I had a discussion with a colleague today, in which he claimed that consumer data will soon become a driver of mobile revenues in Africa. Suffice to say, I disagree, but he wasn’t to be shifted. So I did the following calculation to back up my point of view.
Taking Safaricom as a good baseline (because it is large, technically and commercially sophisticated and based in Kenya, where the populous is relatively rich and educated), I first calculated the maximum possible capacity of their 3G network. Safaricom has about 1,300 3G sites. After overheads, those sites can physically serve about 86TB of data in a year if they are 100% utilised all the time.
Safaricom charges about 1c per MB for data, meaning that if fully utilised all the time, the maximum return the network could produce today is $860Mn. There are 5.1Mn 3G subscribers on Safaricom, so data ARPU in this scenario is about $14.
All well and good, except that the maximum utilisation of the network is about 60% in the real world – demand is lower at certain times of day and in certain locations. With this in mind, data revenue falls to about $500Mn. Still a big number, and about half of Safaricom’s billion dollar revenues.
But there’s a further problem. In real terms, mobile data is about 20 times more expensive in Kenya than in the UK. Although it has great utility, it’s very hard to see Kenyan consumers wanting to pay such high rates. Either they will pay less per unit or use less – commercially the same difference. Being conservative and saying that a Kenyan would pay a 10x mark-up on the UK price gives a revenue for that 3G network of about $50Mn.
Safaricom’s current data revenues were actually about $25Mn in H2 2011 – this includes EDGE as well as 3G – but shows that the above assumptions are in the ballpark. Small change against the backdrop of the investment being put into Africa. It’s also worth noting that Safaricom is a great success story, with a wonderful reputation in-country and substantial market share. Smaller operators would not do so well in price or customer mix terms.
I’ve done a bit of modelling based on the above assumptions and the size of the markets in other geographies. From this I estimate that in 2012, mobile cellular data revenues in Africa will be around $1.5Bn – an increase of over 50% from 2011.