Where Africa’s compute lands.

“What transferred was the basic principle: when multiple international networks converge in one location, there is value in creating a genuinely carrier-neutral facility where those networks, content providers, enterprises and other operators can interconnect.

Marseille had demonstrated that a cable-landing city could become much more than a place where submarine cables terminate and traffic simply transits. At the time, we thought of it as analogous to a container port: ships arrive and depart, while factories nearby process the goods. Data arrives, is exchanged, processed, cached or hosted locally, and then distributed onwards.

What did not transfer one for one was the market depth. Marseille already sat inside a large European connectivity and enterprise market. Mombasa had significant subsea connectivity, but the local demand base was much smaller and much of Kenya’s commercial and technology demand was concentrated in Nairobi. You could not simply copy a European data-centre business plan and expect the same utilisation curve.

There were also practical challenges: incomplete land-registration information, an evolving regulatory framework, and the need to establish reliable infrastructure data. Interestingly, Mombasa already had a fully digitised overview of its power infrastructure.”

“Because submarine cable capacity and end-user demand are two different things. Mombasa was the international gateway, but Nairobi was where most of Kenya’s corporate headquarters, banks, technology companies, government institutions and the larger concentration of users were located. The fibre carried capacity inland to where the economic activity was.

That does not diminish Mombasa. It is strategically important precisely because the cables land there, and a carrier-neutral facility close to those landing points improves resilience, interconnection and the economics of moving traffic onwards.

Nairobi also had a climatic advantage: at around 1,800 metres above sea level it is significantly cooler than tropical Mombasa, which can reduce the cooling burden.”

“One, anchor customer or credible demand. Two, power and grid connection. Three, suitable land and fibre connectivity. Four, capital.

The anchor customer comes first because without credible contracted demand it is increasingly difficult to raise either equity or debt for a greenfield project. Power is then a fundamental constraint: you need sufficient capacity, reliability and a credible timetable for connection. A data-centre site without dependable power is simply not a viable site.

Land matters, but it is not only about acquiring a plot. It needs the right power access, fibre routes, security, planning environment and expansion potential.”

“The test should be whether the market has the same combination Kenya had around 2015: improving international connectivity; rapidly growing domestic internet demand; limited existing carrier-neutral capacity; a sufficiently large commercial centre; improving fibre; credible power availability; regulators willing to enable competition; and enough early customers to support the first facilities.

Several markets have some of those characteristics today, but I would hesitate to call any one of them the next Kenya. If I had to nominate one to watch, Morocco is interesting, although it is not a perfect comparison. It combines strong international connectivity, a sizeable commercial economy, proximity to Europe and an evolving digital infrastructure market.

The relevant question is less which country is the next Kenya, and more where connectivity, power, demand, regulation and execution start to come together.”