Wednesday, January 17, 2007

Timing the tipping point

A couple of previous posts on this blog were intended to give a flavor of how conventional bandwidth demand projections might be upset, or their time frames compressed, by "unforeseen developments". As one example, BSkyB has publicly stated a working assumption that end-user bandwidth demand may double every five years. Yet I have to question how these kinds of assumptions are arrived at, and how closely connected the forecasters are to the end-user behaviors driving the changes (check out this article on Joost from NRC Handelsblad, in which the CTO complains that provisioning fiber to the development office in Leiden took an eternity because the vendor couldn't grasp why anyone would possibly need so much bandwidth).

Just looking at the issue from a historical perspective also makes me very uneasy with some of the projections out there. (Here I am uncomfortably reminded of other previous industry predictions which proved wide of the mark - "no one will ever need or want their own computer," "SMS has no potential as a consumer service," "mobile penetration will peak at 20%," "the public internet will never support acceptable quality for voice," etc.) Five years ago, broadband in most of Europe meant typically 512k or 1Mbps, but today in France it may mean 28Mbps. It's still hugely asymmetrical, which is a big problem in my view, but nevertheless, theoretical downstream capacity has grown by a factor of as much as 56 times.

Capacity is not the same thing as consumption, but it is still noteworthy that the "capacity to consume" has grown many times faster over the past five years than the rate projected by BSkyB for the next five. I'm not singling out Sky for abuse here, rather it's just a convenient example, but I don't understand why future expectations should be more conservative than has been the case historically, when the evidence I have seen seems to point to something more steeply linear. Nor do I think there is any fundamental difference between French and UK internet users which suggests that the UK needs less bandwidth. Yet one view in the UK is of demand doubling every five years, while in France we already see a case for jumping straight to FTTH.

I think the attempt to measure rates of change, to define the critical tipping point, will be a hotly debated issue over the next couple of years. And where there is fodder for debate and industry/investor anxiety, there will be dramatic research, such as this newly released survey from ABI Research, which appears to cover different approaches to addressing the "bandwidth crunch," and concludes that MSOs will need to spend $80bn globally to remedy it.

I certainly agree with the drivers they identify, and I am not in a position to confirm or dispute the numbers, but the last sentence caught my eye: "In practice, says the study, network upgrades will naturally start in the major urban centers and gradually spread to less densely-populated regions." This may be true for the established players in cable, but this very bias would seem to offer incentives for projects of the kind we have seen in the smaller Dutch towns of Hillegom and Nuenen - where entrepreneurial capital and local consensus have converged to leapfrog offerings by the telco/cable duopoly, and apparently generate fiber envy in neighboring areas.

So let's add to the list of potential "unforseen developments" driving higher bandwidth consumption the idea, often covered on this blog, that local self-determination also has a role to play. (Incidentally, I recently came into contact with a company with a very interesting model for addressing just this space - please contact me if you are interested in speaking with them.)It's not necessarily all in the hands of the incumbents to determine what our demands will be and deliver supply in an incrementalist trickle.

UPDATE: Interesting article from Level3 here seems very much in line with my sentiments.

UPDATE ON 25TH JANUARY: Read this one and weep!

No comments: