» ISP News » 

The UK Top 7 Fastest Big Fixed Line Home Broadband ISPs – March 2014

Wednesday, April 2nd, 2014 (11:18 am) - Score 1,303

The average home broadband ISP download speeds for the largest seven Internet access providers in the United Kingdom managed to reach 20.94Mbps in March 2014, while uploads hit 4.79Mbps. As usual BT and Virgin Media continued to hold the top spots for downstream performance.

It’s very important to highlight that there’s been a big change this month because we’ve decided to experiment by using Ookla’s data from Speedtest.net for our monthly summaries. In the future this will allow us to expand and cover more ISPs but for this month we’ll retain a little consistency by focusing upon only the largest mainstream providers (TOP 10), although for a comparison we’ve added Zen Internet into the mix.

Top 7 Big UK ISPs – Average Download Speed (Megabits per second)
1. Virgin Media – 42.60Mbps
2. BT – 25.41Mbps
3. PlusNet – 21.16Mbps
4. Zen Internet – 20.15Mbps
5. TalkTalk – 12.65Mbps
6. Sky Broadband – 12.38Mbps
7. EE – 12.27Mbps

Top 7 Big UK ISPs – Average Upload Speed
1. BT – 7.44Mbps
2. PlusNet – 6.51Mbps
3. Zen Internet – 6.15Mbps
4. Virgin Media – 4.91Mbps
5. Sky Broadband – 3.29Mbps
6. EE (Orange) – 2.94Mbps
7. TalkTalk – 2.29Mbps

Please take anecdotal data like this with a big pinch of salt. Every home is different and performance can be affected by all sorts of issues, many of which are beyond the ISPs ability to control (e.g. slow wifi or poor home wiring). We do not consider the above data to be a reliable barometer for individual users but it can help to highlight general changes in the market.

In addition, Ookla’s data attempts to reflect the fastest sustainable throughput performance by dropping a large chunk of the slowest and a smaller slice of the fastest results for each ISP. This may be useful for certain comparisons but it can also skew the results, especially since slower connections tend to exist in rural areas. The impact is most noticeable for upstream speeds, which we believe to be too high (we might not use their data for upstream in the future).

On top of that not everybody takes the fastest speed available from their ISP and the latest “super-fast” connections (FTTC, FTTP etc.) can have a disproportionate impact compared with the older and slower copper ADSL2+ lines etc.

Leave a Comment
8 Responses
  1. Avatar Steve Jones says:

    Rather than manipulate the average through dropping “outliers”, it would be better if they published percentile type information. That is the median together with quartiles and possibly some of the deciles. This overcomes the known problem that averages can be highly misleading when a distribution is highly skewed. For example, a single users on a 300mbps FTTP connection could have a wholly disproportionate effect on the average giving completely the wrong impression of overall speeds. Percentile type stats give an idea of distribution.

    Of course, it may still be necessary to remove true outliers where “impossible” results appear, but that’s very different to artificially manipulating averages.

    1. Avatar Steve Jones says:

      For those who want to know why the arithmetic mean is a wholly inappropriate metric for characterising heavily skewed distributions (like broadband speeds) this is just one of many explanations.


      Means should be kept for data that is normally distributed.

      nb. there is a real technical problem with manipulating medians, in that you cannot combine separate sets of figures apart from going back to the original stats (that is you can’t take two different sets of data with a computed median and then combine them to produce a combined median). However, there is an alternative approach and that is to use a series of pre-defined “bucket” ranges and counting the number of observations that fall in each. Then you can combine values providing the bucket values match. Not quite percentiles, but an excellent way of looking at QoS targets like x% of observations are better than a certain threshold value and how that changes over time.

      This boring, but important stuff, dates back to my days of having to measure and manipulate response time characteristics of 100s of millions of computer transactions over long periods of times. Personally I doubt there’s a single area of public policy which isn’t misrepresented through a combination of misleading stats and the public (and the media’s) woeful understanding of basic statistics.

    2. Avatar DTMark says:

      These are good points.

      I think for the numbers to have any real meaning they need breaking down to a degree which can’t be achieved.

      Firstly, split by connection technology (3G, 4G, cable, ADSL, VDSL) – this might just be possible. Then split by residential/business.

      Then you can get a couple of tables of raw outright speeds and performance. What it doesn’t accurately reflect is what speeds are *possible*, only what is achieved. For example you could have someone with ADSL who could have 4G or cable and faster speeds.

      Next you’d need to split by package speed. Don’t see how this can be done.

      And for DSL, you need the IP Profile. Now you can see how each ISP performs versus how they *could* perform. This would expose, for example, congestion.

      Then express the throughput as a percentage of the maximum throughput. So for cable, say, 30Mpbs of an up to 30Mbps package = 100% – relevance – you get 100% of the headline speed, with VDSL @ 30Mbps on an up to 80Meg package = you get 38% of the headline speed. This would be useful in ascertaining how technologies perform versus advertising.

      The raw data as shown here is of some interest, but in the context of any one property is fairly meaningless in terms of which ISP might perform best for you.

  2. Avatar No clue says:

    Maybe produce a chart next month MarkJ of various organisations data Ookla, neflix index, and the previously used Broadband.co.uk as well as any others you see fit and place them side by side to see how the speed figures and placing of ISPs compare.

  3. Avatar RLP says:

    These results are incredibly unreliable.

    Too many people will run speed tests over WiFi and get terrible results purely because their device is not connecting at the correct speed, let alone the WiFi slowing things down.

    Also the servers used by Ookla vary hugely. I have also found that different computers provide different results on the same connection.

    I have a connection with full 80/20 connection speeds. To illustrate some of the above statements please look at the results below that I have taken from the same computer on different servers. The last was taken a few days ago:




    Old result:

    The White box which is available from SamKnows is about the only reliable speed indication which should be used in a report like this.

    1. Avatar Raindrops says:

      Perhaps run your test when you are not uploading, that normally helps LMAO

    2. Avatar RLP says:

      Strangely enough raindrops I wasn’t uploading anything when I ran the tests this morning.

    3. Avatar George says:

      That must be a fault then because there is no other way you would still be getting 75Mb on your downstream and only 1.2Mb on the upstream. The only way that happens is a fault or you are using your connection in some manner when performing the test, which would make more sense given the results you provided.

Comments are closed.

Comments RSS Feed

Javascript must be enabled to post (most browsers do this automatically)

Privacy Notice: Please note that news comments are anonymous, which means that we do NOT require you to enter any real personal details to post a message. By clicking to submit a post you agree to storing your comment content, display name, IP, email and / or website details in our database, for as long as the post remains live.

Only the submitted name and comment will be displayed in public, while the rest will be kept private (we will never share this outside of ISPreview, regardless of whether the data is real or fake). This comment system uses submitted IP, email and website address data to spot abuse and spammers. All data is transferred via an encrypted (https secure) session.

NOTE 1: Sometimes your comment might not appear immediately due to site cache (this is cleared every few hours) or it may be caught by automated moderation / anti-spam.

NOTE 2: Comments that break our rules, spam, troll or post via known fake IP/proxy servers may be blocked or removed.
Cheapest Superfast ISPs
  • Hyperoptic £22.00
    Speed 50Mbps, Unlimited
    Gift: Promo Code: HYPER21
  • Plusnet £22.50 (*36.52)
    Speed 36Mbps, Unlimited
    Gift: £60 Reward Card
  • Onestream £22.50 (*27.99)
    Speed 45Mbps, Unlimited
    Gift: None
  • xln telecom £22.74 (*47.94)
    Speed 66Mbps, Unlimited
    Gift: None
  • Vodafone £22.95
    Speed 35Mbps, Unlimited
    Gift: None
Large Availability | View All
Cheapest Ultrafast ISPs
  • Vodafone £26.00
    Speed: 100Mbps, Unlimited
    Gift: None
  • Virgin Media £27.99 (*51.00)
    Speed: 108Mbps, Unlimited
    Gift: None
  • Hyperoptic £29.00 (*35.00)
    Speed: 150Mbps, Unlimited
    Gift: Promo Code: HYPER21
  • TalkTalk £32.00 (*39.95)
    Speed: 145Mbps, Unlimited
    Gift: None
  • Giganet £35.00
    Speed: 200Mbps, Unlimited
    Gift: None
Large Availability | View All
The Top 20 Category Tags
  1. FTTP (3050)
  2. BT (2886)
  3. FTTC (1835)
  4. Building Digital UK (1815)
  5. Politics (1777)
  6. Openreach (1710)
  7. Business (1544)
  8. FTTH (1351)
  9. Mobile Broadband (1345)
  10. Statistics (1321)
  11. 4G (1158)
  12. Fibre Optic (1109)
  13. Wireless Internet (1081)
  14. Virgin Media (1072)
  15. Ofcom Regulation (1069)
  16. EE (767)
  17. Vodafone (747)
  18. TalkTalk (717)
  19. Sky Broadband (704)
  20. 5G (630)
Helpful ISP Guides and Tips

Copyright © 1999 to Present - ISPreview.co.uk - All Rights Reserved - Terms , Privacy and Cookie Policy , Links , Website Rules , Contact