According to a new study by Epitiro (no, me neither), the average mobile broadband download speed in the UK is 900Kbit/s. On average users are receiving 24% of the advertised speed, web browsing is 34% slower than on DSL and latency is such that game-playing is not possible. The test seems to have had a pretty robust methodology. 1.4 million tests from 1300 users.
A couple of things spring to mind. Firstly, 900Kbit/s sounds OK. Certainly good enough for browsing. Not quite enough to comfortably stream over-the-top video or play games, but my assumption was always that people who wanted to use high bandwidth services would need to maintain a DSL/cable connection anyway. For these users MBB is a complement, not a replacement.
Secondly, I'm sure there is wild variation within that average. If everyone received 900Kbit/s 90% of the time and there was very little variance there would be a strong argument for MNOs advertising THAT rate. As it is, the variance is massive. Personally, if I could get a guaranteed 100Kbit/s on my mobile broadband, I'd be happy. For this reason the only logical approach is to advertise the theoretical maximum and attach caveats that the signal is likely to degrade due to various factors and you won't always get that, as per the recent announcement by the UK Mobile Broadband Group. Attempting to advertise actual speeds is impossible.
Interestingly the recorded speeds increased by 11% over the period of the study (Dec 08 to May 09) indicating that fears of demand massively outstripping supply have been unfounded. On a related note, Kenneth Karlberg of TeliaSonera commented yesterday at the Open Mobile Summit that data networks will crash if MNOs continue to sell flat-rates. Hence, presumably TS's decision recently to increase prices. This does not seem to be the case in the UK.
Will IKEA disrupt the smart home market?
1 week ago