A roadmap for Africa’s data centre market: Part 1
The African data centre landscape is inconsistent, opaque and complex, but there is a trend…
Where does your firm sit in this landscape?
Can you leverage colocation or cloud data centres?
By any measure, the African data centre landscape is under-served and will see massive growth over the next decade.
Africa is seeing growth in regional hot spots that are forming around a hub and spoke model, the regional hubs being:
- South Africa (south region).
- Egypt (north region).
- Kenya (east region).
- Nigeria (west region).
These hub countries are where most of the development is currently, but over the next couple of years this will spread to the surrounding areas (spokes) as demand for data grows here.
If you operate in these regions, now is the time to understand how and where your business can make the most of this opportunity.
Data centre types
In my post, What is a data centre?, you saw how data centres evolved from a room of servers and switches managed by an organisation (enterprise) at your office (on-premise), to huge, dedicated buildings managed by third parties such as colocation or cloud providers. As a result, an organisation with IT requirements will fall into one of these categories:
- Enterprise data centre – Small, medium and large data centres managed by organisations on or near their office premises. These data centres are commonly called ‘on-premise’ or ‘enterprise’ data centres as they are owned and managed by ‘the enterprise’ i.e. a company, university or government organisation.
- Colocation data centre – Owned and managed by organizations specialising in operating data centres. They lease space for you to put your servers and switches, in racks. This reduces or eliminates your need for enterprise data centres, e.g. Equinix or Africa Data Centres.
- Cloud data centre – Cloud operators like AWS, Salesforce, etc. provide not only the building, but the IT hardware and software as well. These operators have huge customer bases and to best serve their customers, they build and operate their own data centres.
- Edge data centre – This is a relatively new segment of the market, growing as the demand for low latency applications rises in industries such as Internet of Things (IOT) and 5G. This drives the development of smaller data centres built closer to customers i.e. at the edge. I’ll deal with this subject in detail, in a later post.
Now let’s take a look at how the African data centre landscape has developed over the years.
What does the African data centre landscape look like?
In the context of the African data centre landscape, each country is on its own journey, but there are general themes that draw the landscape together. From my post on the drivers in a data centre market, we saw that the data centre landscape develops when:
- The telecommunications market in a country opens up and becomes de-regulated. This encourages competition amongst the incumbent telecom operators and encourages new entrants, becoming carrier-neutral.
- Data usage and the need to minimize network latency grows, causing the location and size of data centre demand in a region to also grow.
- As a result, colocation data centre operators enter the market. These operators offer enterprise users a managed, carrier-neutral, off-premise data centre solution that allows enterprises to concentrate on their core business.
- As the market matures, cloud providers enter the market to offer enterprise clients cloud solutions as well.
- Finally, as companies take up technologies such as the Internet of Things (IOT) and 5G, data centres move close to the customer and edge data centres develop.
On this assumption, we‘re seeing the following market dynamics play out in Africa.
Hub & Spoke
The hub & spoke model originated in the airline industry. A central hub is established – that’s a concentrated, central and developed location, connected to smaller locations via spokes, much like a bicycle wheel.
The consensus has been to view the developing African data centre market as a regional hub and spoke model. That’s because the distance between these hubs is big enough that as the market grows, latency will demand hubs radiating to spoke locations. This is clear when we look at the level of development in the four hubs across Africa.
Coastal v Inland
Another theme in the growth of Africa’s data centre landscape centres around the arrival of submarine cables and their effect on the access and cost to telecom networks. As these massive fibre pipes roll out around the continent, they provide increased connectivity for Africa to the rest of the world, which increases bandwidth (bigger pipes) and eventually reduces the cost of network access in that location (more competition).
In the locations where these cables come to land, they’re terminated at a Cable Landing Station (CLS). This is a key connection point and various things happen here:
- Increased bandwidth speed and reduced network costs for consumers.
- Which encourages people to consume more content, content providers (e.g. Netflix) then needs to process data locally at local data centres to ensure the customer experience is maintained.
- Businesses use more cloud applications, as this grows, critical mass is achieved, cloud providers then locate servers in this location as well.
Over time, this location and the surrounding area becomes more connected to the land (terrestrial) network which brings greater access to the inland areas, and their cost of data usage comes down in these inland locations as well. So, we see growth from coastal to inland.
Where is it going?
Each region is at different stages of development. We can see this from the activity going on in the hub of each region. To do this, we can look at:
- How open (carrier-neutral) the telecom space is. Is there competition in the network space?
- Are there colocation players in the market and is it growing?
- Are the global colocation players present in the market?
- Are the hyperscalers present?
We can then rank them on a scale of 1 to 5 on the following criteria:
| Not yet | Thinking about it | Getting started | Established | Growing |
|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 |
Using this as a basis, here’s how the hubs in each region score:
| Region | South | North | East | West |
|---|---|---|---|---|
| Country | South Africa | Egypt | Kenya | Nigeria |
| Open ‘carrier neutral’ telecom market? | 5 | 2 | 4 | 3 |
| Is there a colocation market? | 5 | 2 | 4 | 3 |
| Are the global colocation players present? | 5 | 2 | 3 | 3 |
| Are the hyperscalers present? | 5 | 3 | 3 | 3 |
| TOTAL | 20 | 9 | 14 | 12 |
From the results we can see:
- South Africa is a fully developed market, that is just starting to hit its stride.
- Nigeria and Kenya are growing however are still in the developmental phase.
- Egypt has a long way to go to open up, however the sheer size of the market is pushing demand along.
In my next post I’ll look at each region in more detail.