Africa's Deep-tech Investment Story Doesn't Hold Up to Scrutiny
In a previous piece, I argued that Africa's digital future risks becoming another missed opportunity because we keep reaching for frameworks that are incompatible with our realities.
The cost of getting this wrong is not just another failed investment cycle. It condemns the next generation to a cycle of dependence that is already extremely difficult to break.
This piece is a different, more specific concern. It is directed at the narratives being constructed around African deep-tech investment, and the gap between how that story is being told and what the ground reality actually looks like.
A recent article covering SWEAT Africa, a tech gathering in Stellenbosch, South Africa, offers a useful case study because it is very typical of the discourse coming out of this sector. It captures, almost perfectly, how investment narratives about Africa get assembled: real problems, plausible logic, credible speakers, and conclusions that do not quite follow from the evidence.
The Architecture of an Investment Narrative
The article makes a case for deep-tech as Africa's next major venture capital frontier. It cites investors, founders, and fund managers, all speaking at what is essentially a gathering of the African tech investment community.
The arguments offered are, on the surface, compelling: AI is commoditising software, making deep-tech's patent moats more defensible; Africa has a young and genetically diverse population that could be a research advantage; limited hospital capacity creates demand for preventative, AI-driven care; and high drug resistance rates in Africa justify building locally-tailored pharmaceutical solutions using AI.
Each of these claims deserves scrutiny, because the logical leaps embedded in them are the entire argument. If they fail to hold true under scrutiny, the entire thing falls apart, and we would have spent more years wasting resources, while the rest of the world is moving forward at light speed.
Where the Logic Breaks Down
Take the hospital capacity argument. The article states that with limited hospital capacity across Africa, predictive models could shift care from treatment to prevention. The implicit logic is: scarcity of treatment infrastructure creates demand for prevention infrastructure. But prediction-driven prevention is not cheaper or simpler than treatment to deploy. It requires clinics capable of acting on predictions, staff trained to interpret model outputs, patients with reliable access to follow-up care, and data pipelines sophisticated enough to produce meaningful signals in the first place. The article presents the problem as though it implies the solution, which it does not.
More jarring is the claim that Africa's young population and linguistic diversity constitute a competitive advantage in deep-tech. These are presented in rapid succession, as though their connection is self-evident. It is not. The genuine potential advantage in AI health research is genetic and phenotypic population diversity — which is real and scientifically significant. But that has nothing to do with median age or language count.
The conflation of three unrelated demographic facts into a single "advantage" narrative is more of rhetorical momentum than any real competitive edge.
The drug discovery argument is the most structurally honest of the claims, but it still fails on execution. The problem it identifies is real: drug development has historically been centred on Western populations, and resistance rates in Africa are partly a consequence of therapies not designed for local disease profiles.
The conclusion, however, that this gives Africa a massive advantage in using AI to develop locally does not follow from the problem. Having an unmet need does not create the capacity to meet it. That capacity requires research-grade data infrastructure, computational resources, regulatory frameworks capable of approving locally-developed therapeutics, and clinical trial ecosystems.
None of these exist at meaningful scale across most of sub-Saharan Africa.
Identifying the gap is not the same as being positioned to close it.
The Expert Knowledge Problem
What is notable about these statements is not that the people making them are uninformed in general terms. The investors quoted in the article clearly understand venture capital mechanics, portfolio theory, and the broad logic of defensible innovation. The problem is domain specificity.
Understanding that patent moats create durable competitive advantages is a general investment principle. Whether those moats are buildable in a specific regulatory, scientific, and economic context is a local execution question, and it is a different kind of knowledge.
This distinction matters because the social dynamics of events like SWEAT Africa tend to suppress it. When credible people in a room share the same optimistic framework, the absence of local operational expertise goes unnoticed.
Nobody is lying. But nobody is asking the questions that would require living in Lagos or Nairobi or Accra to answer:
- What does Nigeria's current inflation environment actually do to the institutional buyer layer that B2B deep-tech depends on?
- How solvent are the insurers and banks that climate intelligence startups are selling to, when their consumer base is contracting?
- How reliable is government support for research infrastructure across election cycles?
Leapfrogging is real. Kenya proved it with M-Pesa. But leapfrogging requires understanding exactly which layer you are skipping and why.
Simply transplanting an advanced solution into an environment that hasn't built the foundation beneath it is not leapfrogging. That's gambling.
Nigeria offers the clearest illustration of this fragility, but the dynamic is not unique to it. Over 7 million MSMEs closed between 2023 and 2024 alone — conservatively representing over ₦1 trillion in lost market value, purchasing power, and institutional depth. A direct consequence of policy choices that compressed margins across every layer of the domestic economy. The country's inflation has been severe enough to compress margins across consumer-facing businesses and the institutions that serve them.
When the B2B buffer — the solvent institutional middle layer between a deep-tech product and a low-income end user — is itself under macroeconomic pressure, the entire investment thesis becomes fragile in ways that are invisible from a conference room in Stellenbosch.
The VC Model and Its Mismatch
There is a structural problem underneath the narrative one. Venture capital is optimised for a specific return profile: high growth, large addressable markets, scalable unit economics, and exits within a defined time horizon.
This model works well for software, fintech infrastructure, and platforms with network effects, but it is a poor fit for most genuine deep-tech, anywhere in the world, and an especially poor fit in African contexts.
Even fintech — the one sector where the African VC thesis has been most consistently validated — is showing structural strain. Fintech business models are shifting from genuine financial inclusion toward fee-heavy payment and lending services. The bridge from saving to investing is narrowing. If the proven sector is under this kind of pressure, the argument that deep-tech represents the next frontier requires more than optimism to hold up.
Deep-tech in Africa that targets local problems with low-income end users faces a compounded challenge: the scientific risk is high, the regulatory pathways are underdeveloped, the capital requirements do not scale down just because you are operating in a lower-income environment, and the purchasing power of the local market cannot support the pricing necessary to generate VC-grade returns.
The article acknowledges this indirectly, as one analyst mentions that without market traction, startups need slower, patient capital such as grants — but does not follow that observation to its logical conclusion: that most of what is being described is grant-dependent development sector innovation dressed in venture capital language.
That is not a dismissal of the work.
Development sector innovation is real and valuable. But calling it venture-backable deep-tech creates a narrative that is exciting to investors, fundable through VC vehicles, and legible in the language of global capital markets. The question is whether that narrative accurately represents the underlying reality, or whether it is optimised for fundraising.
Genuine Exceptions
To be precise: there are categories of African deep-tech where the local context is a genuine, non-negotiable advantage rather than a cost-reduction strategy. Where the research must happen in Africa because that is where the patients, the crops, the climate patterns, and the disease profiles are.
These cases are real. But they share a common feature: they require a global exit thesis from the beginning. The product may be developed in Africa, tested on African populations, and first deployed in African markets, but the return on investment depends on export, foreign acquisition, or licensing to markets with higher purchasing power.
That is a fundamentally different story from "Africa's diverse population is an advantage for building African health solutions for African markets." It is a story about Africa as a strategic research and development location for globally scalable products. Those two narratives are not interchangeable, but in rooms like SWEAT Africa, they frequently get merged.
The Uncomfortable Question
I want to end not with a prescription but with a question that I think the African tech investment community has not seriously confronted:
Who is this narrative for?
If it is for foreign investors who do not know the local realities, then the narrative is optimised for legibility to outsiders, not for accuracy. A foreign VC who hears a few respected voices describe African deep-tech as the next defensible frontier, who does not have the local knowledge to interrogate the execution assumptions, and who is already primed to look for the next emerging market story, is in exactly the position where social proof substitutes for due diligence. The narrative does not need to be false to be misleading. It only needs to omit the questions that matter most.
If it is for African founders, to give them visibility, attract capital, and accelerate the ecosystem, then the question becomes whether a narrative built on overstated investment logic ultimately serves them.
Founders who raise capital based on theses that the local market cannot support are being set up for a particular kind of failure that will make the next generation of African deep-tech founders less fundable.
And if it is for the continent — for the genuine long-term development of African science, technology, and productive capacity — then perhaps the most useful thing the people in that room in Stellenbosch could do is say, clearly and on the record, what African deep-tech actually requires to work: patient capital with development-oriented return expectations, government investment in research infrastructure that is insulated from political cycles, technology transfer offices capable of moving university research into commercial pipelines, and regulatory frameworks built for locally-developed products.
That is a harder story to tell in a conference room. It does not generate the same excitement. But it is the story that might actually be true.