make continuous monitoring materially more achievable than in 2022 - capacity constraints no longer hold with the same force. The substantive treatment of how AI changes capability for ESO operations sits in What AI changes about African scaling; the implication for this goal is that the cost of continuous, honest, comparative measurement has fallen substantially. The institutional barrier is no longer technical capacity. It is the willingness to surface comparative performance data that the donor architecture has historically discouraged.

More honest reporting on failures and lessons learned is a prerequisite, modelled on Mercy Corps Ventures' annual impact reporting which explicitly publishes failures alongside portfolio highlights. Africa needs the equivalent at the country and programme level. The structural barrier to honest reporting is not technical - it is the conflict-of-interest treated in Political Economy Ecosystem. Solving the M&E problem requires solving the governance problem that has historically prevented honest M&E from being safe to produce.