Every few years a new template arrives for how to build a technology company in Africa. Raise a seed round, pick one wedge, grow until the wedge becomes a platform. It is a good template. It was built for a market that already has the things a single-product company needs to survive: distribution it can rent, talent it can hire, regulation it can read a year ahead.
West Africa does not reliably have those things yet. So the template keeps producing companies that are correct about the problem and wrong about the conditions.
The assumption underneath the startup model
A single-bet startup is a wager that everything outside the bet is a solved problem. That the engineers exist and can be hired. That payment rails will stay where they are. That the regulator's position in eighteen months is roughly its position today. In a mature ecosystem those are reasonable assumptions, because someone else has already absorbed the cost of making them true.
I learned what happens when they are not true. I founded Zuberi to lend to Ghanaian government workers — people with verifiable, predictable income and almost no credit infrastructure built around them. The thesis was right. We reached roughly twenty thousand customers. Then the regulatory weather changed, and growth stopped. Not because the product was wrong, and not because the customers went away, but because a single-product company has one relationship with the conditions around it, and when that relationship breaks there is nothing else holding the weight.
That is the part the template does not price. A startup is structurally fragile in exactly the environments that most need new companies.
What an operating group does differently
A holding company is not a fund and it is not an accelerator. It is a set of operating businesses under one roof that are deliberately arranged to need each other.
BUILD Group runs three. BUILD AI Academy trains and certifies African professionals in applied AI. BUILD Studio takes those operators and puts them into real client work — product design, automation, software that ships. Cactus Talent places human and AI teams into British companies from the UK. Training feeds deployment. Deployment feeds placement. Placement tells training what the market actually pays for.
Each of those is a business on its own terms. Together they do something a single company cannot: they build the ecosystem they depend on as a byproduct of their own operations. The Academy is not a marketing channel, it is how the talent shortage stops being someone else's problem to solve. The Studio is not a services arm bolted on for cash flow, it is where training gets tested against a client's deadline.
The unfashionable part
None of this reads well in a pitch. Holding companies are slower to explain than a single product. They do not produce a clean chart. They demand that you actually run several businesses rather than one, which is harder and less romantic than the founder mythology suggests.
But the arithmetic is straightforward. In a market where the surrounding infrastructure is thin, the returns go to whoever is willing to own more of the stack — and to treat the gaps as the opportunity rather than as excuses. Real classrooms. Real client work. Real revenue.
The next decade in Ghana, and in the wider West African corridor, will reward groups that own the talent layer end to end. Not because owning more is inherently virtuous, but because in this market the alternative is renting something that does not exist yet.