Africa has attracted major investment into subsea cables, data centres and cloud infrastructure. But digital buildout will depend on how effectively these assets are connected – and whether capacity can move efficiently from international networks to data centres, businesses and ultimately users.

Africa’s digital infrastructure map is filling up.

New submarine cables have dramatically increased Africa’s international capacity. Two of the largest systems alone – 2Africa and Equiano – provide up to 180 Tbps and 144 TBPS of design capacity respectively. Data centre operators are expanding facilities in markets including South Africa, Nigeria and Kenya. Cloud providers are establishing infrastructure on the continent, while investors are committing more capital to fibre, computing and connectivity.

But building more individual pieces of infrastructure does not necessarily produce a functioning digital ecosystem.

A submarine cable can land enormous capacity on Africa’s coast. A data centre can provide megawatts of computing capacity. A mobile network can connect millions of subscribers.

The economic value comes when those assets are connected.

That was one of the less obvious conclusions from discussions at ITW Africa in Nairobi: Africa’s next digital infrastructure challenge is not simply building more infrastructure, but connecting the infrastructure it already has.

For Chris Wood, Chris Executive Officer at the WIOCC Group, one of the largest gaps is increasingly at the metropolitan level.

“I think there’s going to be a lot more subsea cables coming to Africa,” Wood told the conference, pointing to investments by major technology companies.

But he argued that the more immediate requirement lies elsewhere.

“Probably the biggest area that is required, I think, is metro networks.”

The missing middle

Africa’s connectivity story has often been told through submarine cables. Over the past two decades, successive cable systems have transformed the continent’s international connectivity, adding capacity and creating greater diversity between Africa and the rest of the world.

But landing capacity on the coast is only the beginning of the journey.

Traffic still has to move from landing stations into national backbone networks, through cities and into data centres, internet exchanges, mobile networks, businesses, homes and other places where connectivity is consumed.

That makes the terrestrial network between international capacity and the end user increasingly important.

Wood noted that major cities in markets such as Kenya are relatively well served, but said significant metro infrastructure still needs to be built, particularly beyond Africa’s leading digital infrastructure markets.

South Africa, Kenya and Nigeria have attracted substantial investment. The challenge becomes more pronounced when moving beyond those countries and their largest commercial centres.

This is Africa’s digital infrastructure missing middle.

The continent may have international capacity at one end and millions of potential users at the other, but without sufficient metro, backbone and interconnection infrastructure between them, the two cannot be efficiently connected.

A cable landing is not connectivity

This distinction matters because infrastructure announcements can create a misleading picture of actual connectivity.

A country can have several submarine cables landing on its coast and still experience expensive domestic connectivity – as Nigeria illustrates, where carrying capacity from Lagos to Abuja can cost up to four times as much as carrying it from Lagos to London.

Similarly, a country can host modern data centres while businesses outside its principal commercial districts struggle to access affordable, high-capacity fibre.

It can have an internet exchange while networks in other parts of the country must travel long distances to reach it.

The infrastructure stack is interconnected: submarine cables bring international capacity into a market; national backbones transport it across the country; metro fibre distributes it within cities; internet exchanges allow networks to exchange traffic locally; data centres host applications and computing capacity; and access networks ultimately connect businesses and consumers.

Weakness at any point reduces the value of the rest.

This is why measuring Africa’s digital infrastructure development only by the number of submarine cables, data centres or mobile subscribers can miss an important part of the picture.

The more relevant question is how effectively those assets work together.

The next $300m

WIOCC Group‘s recently announced $300 million investment provides an illustration.

Wood said part of the capital would go towards expanding existing data centre capacity, including in markets where the company sees strong demand.

But the investment strategy also encompasses fibre.

That reflects a broader shift in the African infrastructure market.

The opportunity is increasingly not simply to own one category of asset, but to ensure that the different layers of infrastructure required to move, store and process data can operate together.

Wood also put the scale of the investment into perspective.

For a company attempting to build infrastructure across the continent, $300 million is substantial but nowhere near enough to meet Africa’s total requirement.

That raises a more important question about where the next billions should go.

More submarine capacity will be needed. More data centres will be needed. So will more computing infrastructure.

But there is also a strong case for directing considerably more capital towards the networks connecting those assets.

Beyond Lagos, Nairobi and Johannesburg

Africa’s infrastructure concentration presents another challenge.

Investment naturally gravitates towards markets where demand is already established.

Lagos, Johannesburg, Cape Town and Nairobi have therefore emerged as major nodes in the continent’s digital infrastructure ecosystem. Cairo, Casablanca and other markets are also attracting investment.

The economics are logical: investors prefer locations with enterprise customers, financial institutions, connectivity, power, skills and existing infrastructure ecosystems.

But concentration can reinforce itself.

Data centres locate where connectivity is strongest. Network operators build where large customers are concentrated. Cloud providers follow enterprise demand. Businesses, in turn, benefit from locating close to that infrastructure.

The result can be increasingly sophisticated infrastructure in a handful of cities while secondary cities and smaller markets remain comparatively underserved.

Closing that gap requires moving beyond national connectivity strategies towards more granular infrastructure planning.

It is no longer enough to ask whether a country is connected.

Policymakers and investors need to ask whether its major economic centres are connected to each other, whether businesses can obtain competitive fibre routes, whether networks can reach local internet exchanges efficiently and whether data-centre capacity can serve customers beyond the immediate city in which it is located.

Data centres need networks

The data centre investment boom makes this even more important.

A data centre is not an isolated piece of real estate.

Its value depends partly on the density and diversity of networks that can reach it.

Cloud providers, banks, fintechs, content companies and enterprises need resilient connectivity into facilities. Network operators need multiple routes. Internet exchanges need participants. Customers need to reach applications with low latency.

A data centre with limited connectivity is therefore inherently less useful than one embedded within a dense network ecosystem.

The same principle applies in reverse.

Fibre networks become more valuable when they connect to data centres, cloud infrastructure, internet exchanges and large concentrations of enterprise demand.

This creates a reinforcing cycle: more networks improve the attractiveness of data centres; more data centres create additional reasons to build networks.

Africa needs both.

Power cannot sit in a separate room

Connectivity is only one part of the equation.

The broader ITW Africa discussions also highlighted another increasingly important relationship: digital infrastructure and electricity.

Data centres require large quantities of dependable power. Fibre networks require powered equipment. Mobile networks depend on thousands of active sites. AI infrastructure significantly increases the power requirements associated with computing.

One speaker put the issue succinctly during the Nairobi discussions: power planning cannot sit in a separate room from digital infrastructure planning.

That is particularly important as African governments seek to attract larger data-centre and AI investments.

A country may designate land for a data-centre campus, but without sufficient generation and transmission capacity, that land has limited value.

Similarly, power generation in the wrong location cannot easily support computing infrastructure without the grid capacity required to deliver it.

The implication is that digital infrastructure planning increasingly needs to converge with energy planning.

Think in systems, not assets

Africa’s next digital infrastructure phase therefore requires a change in perspective.

Governments frequently treat telecommunications, data centres, cloud, electricity, roads and urban development as separate policy domains. Investors understandably specialise in particular asset classes.

But the digital economy experiences them as one system.

A cloud application hosted in a Lagos data centre may depend on international submarine capacity, metro fibre, an internet exchange, reliable electricity and a mobile network before it reaches a user.

Every layer matters.

This does not mean every company needs to own the entire stack. Neutral infrastructure and open-access models can allow specialised operators to build individual layers while connecting efficiently with others.

What matters is that infrastructure is planned and regulated with the complete system in mind.

From landing capacity to usable capacity

Africa has made considerable progress connecting itself to the global internet.

The next challenge is making that capacity more useful inside the continent.

That means building the metro networks that connect businesses to data centres, extending terrestrial fibre beyond the largest cities, strengthening national and cross-border backbones, deepening internet exchange ecosystems and ensuring sufficient power exists where computing infrastructure is being built.

The objective should not simply be more infrastructure.

It should be infrastructure that connects to other infrastructure.

Because Africa’s digital buildout will not be determined only by how much capacity lands on its shores or how many megawatts are installed in its data centres. It will be assessed by how effectively that capacity reaches the economies it is supposed to serve.