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 co-founded Zuberi, a digital bank built around earned wage access for Ghanaian government workers — people with verifiable, predictable income and almost no credit infrastructure built around them. We built the team, the product and the operation from scratch. The thesis was right. We reached roughly twenty thousand users. 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 group 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.

B57 runs five. BUILD AI Academy is the skills layer — a public programme on a mission to train 300,000 Ghanaians in practical AI. Mansco is the systems layer, a frozen-protein import and coldstore business at Tema port whose entire operation runs on agents we built. Poligence is the government layer, building sovereign AI services for the state on infrastructure the state controls. Cactus Talent is the talent layer, positioning Ghana as a hub for AI-ready digital talent. The Collaborative Skills Economy, launched at the UK Parliament, is the work layer, carrying that talent into international companies and bringing the value home.

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 skills shortage stops being someone else's problem to solve. Mansco is not a cash-flow business bolted on for ballast, it is where what we teach gets tested against a container that has to clear the port on Tuesday.

The unfashionable part

None of this reads well in a pitch. Groups 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 operations. Real revenue.

The next decade in Ghana, and in the wider West African corridor, will reward groups that own the infrastructure 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.