Tunde Abagun, channel sales manager of Sub-Saharan Africa, Nutanix.

In 2017, when Microsoft announced plans to land the first hyper-scale data centre regions in Africa, it took two years to gain fruition, but it’s considered a turning point for the continent.

Since then, AWS has landed a local data centre region, as have Huawei and Oracle, and plans are afoot for Google and Alibaba to host their cloud services locally too.

We’ve also seen an explosion of investments by colocation providers to build the physical facilities to house the equipment needed to run these services and announcements about skills development in the local market.

While these past six years have been formative in the local growth of cloud, the common denominator is that all these activities have taken place in South Africa. But what about the rest of the continent?

With slowing growth in more mature markets, and a current population of 1.2 billion people, which the UN forecasts will reach 2.5 billion by 2050, Africa’s cloud opportunity is finally beginning to be recognised.

Winston Ritson, chief operations officer of Liquid C2, which operates in 16 African countries, says one way to define the opportunity across the continent is ‘cloud-ready’ markets and ‘cloud nearly ready markets.

“In cloud-ready markets, there’s good digital literacy, developed underlying broadband infrastructure – so fibre adoption – and good redundant external connectivity into and out of the country,” he says.

He believes only South Africa and Kenya fit the cloud-ready category, with Nigeria being ‘borderline’ but more of a nearly-ready market.

Markets that he thinks are cloud nearly-ready include Egypt, Zambia, Rwanda, Ghana, Tanzania and Angola.

“We believe there’s huge potential demand, but one factor is currently missing in each market, so the broadband connectivity might not be up to par, or digital literacy is lagging, or, from a macroeconomic perspective, the companies operating there might have legacy investments in infrastructure,” Ritson says.

Andrew Cruise, MD, Routed, says a key consideration is that fast, cheap, reliable internet is necessary for cloud growth in a country.

“We need to be talking about 90% plus fibre coverage in the country before you can start thinking the cloud is going to be successful.”

Ritson agrees. “Ubiquitous terrestrial broadband or fibre connectivity across the continent is what will drive the continued adoption of cloud,” he says.

Submarine cables

Connectivity and the cloud are inherently intertwined and, like the growth of fibre within the territory, the number of undersea cables that increase connections to the continent has also grown over the past decade.

The arrivals of 2Africa and Equiano in the past year are cases in point.

Incidentally, both are backed to a greater or lesser degree by a big tech player (Facebook and Google, respectively), although Stephane Duproz, vice president, of the Africa Data Centres Association, notes these players’ motivations are largely to serve consumer markets, rather than enterprise cloud purposes.

Further, he notes, the cost of broadband access is also a consideration, and lowering these costs is a key driver, especially for consumer adoption of cloud services.

A study released by the ITU last year showed that Africans paid more than three times the global average for mobile broadband services and over five times the global average for fixed broadband in 2021.

“When you have virtually unlimited usage of data, you don’t care and just use the data. But when you pay for it by the unit, you try not to use as much data as possible before you can get connected to a WiFi service, at which point you start using the data again. As Africans, we need to work on easing up data consumption on the continent for the development of the consumer-oriented data operators,” he says.

As well as the high cost of connectivity, locally both enterprises and consumers also suffer financial challenges because most cloud services are priced in US dollars. “In Africa, in terms of foreign exchange between local and foreign currencies, most times enterprises end up paying twice the actual value of what they get,” says Tunde Abagun, channel sales manager of Sub-Saharan Africa, Nutanix.

Currency fluctuations also make it difficult to budget cloud effectively. We can’t talk about the growth of the cloud without considering the role of governments and regulators.

Duproz says that the cloud works best in deregulated telecoms markets, where multiple providers can offer connectivity. “Deregulation is critical,” he adds.

Ritson says if a country’s government is progressive in its thinking and actions, the private sector tends to take the lead and grow the market. He also cites policy certainty as a key factor, as this encourages investment in the necessary infrastructure.

On the opposite side of that, in 2021, there were 16 incidents across Africa where a country’s government imposed internet disruptions, and 13 last year, according to Surfshark. This level of state-controlled censorship impacts the potential for cloud adoption.

Ethiopia is an example of a country that’s experienced severe regulation and government control, where internet shutdowns are often enacted, including a two-year blackout in the Tigray region.

“There are a lot of stringent rules in Ethiopia,” says Abagun. “You can imagine what that would mean for a business that has critical applications hosted in the hyper-scale cloud. Because of that, we find that pretty much all African countries must be more creative with their cloud adoption approach.”

Data sovereignty

While international eyes might think the landing of a hyper-scale cloud in South Africa means large parts of Africa can be served from there, it ignores individual countries’ data sovereignty and data residency requirements.

Cruise highlights the fact that the three big hyper-scalers are American and subject to US laws, and this presents a challenge for national governments. He also believes African governments are likely to want to have sensitive data held in their own countries.

“The likes of AWS and Microsoft have made massive investments to put down a point of presence in the region. I think they’d like to serve most of their customers out of South Africa because there’s enough demand in South Africa to make the investment worthwhile, then push the tentacles further off the continent. But they are going to have to be present in other countries, even if it’s a small presence, to satisfy those countries’ sovereignty requirements.”

Ritson says a lot of purchasing decisions have typically been driven by price or latency, although data privacy and sovereignty are starting to impact decisions, and this varies depending on the sector.

For example, a lot of data privacy and sovereignty laws impact financial services institutions, he says. “In markets where there are financial services institutions and some data privacy/sovereignty requirements, we do see them driving the hybrid conversation and a Lepod conversation, so they meet central bank regulations. We’ve also seen similar demand in Nigeria around the oil and gas industry; where geological survey data must reside in-country, and the data must be processed in-country.

Lepod is Liquid’s Edge solution that it’s rolling out across African markets, using Azure Stack, to provide an in-country node to the Azure service. Given the investments needed to deploy data centre regions in other countries, this provides a public cloud at a local level.

Liquid says Lepod solves the data sovereignty, residency, latency, and bandwidth issues, and is currently available in Kenya, Rwanda, Tanzania and Zimbabwe.

In a similar vein, AWS recently launched a Local Zones presence in Lagos, Nigeria, and has announced plans for one in Nairobi.

Ritson believes there may be a handful of ‘key markets’ where the hyper scalers might invest directly.

“I think their overall strategy will be twofold: they’ll put down acceleration nodes or caching nodes, in bigger territories. The second strategy is to work with partners who can invest in hardware and build out an edge-based strategy for the continent.”

For us, by us

So with the international cloud providers likely only to make investments in select markets, this does present an opportunity for local cloud players.

“I think the fact that we may not be seeing hyper scalers come in anytime soon does lend credence to that opportunity,” says Abagun.

“There is an opportunity for colocation providers, offering the brick and mortar, providing the connectivity and the power and then enterprises in Africa can bring their infrastructure. There are bare metal providers, where in addition to the brick and mortar, they also provide the service, storage, networking switches and so on. Above that are hosted service providers, then managed and cloud service providers.”

He adds that there are also opportunities in countries where there’s a challenging foreign exchange environment.

“The ability to bill in a local currency could be a differentiator as it eliminates a lot of the macroeconomic challenges involved with trading business internationally.”

Cruise believes international hyper scalers typically don’t distinguish or differentiate their approach between regional markets.

“They tell the same story in all of the regions, but that story doesn’t work in Africa. For example, taking data and backing up to the cloud, you just can’t do that when your internet connections are slow and unreliable.

“Internationally, people are beginning to realise that there are viable alternatives to the big-name faceless global corporations.

There are local providers, local clouds, that give you more local control, local residency, local sovereignty, local currency, and local support, not a big nameless, faceless organisation that you can only contact through email, but someone you can see face to face and speak to.”

Does he see opportunities? “Absolutely, because it’s not necessarily the top of the stack, it’s not necessarily the most complicated stuff that you’re competing on. You know, the demands in Africa are much simpler.”

So, the cloud markets are evolving across Africa, but the underlying infrastructure requirements need to be met first, skills will be key, but it will be interesting to see if the US hyper-scale players continue their market dominance, or if local companies who better understand the local market can serve Africa better.

* This article first appeared in ITWeb Brainstorm’s State of the Cloud 2023 supplement, which can be found here: https://brainstorm.itweb.co.za/archive/supplement