There are a number of misconceptions that are perpetuated in this article, but it certainly suits Derek that his customers don't really know what's going on.
The first question is what one means by "national bandwidth". Physically, the connectivity between cities in South Africa costs what it costs - roughly R2bn to build a fibre network. So far, there are only two of these - Telkom and Neotel/Infraco. There will be more one day when the Neotel/MTN/Vodacom network is completed, and perhaps, further into the future, when FibreCo gets there. (There is one national DFA route, but it's only about 500 km out of about 10,000 km needed to connect the country). At that point, these players will all be trying to sell bandwidth at a small margin above cost, just to stay in the game. In all likelihood, if there are four or more networks, at least one of them will not be financially viable - except as part of a larger services-based operator. This is the unfortunate reality check for South African Internet users - there simply aren't enough consumers to make it worthwhile to build national networks (FibreCo take note). It's all a lot easier in China or India. The good news for consumers is that most of the cost of national networks is covered by services for large businesses. Again, this would be easier in Europe or the US, but it's just about sustainable. The laws of supply and demand apply here just as anywhere else - prices would drop if there were more demand (in the economic sense, not the "I want cheaper bandwidth" sense.)
In the mean time, it's worth noting that the wholesale price of national links is more than 90% lower than it was three years ago, as a result of the competition between the existing two networks. The statements in this article make no sense in this context. Bandwidth prices in metros, where there are multiple fibre deployments, are also plummeting.
This is where the confusion starts. Local transit (sale of an Internet connection that gets you visibility of the whole of the Internet in South Africa, but not the rest of the world) is not "bandwidth" in the sense above. ISPs (notably IS) charge much more than the price of the underlying bandwidth for this, just as they charge much more for international transit than the cost of international bandwidth. In itself, this would be reasonable, and obviously how ISPs stay in business, provided that the incremental cost is fair. What we will see over the next few years is a normalisation of the South African Internet, as the distinction between local and international blurs, and prices for transit drop closer to what they are in major hubs. Prices will probably never be as low as they are in those hubs, but we are by no means unique. In Internet terms, other countries that are in the outer spiral arm of the Internet galaxy have similar frustrations, but it will get better over time.
Sign up with Google