Broadband News

Broadband Speed Test Results for February 2016

The world of broadband speed testing is gradually developing a split that has the FTTH providers occupying the top slots and what is interesting is the decision making process people go through on places like our forums when asking about whether upgrading from a fairly average ADSL2+ connection speed to any of the faster technologies. It is clear that if the UK wants to put itself at the top of the International speed tables that the price of superfast and ultrafast connectivity needs to drop significantly, the killer app of being able to stream iPlayer, Netflix, Amazon Video etc is all that the majority care about.

The main table has been expanded to 25 entrants so that we can get a few more of the household names into the chart. Venus has ousted Hyperoptic from the top slot, though if we ordered the table on any other metric Hyperoptic would be in the top slot. Venus offers pure fibre to places like flats in the Barbican.

The 25 Fastest UK Broadband Providers in February 2016
(ordered by median speed)
Smaller providers without enough geographic data samples are not included
Provider Median Download Mean Download Median Upload Mean Upload Download Speed of top 10%
Venus 90.6 Mbps 84.9 Mbps 72.1 Mbps 88.4 Mbps 145.8 Mbps
Hyperoptic 55.1 Mbps 118.8 Mbps 43.1 Mbps 110.9 Mbps 367 Mbps
Gigaclear 51.9 Mbps 98.4 Mbps 51.6 Mbps 99.9 Mbps 336 Mbps
Virgin Media 41.1 Mbps 49.3 Mbps 5 Mbps 6 Mbps 100.9 Mbps
Vodafone Broadband 29.8 Mbps 31.1 Mbps 8.9 Mbps 9.5 Mbps 60.7 Mbps
Wessex Internet 29.2 Mbps 36.7 Mbps 4 Mbps 12.9 Mbps 58.7 Mbps
Metronet UK 27.9 Mbps 45.3 Mbps 19.6 Mbps 31.4 Mbps 84 Mbps
Keycom 21.9 Mbps 36.3 Mbps 18.1 Mbps 27.8 Mbps 95.2 Mbps
IDNet 16.9 Mbps 25.7 Mbps 1.3 Mbps 6.7 Mbps 66.3 Mbps
Zen Internet 16.9 Mbps 25.7 Mbps 4.2 Mbps 7.4 Mbps 65.6 Mbps
AAISP 16.5 Mbps 27.4 Mbps 1.6 Mbps 7.7 Mbps 73.8 Mbps
BT 14.5 Mbps 19.6 Mbps 1.7 Mbps 4.9 Mbps 40.8 Mbps
EE Mobile (3G/4G) 14.3 Mbps 19.3 Mbps 1.3 Mbps 4.2 Mbps 42.3 Mbps
Vodafone Mobile 13.4 Mbps 16.1 Mbps 2 Mbps 3.3 Mbps 33.6 Mbps
Relish 12.5 Mbps 16.2 Mbps 1.7 Mbps 2.3 Mbps 36 Mbps
Claranet SOHO 12.3 Mbps 24.8 Mbps 1.5 Mbps 11.6 Mbps 68.5 Mbps
Europasat 11.2 Mbps 11.4 Mbps 0.1 Mbps 0.2 Mbps 18.7 Mbps
Plusnet 9.8 Mbps 16.6 Mbps 0.9 Mbps 3.9 Mbps 38.4 Mbps
TalkTalk 8 Mbps 13.5 Mbps 0.7 Mbps 1.7 Mbps 34.2 Mbps
Sky 7.9 Mbps 11.8 Mbps 0.8 Mbps 2.5 Mbps 28.9 Mbps
O2 Mobile 7.4 Mbps 12.5 Mbps 1.2 Mbps 2.8 Mbps 30.9 Mbps
Three 7.2 Mbps 12.2 Mbps 1.1 Mbps 3 Mbps 29.1 Mbps
KC 6.9 Mbps 21.1 Mbps 0.7 Mbps 4.6 Mbps 67.9 Mbps
EE (ADSL2+/FTTC) 6.8 Mbps 11.9 Mbps 0.8 Mbps 2.6 Mbps 33.5 Mbps
Eclipse 5.6 Mbps 11.1 Mbps 0.7 Mbps 2.3 Mbps 33.8 Mbps

For those providers that had enough tests to talk about but did not make the table their median download speeds are Demon 5.5 Mbps, TalkTalk Business 5.1 Mbps, BT WiFi 4.6 Mbps and Post Office 2.4 Mbps.

Of those providers where the sample size is too small to feature in the main table still worthy of a quick mention is B4RN (median down 56.1 Mbps), IFNL (33.6 Mbps), Ask4 (27.9 Mbps) and WightFibre (27.1 Mbps).

Large Provider Fibre Based Connection Speed Tests February 2016
Provider Median Download Mean Download Median Upload Mean Upload
FTTC Overall (excludes Virgin Media) 28.4 Mbps (+0.2) 29.6 Mbps 6.7 Mbps (+0.1) 7.5 Mbps
BT 31 Mbps (+0.2) 32 Mbps 8.3 Mbps (+0.1) 8.6 Mbps
EE 29 Mbps (+2.2) 27.8 Mbps 8.2 Mbps (+0.3) 7.5 Mbps
Plusnet 31.4 Mbps (+0.2) 32.3 Mbps 7 Mbps (-0.3) 8.5 Mbps
Sky 24.8 Mbps (+0.2) 24.6 Mbps 6.8 Mbps (+0.2) 6.6 Mbps
TalkTalk 27.5 Mbps (-0.3) 28.4 Mbps 1.8 Mbps (=) 4.2 Mbps
Virgin Media 41.1 Mbps (+0.3) 49.3 Mbps 5 Mbps (+0.1) 6 Mbps
Vodafone 32.8 Mbps (+3.2) 35.2 Mbps 9.1 Mbps (+0.2) 10.8 Mbps

The observant will notice how the Vodafone FTTC speeds are not massively different from the Vodafone Home Broadband speeds and this is down to the reality that the customers the new fixed line broadband ISP is attracting are generally connecting on one of the VDSL2 products.

ADSL/ADSL2+ Connection Speed Tests February 2016
Provider/Area Median Download Mean Download Median Upload Mean Upload
All Providers 5.3 Mbps (+0.1) 6.6 Mbps 0.6 Mbps (=) 0.6 Mbps
BT 4.8 Mbps (+0.1) 6.3 Mbps 0.5 Mbps (+0.1) 0.5 Mbps
EE 4.8 Mbps (+0.2) 6.1 Mbps 0.7 Mbps (+0.1) 0.6 Mbps
Plusnet 5.2 Mbps (+0.3) 6.4 Mbps 0.6 Mbps (+0.1) 0.6 Mbps
Sky 5.3 Mbps (+0.1) 6.6 Mbps 0.6 Mbps (=) 0.6 Mbps
TalkTalk 5.5 Mbps (=) 6.6 Mbps 0.6 Mbps (=) 0.6 Mbps
Rural ADSL 3.8 Mbps (+0.1) 4.8 Mbps 0.3 Mbps 0.4 Mbps


The results for the gigabit-only B4RN just show how heavily home networks influence results.

I vaguely recall there being some capacity issues on a B4RN transit partner that may have been part of the issue but the recollection is vague and I may be mistaken.

  • CarlThomas
  • over 2 years ago

Virgin and the curates egg.
I have the new virgin 200Mb package and am very impressed. Some speeds up to 209 Mb download. Typically between 190 Mb and 205Mb.
However, it drops to 3 to 5 Mb download if I leave my PC constantly connected for a few hours and am forced to restart my connection to get back to the golden numbers again. Is this some sort of "throttling" by Virgin?

  • georgiecoo
  • over 2 years ago

turn off firewall(SPI) on the router some times stops if from going slow

  • leexgx
  • over 2 years ago

Thanks leexgx for your speedy feedback. I don't know how to access the settings to do that. Any help gratefully received. Sincerely, GC

  • georgiecoo
  • over 2 years ago

Always confused by these tables, the main one states "Smaller providers without enough geographic data samples are not included" yet the top 3 are all small providers with very limited coverage. Do they pay to get put in there?

  • Daemon66
  • over 2 years ago

@daemon66 I believe it is down to the number of tests done rather than the area they cover. Many people that get new fast connections do a lot of tests to show how good the connection is. I do not believe TBB take any payments from ISPs to promote them and if they did I believe they would be up front about it.

  • ian72
  • over 2 years ago

Of course if it's the number of tests then the system is open to being "gamed" though initiating automated runs. However, it might be that TBB use more sophisticated systems (like unique IP addresses or cookie tracking) to weed this out. The number of unique users doing tests it probably a better filter than the absolute number of tests.

  • TheEulerID
  • over 2 years ago

I do not believe this is a true reflection as
people only run speed tests hen they have a problem or an upgrade.
Reducing the price of fast superfast will not increase the average speeds, its making it more available that would make the difference.

  • t99del
  • over 2 years ago

There is no payment system to appear in this table. Of the 60+ providers we list perhaps 10 have affiliate based links, and the banner adverts are the other source of income if people click through.

  • andrew
  • thinkbroadband staff
  • over 2 years ago

@TheEulerID We take a number of measures to guard against gaming the system, but you can never be totally bullet proof.

If a provider was persistently gaming things in a provable way then we would not hesitate to out them.

  • andrew
  • thinkbroadband staff
  • over 2 years ago

@t99del This is why if you read through and browse the twitter feed you will find extensive other analysis of the results. This monthly table is just the tip of the iceberg.

Don't believe any other testing body publishes such wide ranging levels of analysis.

  • andrew
  • thinkbroadband staff
  • over 2 years ago

There is no doubt home networks influence results Carl. Also the devices used to test with, many test with ipads which bring the average down. Most folk don't seem to use ethernet any more, its all wifi. There is no problem with transit, and my tests on my fairly good computer on wire are always well over 900Mbps symmetrical. just done this one for you <a href="">
<img src=""></a>

  • cyberdoyle
  • over 2 years ago

The DSLReports test uses multiple servers and as a result will avoid such issues. In your case you used servers in Germany and Dublin.

The average across all ISPs I'd expect to be similarly impacted by WiFi. If anything the MDU providers more heavily due to much higher WiFi network density.

I may be wrong but I definitely recall something about a transit provider having some problems between Manchester and London - this is not in any way indicative of the B4RN network itself, incidentally, which is a very simple set up with little to go wrong.

  • CarlThomas
  • over 2 years ago

'Reducing the price of fast superfast will not increase the average speeds, its making it more available that would make the difference.'

Barely 10% of the UK cannot get a 'superfast' service. I don't see how making it cheaper for nearly 90% would have less of an impact than increasing availability.

Our broadband is quite cheap enough to be honest. Ideally more expensive and less reliance on line rental to cross-subsidise would be nice.

  • CarlThomas
  • over 2 years ago

Chris - are you ready to fear my broadband? ;)

20.3/8.16 megabit/second


  • CarlThomas
  • over 2 years ago

Hi Broadband Watcher.
Over the last six months TBB results have improved but as they are using an average results this does not portray the best service that a customer can receive on a post code it tends to under value it to the ISP,s and customers.
I still think that are giving a good service by including more ISP,s in the list thus generating compertion in the market. Just keep checking on the (6--12 month average results) on post codes.

  • Blackmamba
  • over 2 years ago

@blackmamba If the world could cope with a much larger table we could do top and bottom percentile and quartile speeds too. We squeeze in top 10% on the ISP table already.

  • andrew
  • thinkbroadband staff
  • over 2 years ago

Hi Andrews Staff.
If the ISP is not in your list top 10% I would think they are starting to decline and could be loosing market position so it is up to them to respond via their CEO or top managers to take action. It also works the other way if you drop out of the 10% and not beening advertised in your format.

  • Blackmamba
  • over 2 years ago

@BlackMamba You have misunderstood, look at the top 25 table there is a column named "Download Speed of top 10%" that is what I was referring to.

Nothing to do with top 10% of the market and management or provider size.

  • andrew
  • thinkbroadband staff
  • over 2 years ago

I have evidence that BT Wholesale is not pursuing REIN noise polluters who are lowering ADSL speeds. By not pursuing them they force people onto fibre and because its FFTC it continues to degade fibre speeds.These speed tests do not check or report the presence of REIN noise pehaps they should.

  • SNR12
  • over 2 years ago

@SNR12 I'd be interested to hear how you spot REIN from a TCP level test?

Where REIN is affecting throughput the if a user runs a speed test the lower result will be recorded, but we cannot make a link between their speed test and it definitively being REIN.

  • andrew
  • thinkbroadband staff
  • over 2 years ago

Just tested again Carl. - if you disable the London server you get a higher speed. Germany and Ireland are faster than london. I can feel your pain. London isn't doing good is it? We find Manchester is faster, but closer helps I guess. 2ms ping from there.

  • cyberdoyle
  • over 2 years ago

Post a comment

Login Register