Skip Navigation

Coverage of 2 Mbps broadband increases to over 95%
Thursday 16 May 2013 12:33:10 by Andrew Ferguson

The UK scorecard for broadband coverage has just got a lot better if the latest data from Ofcom is correct. In 2011, the coverage of 2 Mbps standard broadband was listed by Ofcom as 86%, this rose in November 2012 to a figure of 89.9%, but now in just six months has seen a significant jump to 95.3% according to the latest figures from the regulator.

The truth is buried deeper into the report, page 87 to be precise, which tells us we need to stop comparing the figures to the previous data as the methodology has totally changed.

"Data on the availability of standard and superfast broadband differ from what was reported in the “Infrastructure Report, 2012 update.” The estimates in this report were calculated by considering a postcode not to have standard broadband if the median or mean average speed of connected premises in that postcode was less than 2 Mbit/s. If a postcode had an average/median speed of less than 2 Mbp/s, all premises in it were assumed to be at risk of experiencing low speed. This calculation differs from the “Infrastructure Report, 2012 update,” where a postcode was considered not to have standard broadband if any connected line in it experienced a speed below 2 Mbit/s. The underlying data file on broadband speeds provided information on the number of residential and small-business lines with a broadband connection in each postcode. This information was not provided for those postcodes with only large businesses or where there were fewer than three residential or small-business lines for reasons of confidentiality. The restriction to postcodes with a minimum of three connected lines also explains the slight discrepancies in the number of premises with superfast-broadband availability."

Ofcom note on changes to how 2 Mbps coverage is measured (our emphasis)

The previous system of labelling a postcode as slow if data suggested one property was slow could be seen as giving a pessimistic figure, or from the view of people living in those areas it could be seen as a more realistic figure. Certainly the new system is politically more pleasing when you want to report a figure of 99.99% can get 2 Mbps broadband. In fact Ofcom has started to say 'Headline download speeds of at least 2 Mbit/s are available in almost all premises in the UK.'.

Even with the new system almost halving the number of properties not meeting the Universal Service Commitment figure, the situation is still not just a rural problem, there are 24,131 urban properties in Greater London listed as not meeting the 2 Mbps target. Interestingly while the talk around the BDUK and USC debate is often about rural areas, of the 1.2 million properties across the UK that cannot get 2 Mbps, some 423,751 are in semi-urban locations, 661,361 in rural locations and 156,725 in urban areas. To give an insight into the split that Ofcom is using for rural/urban, 35% of the UK population live in urban areas, 51% in semi-urban and 14% in rural areas.


Posted by Dixinormous over 4 years ago
A real pleasure to be one of the 156,725.
Posted by 21again over 4 years ago
> 'Headline download speeds of at least 2 Mbit/s are available in almost all premises in the UK.'

It's one thing having the router syncing with the exchange at 2Meg but it's another being able to download at that speed :p
Posted by WWWombat over 4 years ago
I'm trying to get my head round this "cheat" for saying how much of the UK has 2Mbps, by choosing how to gather stats for each postcode.

But it struck me... if they have the data to work out a mean or median speed for a postcode, they have all the speeds anyway, so why gather stats at the intermediate postcode level at all?
Posted by mikejp over 4 years ago
We must be careful about where this 'data' is coming from. Locally, until I whinged at Ian Livingston, the BT availability checker was 'promising 4mb' to all the houses here who could only get dial-up. After I whinged they are now 'promising' 750kb - closer to the truth, but still far from accurate, with speeds of 30kb common. The figures I have seen for the odd BT postcode are often wildly over-optimistic when you compare them to the Intercai-Mondiale survey.

Posted by mikejp over 4 years ago
Part II:
If this same inaccurate 'performance' is being used we need to be aware. Would the '4mb folk' on my exchange have been put into the '>2mb box' had I not whinged? Surely we need some sort of nationwide 'survey' to let customers actually tell us the real sync speeds after connection?
Posted by New_Londoner over 4 years ago
But Ofcome does the real world speed testing through its sample of (IIRC) around 2000 premises using dedicated hardware. This avoids user-induced problems like not using the master socket, connecxting via wi-fi, running the test on a slow pC etc.

DOing some sort of user-driven speed testing woudl lead us to the sort of nonsense published monthly by uSwitch. If you took decisions based on this you might as well burn the money!
Posted by mikejp over 4 years ago
How can 2000 be enough to assess the millions of properties involved? Surely asking users for sync speeds is not that unreliable?

All the problems on your list evaporate with sync speed apart from the master socket - and that should be fairly easily catered for with clear instructions, and enlisting the help of the significant number of 'local helpers' to check where speeds are reported as deficient.
Posted by andrew (Favicon staff member) over 4 years ago
People are going down the wrong path, the 2000 is not used for this 95% figure at all.

The 1,200 to 2,000 locations is used for the UK average speed, and some modelling is used to extrapolate. (A big debate to had there).

This assessment today is on line connection speed data supplied by providers, which gives them a LOT more data points to work with.

Just stopping a myth from starting.
Posted by mikejp over 4 years ago
"This assessment today is on line connection speed data supplied by providers" - my point, exactly!
Posted by herdwick over 4 years ago
Not really mikejp, you were talking about the availability checker estimate. OFCOM are using data off connected lines. There is no 30 kbit/s broadband sync speed.

Makes sense not to base categorisation on a single outlier, more sound statistically. Otherwise the guy near the exchange feeding his modem through the wrong side of a filter gets to condemn the whole postcode sector.
Posted by andrew (Favicon staff member) over 4 years ago
Actually the availability checker does have some level of loop back data, as BT Wholesale used to and probably still do feed back connection data into the database.
Posted by Plankton1066 over 4 years ago
They need to adjust these figures, as last week I could get 2.4 M/b now only 1.7 M/b.I suspect that the crows have pecked of the gutta-percha and tarred hemp insulation.
Posted by camieabz over 4 years ago
From a certain perspective, if the target was 4 Meg universal, we would be fairly confident of 2 Meg universal being achieved.

When the 2 Meg target was set we were in a world of 240p or 360p vids on YouTube as the norm. Now we're expecting 480p to 720p (or higher).

2 meg is not sufficient. Websites are heavier too. JQuery is heavy duty, and present on many sites. We should be looking at shifting the target regularly to being a set minimum of a typical range of products (i.e. 8 Meg as the current minimum). If the broadband speed range shifts upwards, so should the target USC.
Posted by WWWombat over 4 years ago
Agree with @camieabz there - the USC target should be shifting upwards.

Otherwise those on "basic" service will be forever having to beg the council for the next Mbps. They'll get treated the same as the potholes on the roads - once in a blue moon when extra funding arrives.
Posted by mikejp over 4 years ago
herdwick - you missed it! 30kb is dial-up.

Also "OFCOM are using data off connected lines." read Andrew "The 1,200 to 2,000 locations is used for the UK average speed"

To my mind this is like taking your car to a garage for a service and then having the garage fill in the customer satisfaction survey.
Posted by themanstan over 4 years ago
It's all very well and good saying there should be a USC that is faster. But, how exactly is this to be acheived? Are all ISPs with infrastructure required to provide this or just a couple (VM/BT)? Because in reality most of them could if they wanted to, but simply don't because of their business model. And from whom do the infrastructure providers recoups the costs? Do you give them an absolute monopoly to recover costs or do you retain wholesale?
Posted by andrew (Favicon staff member) over 4 years ago
USC is very different to a USO.

So no requirement on providers, but simply put a best efforts push by local authorities. Which with no funding for London and USC levels leaves a big who and how question?
Posted by themanstan over 4 years ago
Indeed....however people are quoting it like USO and have the same expectations. The reality is USC has no teeth and they should have no expectations.
Posted by herdwick over 4 years ago
OFCOM have connection speed data from ISPs that is mapped and accessible to postcode level, providing there re enough users per postcode to allow it to be displayed. The 1200-2000 locations are the samknows monitoring boxes which can give some understanding of typical speed tests vs sync speeds.

There are *two* datasets.

We don't need an end user with a speedtest to tell us what the DSLAM can tell us accurately.
Posted by rogerjowett over 4 years ago
what a pack of lies
you are using broadband to hack our computers and set fire to homes
i got less than 3% of the advertised connection speed and bills for over £400 a quarter - you hacked my computer so that it was unusable you are criminals and guilty of manslaughter
itu t v44+IPR=921600 v92 connects at what speed? 1% of its capability? crooks the lot of you
data usage -minus all the malicious software you have transferred to our computers?
Posted by philbr over 4 years ago
There needs to be some kind of reality check here. SKY say I should get 2meg. In fact it is now less than 1 meg. Where do the statistics come from? If from the suppliers then that leaves some reason for scepticism! As a domestic user it seems that the truth is not out there for those of us not in the 'big cities'.
Posted by 21again over 4 years ago
At the end of the day the government and the guys in the dark suits simply want mass suitable BB that allows forms to be sent, filled in and returned online, anything else is a bonus :p
Posted by AspieMum over 4 years ago
If the postcode it uses for its avaerage is the first half of the postcode they would think I can get about 3 to 5MBs but our estate can't get 2MBs but only if you use the whole postcode will you spot that.
Posted by AspieMum over 4 years ago
If they want that 21again they need to allow you to save part way though considering you'll need to access your bank account and so on while trying to complete the form on less than 2MBs (unless they ensure everyone can get 2MBs people will have to manage on less than 2MBs) with your kids having to use the internet to do homework & the connection dying on you from time to time.
Posted by andrew (Favicon staff member) over 4 years ago
@aspiemum it is the full postcode.
Posted by dragon1945 over 4 years ago
I used to get 2MB. Speed test a minute ago showed 1.81 MB - but that was after I rebooted the Router. Just before that I was getting 1.3 MB. Sky offered me Entertainment Extra. My Provider offers a TV Package. Wonderful. My BB won't even support BBC I-Player. Copper cables outside over 47 years old. Fibre? Not in my lifetime.
Posted by jonahfish over 4 years ago
i look out across the river to Canary Wharf almost a stones throw away every day and wonder why i can only receive 1.5 MBs max. Is it something i said to BT?
Posted by andrew (Favicon staff member) over 4 years ago

It is because the phone wiring follows the old routes and the exchange is a long way from you.

In the regeneration in the 1980's so long as phone worked no-one cared about distance.
Posted by andrew (Favicon staff member) over 4 years ago
@dragon1945 Age of copper is not that big a thing, look at the copper stuff dug up that is 1000's of years old, still pretty good condition.

For ADSL it is distance to the exchange, if you attenuation is under 60dB then you need to do some checks at the test socket and ensure Sky do not have you on a high target margin.
Posted by mikejp over 4 years ago
According to OfCom, my post code has a 'top' speed of 21mb with an 'average' of 4. Since we are 4km+ from an ADSL2+ exchange, and I don't think anyone gets more than 5mb, what use are these figures?

See post #4
You must be logged in to post comments. Click here to login.