At the end of March, Gateway Communications invited a panel of people and a selected industry audience to look at the “Future of African connectivity”. With the exception of a small amount of disquiet over shared infrastructure, there was almost universal consensus about the three items picked out in the headline above.
The future of telecoms in Africa is always a funny thing because it consists of two stereoscopic images that don’t’ always come into focus: what the developed world is doing and what Africa does differently. The same things that happen elsewhere inevitably happen in Africa: everything from markets being liberalized to consumers wanting gadgets like iPads. But all sorts of things happen differently from elsewhere: it’s a pre-pay market and M-Pesa is still well out ahead of elsewhere because the market is largely unbanked.
Sometimes this means that Africa’s operators lumber along, seemingly endlessly constrained by African realities and at other times as with M-Pesa and the rise of social media, the continent bounds along, leap-frogging old attitudes and technologies.
The panel addressing what will happen to the “Future of African Connectivity” were a selection of the brightest and the best from different parts of the industry. Andile Ngcaba has done time in Government as a senior civil servant, served as Chair at Didata and through Convergence Partners has invested in Intelsat’s New Dawn Satellite and the Seacom cable. Stafford Masie worked for Telkom and Didata, set up Google South Africa and is now a technology investor and consultant. IbrahimaGuimbaSaidou used to work for Intelsat and now works for SES. Chris Wood used to work for Flag and is now CEO of WIOCC, the JV initiated by the World Bank within EASSy. Nzioka Waita has worked for Safaricom for over eleven years and is now its Head of Corporate Affairs.
Many things were said and what follows is not a verbatim account. But there was a surprising consensus on the arrival of big data, the death of voice revenues and need for shared infrastructure to deliver Over-The-Top services.
So what is big data? It’s the seemingly relentless requirement for more data capacity to do the kinds of things that data enables. You Tube has been in the Alexa.com rankings for at least three years on all the African countries they provide information for. Since back then it was like eating a three-course meal through a straw, it took some determination to do it. Now it’s sort of OK but not great.
But the same people waiting patiently at their PCs while the clip buffers will soon be joined by millions wanting to see Just a Band or P Square on their mobile phones. A large amount of existing international bandwidth is being soaked up in this way and mobile networks are all being upgraded to take the strain. But this is just the start.
Wait until there’s a complex ecosystem of content providers all uploading content for download: they will all need better upload speeds. On a reasonable connection in the UK, a 10-15 minute HD video clip can take 2-3 hours to upload. Stafford Masie gave one of those Moore’s Law trajectories which ended up with almost every item of knowledge that had ever been published on the equivalent of an iPad. Mike Van den Bergh of Gateway made the point that whilst there were people complaining now about what seemed like an over-provision of bandwidth, it would in time turn out to be not enough.
This is currently deeply disquieting for mobile operators because data like this takes ever larger amounts of bandwidth compared to voice but does not produce the same amount of money per mb or whichever unit you’d care to measure it in. The crossroads for the mobile operators are not clear: should they try and control their own content in some seamless way (as Apple does) that allows greater profitability? Should they use things like apps to provide some pushback against falling revenues? Currently African mobile operators have neither the skills to make what they offer seamless, nor the content rights to make mobile the ubiquitous media.
As if those thoughts were not unsettling enough, Andile Ngcaba said that he thought voice revenues would be dead by 2017 and compared the mobile operators to the old ice factory owners, saying that fridges would never catch on. Nzioka Waita from Safaricom felt it might happen before that date. This afternoon I have had a call with someone in South Africa using a free mobile calling app called Telfree and it was as clear as a bell. Once the Skype-effect starts to lay waste international and domestic calling on mobiles, the voice revenue figures will move relentlessly downwards.
Andile Ngcaba made the point that it was no good the operators complaining that “Over-The-Top” services (jargon for things delivered over the Internet and often free at the point of access) were not producing income for them. They had to refashion their business model to make it work. This is a tough challenge and mobile companies are wont to complain that they don’t want to become just a “dumb pipe”. But the point about selling bandwidth is that if you sell enough at commodity prices, you will make money but then that doesn’t leave room for everyone to be a winner.
So if the future requires an extremely large network investment, then shared infrastructure begins to look a very attractive prospect. The fears expressed in the room were whether these shared infrastructure structures would simply allow the existing “big guys” to entrench their position. Nzioka Waita described Safaricom’s current tussle with the Kenyan regulator over what spectrum should be used for the proposed Open Access LTE network in Kenya. So what if mobile operators go from being vertically integrated to just being service layer provides? Goodbye, masters of the universe?
So the choice for the mobile operators is do they stay in network operation with the increasing mismatch between wholesale price return and CAPEX investment? At a very practical level, a small country like Niger needs a second fibre connection. On current traffic, this will only happen if the existing operators come together to build it and work with Nigerian operators to make it a reality.
But if all of this sounds too far away for you to worry about, there is another problem, something development economists used to call “uneven development”. There is an Africa that is threaded into the future described above but it is largely urban and it moves fastest in a handful of markets.
There is another Africa where the future dares not speak its name because there is effectively only one telco operator or two and they are still struggling to get into the back end of the twentieth century, not the 21st century. There is also a rural Africa where none of us who cherish a future Africa that is different from today have yet described adequately what will make a change.
