Skip Navigation

Speed test results for December 2014
Wednesday 07 January 2015 17:24:00 by Andrew Ferguson

2014 may have finished, but that means we can look at the speed test results for the final month of 2014.

Two main flavours of our speed test are available, our traditional flash based test or our new test that is quicker to run, and has recently had a test history section integrated into it.

Broadband Speeds December 2014 - Large Sample Size
Provider Median Download Mean Download Median Upload Mean Upload % of tests at superfast speeds
Virgin Media 41.7 Mbps 50.2 Mbps 3.5 Mbps 5.4 Mbps 63.5%
BT Consumer 16.4 Mbps 23.1 Mbps 2.7 Mbps 5.4 Mbps 34.2%
Zen Internet 14.2 Mbps 24.3 Mbps 1 Mbps 5.5 Mbps 33.2%
PlusNet 12.6 Mbps 21.6 Mbps 0.9 Mbps 5.4 Mbps 30.9%
KC 9.6 Mbps 31.3 Mbps 0.7 Mbps 5.5 Mbps 16%
Sky 8.6 Mbps 13.4 Mbps 0.9 Mbps 3.6 Mbps 13.9%
TalkTalk 7.6 Mbps 14 Mbps 0.7 Mbps 1.8 Mbps 17%
EE 6.8 Mbps 12.1 Mbps 0.8 Mbps 2.3 Mbps 12.4%

Median speed is the average where 50% of people get this speed or faster. Mean speed is the arithmetic mean and the difference to the median speed reflects the level of popularity of higher speed products at a provider.

Broadband Speeds December 2014 - Reduced confidence due to smaller sample size
Provider Median Download Mean Download Median Upload Mean Upload % of tests at superfast speeds
Hyperoptic 94.2 Mbps 211.1 Mbps 81.5 Mbps 164.6 Mbps 77.4%
EE/Orange Mobile 15.7 Mbps 20.2 Mbps 1.8 Mbps 3.4 Mbps 22.2%
Relish 9.3 Mbps 12.2 Mbps 1.7 Mbps 2.6 Mbps 0%
Entanet 9.1 Mbps 20.1 Mbps 3.4 Mbps 8.7 Mbps 20%
IDNet 9 Mbps 20.8 Mbps 0.9 Mbps 4.9 Mbps 25%
T-Mobile 7 Mbps 13.2 Mbps 1.1 Mbps 3.3 Mbps 8%
Vodafone Mobile 7.8 Mbps 11.6 Mbps 1.1 Mbps 1.9 Mbps 3.1%
Eclipse 6.6 Mbps 12.8 Mbps 0.7 Mbps 2.6 Mbps 8%
O2 Mobile 6.6 Mbps 11.3 Mbps 1.2 Mbps 2.1 Mbps 10.2%
Three 5.2 Mbps 7 Mbps 1.3 Mbps 1.6 Mbps 0%

We are able to analyse the technology used for each test, which with the current push to get people to buy a fibre based service via the Government funded Superfast Nation adverts is a better indicator of what sort of difference the fibre based services make.

Technology Speeds December 2014
Technology Median Download Mean Download Median Upload Mean Upload
ADSL/ADSL2+ 4.5 Mbps 5.5 Mbps 0.7 Mbps 0.7 Mbps
FTTC 33.5 Mbps 34.5 Mbps 8 Mbps 9 Mbps
FTTH/P 94.7 Mbps 148.6 Mbps 42.7 Mbps 97.7 Mbps

The step change in speeds that FTTC can offer is clearly visible, as is the even higher performance difference seen from the combined might of the FTTH/P/B providers in the UK (B4rn, Hyperoptic, Gigaclear, Openreach and others). The speed for ADSL/ADSL2+ saw a major change in speed in 2011, but has generally remained static with slight rises overtime as ADSL2+ roll-outs via LLU and WBC continue, but to a smaller extent than 2011. The speed of FTTC as a technology has been surprisingly static since Spring 2013 when Openreach rolled out the up to 76 Mbps product option. Two common broadband theories are that the FTTC roll-outs are only to those areas where people can already get fast ADSL2+ speeds, but our speed test data does not support this assertion.

FTTC is far from a perfect technology, but an interesting observation is that bottom 10% of people on FTTC have a maximum speed of 11.3 Mbps which is double the median and mean for ADSL/ADSL2+. Showing that even though FTTC is distance limited the improvements seem to be sufficient to support two to three HD video streams at the same time.


Posted by chilting about 1 year ago
My exchange in West Chiltington,was one of the first to benefit from FTTC BDUK last July. Initial speed estimate from BT wholesale was between 8 and 13Mbps on a line just under 2km from the cabinet. I was actually getting about 5.5Mbps. Now the speed has fallen to 3Mbps with a new BT wholesale predition of between 1 and 2.9Mbps for the line. I use bonded lines, my second line is currently even slower at just 0.6Mbps but it averages about 2Mbps. As the cabinet fills speeds reduce as shown by the BT wholesale speed predictions.
Posted by andrew (Favicon staff member) about 1 year ago
The providers checkers have used positive feedback for some years.
Posted by chilting about 1 year ago
Hardly positive. The speed predictions on other lines closer to the exchange have also fallen. BT seem to move the goalposts to accommodate their own failings!
Posted by TheEulerID about 1 year ago
I suspect that estimate have gone now there's experience with cross-talk (and, maybe, due to more self-installs). It might be that vectoring could ameliorate the cross-talk issue.

VDSL over 2km lines is clearly unsatisfactory. ADSL from cabinets would be better (but ANFP issues). However 1km+ lines really need a better solution.
Posted by TheEulerID about 1 year ago
One thing I would favour trialing is that ADSL from exchanges via selected enabled cabinets is disabled and all DSL services operated from the cabinet (suitably redefined). (All VDSL with a new freq plan would be better, but require customer modem swaps).

I believe I read somewhere this was happening in New Zealand. Easier in market 1 exchanges of course.
Posted by AndrueC about 1 year ago
Is it possible to determine what %ge of FTTC users have chosen a 40Mb/s service when 80Mb/s might be available? Likewise on VM?

Would be interesting to know how much of those figures is down to people choosing not to bother with the fastest option. On FTTC especially.
Posted by WWWombat about 1 year ago
Forcing all ADSL variants from the exchange onto the cabinet (though still as ADSL) would bring dual benefit of shorter line and a potential reduction in the power-masking at the cabinet - and Yes, NZ does this.

Obviously "GEA over ADSL2+" is a step towards this, but the best benefit from long lines (from the cabinet) would come from removing those power masks.
Posted by WWWombat about 1 year ago
On VM, anyone can get 150Mbps (or, indeed, any speed).

Judging by the TBB Blog, around 6% of VMers have chosen a package speed of above 100Mbps, and perhaps 50% above 30Mbps.

It'll be almost impossible to say for FTTC, as the attainable speed would have to be, at best, a statistical guess.
Posted by gerarda about 1 year ago
"but an interesting observation is that bottom 10% of people on FTTC have a maximum speed of 11.3 Mbps which is double the median and mean for ADSL/ADSL2+."

My understanding is that Openreach will not normally accept orders for FTTC where the estimated speed is less than 4Mpbs, so bottom 10% will average faster than if that restriction was not in place.
Posted by chilting about 1 year ago
The rock bottom seems to be 2Mbps. They are still touting for FTTC business from my some of my neighours when the range is 1Mbps - 2.9Mbps but others just a few yards away are just offered ADSL.
Posted by paulmjstone about 1 year ago
not sure about the new speedtest results. Same machine powerline connected.
Old 32/18Mb
New 54/17Mb

New one only shows the x6 speed. So it all depends on how you measure?

Old :
Posted by andrew (Favicon staff member) about 1 year ago
When quoting the old speed test you omitted the httpx6 which is basically the same between the two when I look at the tests

On the old one the tbbx1 and httpx6 should be the same if not then suggests wifi, pc, or other hardware issue or congestion somewhere
Posted by Blackmamba about 1 year ago
Hi Chil.
Last week in Hants on Headley Down exchange customers over 1 mile from the new Cab (FTTC) have not been offered SFB and they are receiving only ADSL service. The change over orders are running at 5 working days you may have noticed that the Openreach New list has not been updated since September on the exchange services. I have a feeling that fibre is going to be pushed passed the 1 mile limit.
Posted by Blackmamba about 1 year ago
Hi Gerarda and Broadband Watchers.
I think you will find that the customers who are receiving aprox 12 meg down will be over the 1.1 mile from the new Cab (FTTC ). This is very close to the Surrey target (15 meg) at the post code.
You must be logged in to post comments. Click here to login.