Search This Blog

Wednesday, 27 July 2011

Ofcom broadband speed tests

Once again, we have an Ofcom report about broadband speeds to peruse.

This blog post can be read at

Samknows has run tests in 1767 homes to create the data upon which this report is based. I know that when the original data set and map was published a couple of years ago, I queried the lack of properties whose lines were being tested in rural areas. This resulted in us being sent the grand total of 5 routers to scatter around the countryside, presumably to balance it out and shut up the rural contingent. Hohum.

It would be good to see the geographic locations of the test sites to grasp whether or not this is (still) a skewed picture of UK broadband speeds by being heavily reliant on urban and short lines.

For instance, if we were to place a similar number of routers into mainly rural homes, and ran 455million performance tests, in particular from 4pm -6pm when the kids come home from school and not from 8-10pm when they should be tucked up in bed, what effect would that have on these results?

It does not take a genius to work out that the national picture would most likely be nowhere near as rosy as this report makes out. Terms such as 'surge', 'boost, and 'superfast' are somewhat misleading to say the least but no doubt the journos will make out all is well in the press over the coming days.....

There has seemingly been a 10% rise in average download speeds from 6Mbps to 6.6Mbps, which should I guess be applauded, even if the symmetry is still missing for upload. The difference between advertised and actual speeds remains alarmingly high. Especially for uninformed consumers who believe what the ads say, and have no way of knowing otherwise until an order is placed and a connection is made. Or not, as the case too frequently is.

The report appears to suffer from creative mathematics, which is never good, except perhaps for those who find that copying and pasting the report without applying even minimal logic makes for good copy. Viz

Superfast broadband services are now available to more than half of UK homes as rollout continues.

Today’s research found that superfast services offer significantly faster speeds than copper ADSL broadband

However, over 75 per cent of UK residential broadband connections are currently delivered by copper ADSL telephone lines.

Superfast >50% homes. Superfast does not equal ADSL. ADSL >75% homes. Spot the problem?!

And as for this sentence:

Download speed is the main performance measure by which broadband services are advertised and is the most important single measure in determining broadband performance.

It leaves me completely speechless. The regulator is really saying this???

At least 25% of those on ADSL are on 4Mbps or below according to this report, although one would suspect that this is a finger in the air figure as BT etc still seem to believe that the so-called commercial sensitivity of line data should come before national interest. That Final Third is supposed to be being addressed by BDUK and local authorities, but there seems to be a distinct lack of vision in doing so even with the plethora of experts this country has available to discuss the topic at the never-ending round of 'talking head' conferences. However, time will tell whether local authorities etc can turn it around and apply common sense (to bypass the hype and spin and guff they are being fed by self-interested parties) to deploy what is actually required in each region.

Meanwhile, NZ are reporting that data usage has increased by a third in the last year, and there is plenty of hard evidence from elsewhere with true next generation broadband connectivity that data usage goes through the roof when you give people quality connections of 100Mbps and above - what the rest of the world calls 'superfast'. Perhaps, Ofcom should start thinking about data caps and how, as Britain gets better broadband, the regulator is going to do something about those so users are not continuously being throttled and prevented from actually using the connectivity they have?


PhilT said...

The >50% superfast comment comes from >45% with cable available and the rest from FTTC VDSL available. Note the available, not "used".

I don't know why ADSL gets 75% of the market when cable is an option for half of it. Obviously speed isn't the primary concern of the buying public in cable areas, or we would see say 80% of cabled area users on cable which would be 40% cable : 60% ADSL across the country.

Cybersavvy UK said...

Not sure abt speed not being an issue as VM just announced that whilst they have lost 36k subscribers (mobile as well as broadband), they now have a higher percentage opting for 40Mbps instead of 30. Dunno if that's a pricing thing, may need to look more closely to find the reasons but CEO seemed to imply that people are willing to pay a premium within the product set.

PhilT said...

VM do indeed have some customers willing to pay, but the numbers are small - the 50M service is below 3% of the userbase IIRC.

The CEO comments referred to new customers joining, current offer is for 10, 30 or 50M headlining at £13.50 £18.50 and £25/month with 50% discount on first 6 months. Phone line required at £13.90/month.

Standalone prices for 10,30,50 and 100 are £21,28.50,35,45/month respectively.

VM's average of 25 GBytes/month per customer was interesting.