» ISP News » 

The Fastest Big UK ISPs According to SamKnows New Broadband Test

Wednesday, December 5th, 2018 (12:01 am) - Score 7,501

Internet speed analyst firm SamKnows has this week posted the results from a limited run of their new “sensationally accurate” broadband ISP speedtest tool, which was tested last month as part of support for some research that the BBC’s Watchdog TV show was doing. Unsurprisingly Virgin Media came top, with TalkTalk bottom.

Arguably SamKnows is better known for its direct method of testing, which delivers very accurate results by installing specially modified routers in homes to test the connection performance (Ofcom’s uses this for their study of broadband speeds). The biggest drawback to this approach is cost and as a result the end-user sample size tends to be quite small (no more than a few thousand premises).

Meanwhile the rest of us tend to conduct a Broadband Speedtest by using one of the many web-based solutions, although SamKnows has claimed that the accuracy of those can sometimes struggle to produce reliable results for the new generation of Gigabit (1Gbps) or even multi-Gigabit “full fibre” (FTTP) ISPs.

Excitingly, we’re testing faster and faster connections. But importantly, we’re finding that traditional web tests struggle to handle such speeds, and their performance can vary depending on the browser or computer its running on. So, six months ago, we decided to create a SamKnows Web Test, making it sensationally accurate,” said the organisation.

NOTE: The results shouldn’t be taken to reflect actual network availability (service take-up and other factors – see further below – can have a big impact).

The first challenge came on 9th November 2018, when the BBC raised some concerns about the accuracy of such tests at “high speeds” and SamKnows agreed to let them use their new cloud-based test for the programme. Starting at 8pm on 9th November, the tester received data from over 15,000 speed tests by the following day.

The sample size from such a limited run is too small to offer a wide overview of the market but it did revealed how the biggest five broadband ISPs faired (i.e. BT, Plusnet, Sky Broadband, TalkTalk and Virgin Media). As you’d expect Virgin Media came top for download speed (superior cable network), with TalkTalk marked as the slowest (including for latency, although they did well for upload speeds).

In Figure 3 we can see a table summarising the results of some of the largest ISPs in the UK. The mean is a simple average over the entire dataset, while the median and the 80th percentile shows the speed that 50% and 80% of the tests achieved respectively.


At this point we have to step in with a little reality check. Nothing that SamKnows does will ever be able to overcome all of the inherent problems with web based speedtests, which can be affected by all sorts of issues, such as slow WiFi, limitations of the testing / server itself, package choice, home network congestion at the time of testing, remote network / ISP peering problems and hardware limits etc.

Meanwhile anybody who wants to give the new test a try can do so over here, although it’s really more intended for testing Gigabit or faster connections and probably won’t tell you anything new on slower lines. Our own results suggested that it was no less or more accurate than any of the other tests, but alas we don’t have a 1Gbps line to play with 🙁 .

NOTE: At the time of writing we could not be certain whether the test was single or multiple thread based (or both), which does have an impact.

Share with Twitter
Share with Linkedin
Share with Facebook
Share with Reddit
Share with Pinterest
By Mark Jackson
Mark is a professional technology writer, IT consultant and computer engineer from Dorset (England), he also founded ISPreview in 1999 and enjoys analysing the latest telecoms and broadband developments. Find me on Twitter, , Facebook and Linkedin.
Leave a Comment
16 Responses

    No speed test is accurate. It’s all a lie. It might show you what your line is theoretically capable of but it isn’t realistic nor does it represent what you can expect to get out of your connection on a daily basis. Why is that ? Well, the majority of ISPs do some kind of QoS / Throttling / Bandwidth or Traffic shaping. They priorities speed tests to make themselves look good but it’s a meaningless test. Also check where the endpoint is. Not talking about this article or samknows but many speedtest (cough .net cough) tests will send you to a node that exists on YOUR OWN ISP which is also meaningless since you probably are interested in internet speed to the INTERNET rather than within the network of your own ISP. These tests are HEAVILY manipulate it and if you believe these numbers then you have been duped. What one ISP gets in terms of speed has no bearing on what it will be like to use that ISP on a daily basis on the internet.

    1. Oggy says:

      I don’t agree to everything. Even though a lot of ISPs will and do do that, i wouldn’t say that all of them are in the same boat. While doing an online speed test, i get shown a speed at which i do download. Pretty much exact. Sure, you got to understand the bit – byte conversion, but all it takes it to divide the speed by 8.

  2. New_Londoner says:

    I think the mean and 80th percentile upload for TalkTalk may have been reversed? Surely the mean can’t be bigger than the 80th percentile?

    1. Mark Jackson says:

      The table is direct from SamKnows but I agree, IMO that specific figure looks to be the wrong way around. I’ll update if / when they correct the original.

    2. EndlessWaves says:

      The mean can be higher if there are less than 20% of users on much faster connections. Say, TalkTalk’s York FTTP trial.

    3. gerarda says:

      Simple example: 10 connections 1 gigabit fttp, 8 80mbps FFTC, 1 5mbps ADSL

      mean is 164.5mbps, 80th percentile 80mbps

  3. Mark says:

    KCom in Hull have a 400mbs service live now and have announced a 900mbs package too.
    They say they are 14 years ahead of the rest of the UK.

  4. Jason says:

    The UK isp prpviders could adequately give 1gbps to cities and busier towns, villages. They just don’t want to as their is no growth and years down the line investment would need to be higher. It also incentivises corporate broadband which is in the 1000-10,000s every month. If they opened this up to consumer for say 100 pound a month then they close that corporate market.

    1. Joe says:

      That’s not really true. The big diff between biz/consumer bb isn’t speed but service levels. There maybe some firms at the margins that it would be tempting to but not any serious players…

    2. cro barr says:

      But it would get to a point that RES charges x RES customers , will exceed the revenue from a handful of corporate customers and their higher charges, also when the price comes down (which it should) then more businesses will come on board too.

      Stack em high , sell em cheap , richer sounds methodology.

      Also i inherently disagree that business customers should pay more for a service , except when it comes to warranty and service guarantees.

      I’m not against people choosing to pay more (like communities) who decide they want to speed up roll out by forking out installation costs

      Running costs are not high once infrastructure is in place , and RES prices should reflect that.

      Also 5G should be available at home with unlimited bandwidth tariffs (but they won’t do that will they as they got rid of almost all unlimited tariffs from mobile contracts )

      5G has so much potential and it will be wasted

      Also fibre to the premesis has so much potential but I fear the same as others here that the uk is well behind the curve on roll out, for whatever reason.

  5. Stephen Martin says:

    I was getting 65mbps down and 15mbps up until a couple of months ago with Plusnet, now it’s dropped by 30-40%,when I asked Plusnet about this they eventually said that BT did it by throttling thr connection.

    1. New_Londoner says:

      That makes no sense given that BT owns Plusnet. I suggest you point out that your contract is with Plusnet (a separate legal entity) and ask the service team to fix your service, assuming that it’s now below your speed estimate – the one provided when you contracted for broadband.

    2. Joe says:

      I don’t buy that explanation. Crosstalk perhaps going from an unfilled cab to a full one

  6. Roger melly (the man on telly) says:

    I’m with BT
    I get 74mbps down
    I get 18mbps up
    I’m 2km from exchange
    I’m close to the green fibre box on my rd

    Tested both Speedtest.net and using a download app for real internet speeds and mine is pretty much spot on each time

    BT have great low latency Fibre, awesome for gaming where latency can make a huge difference to dying in game 🙂

    1. Joe says:

      Assuming you’re on fttc then its the cab not exchange distance that matters.

  7. Neil - SamKnows says:


    I can confirm there are a handful of upload results getting 1Gbps upload (presumably from a York TalkTalk 1Gbps user as per EndlessWaves’ suggestion), but the vast majority are in the 10-20Mbps bracket. So I can confirm the numbers are correct.

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 £15.00 (*25.00)
    Speed 50Mbps, Unlimited
    Gift: None
  • Vodafone £19.50 (*22.50)
    Speed 35Mbps, Unlimited
    Gift: None
  • NOW £20.00 (*32.00)
    Speed 36Mbps, Unlimited
    Gift: None
  • Shell Energy £21.99 (*30.99)
    Speed 35Mbps, Unlimited
    Gift: None
  • Plusnet £22.99 (*38.20)
    Speed 36Mbps, Unlimited
    Gift: £70 Reward Card
Large Availability | View All
New Forum Topics
Two ISPs on one line?
Author: unhappyguy1673
Brawband in Glasgow
Author: Simmie
Contract coming to an end.
Author: LeeWillRockYou
B535-232 + Poynting vs B818
Author: Conspiracy
Cheapest Ultrafast ISPs
  • Hyperoptic £20.00 (*35.00)
    Speed: 150Mbps, Unlimited
    Gift: None
  • Vodafone £24.00 (*27.00)
    Speed: 100Mbps, Unlimited
    Gift: None
  • Community Fibre £25.00 (*29.50)
    Speed: 300Mbps, Unlimited
    Gift: None
  • Gigaclear £27.00 (*59.00)
    Speed: 500Mbps, Unlimited
    Gift: None
  • Virgin Media £27.00 (*51.00)
    Speed: 108Mbps, Unlimited
    Gift: None
Large Availability | View All
The Top 20 Category Tags
  1. FTTP (3505)
  2. BT (3010)
  3. Politics (1925)
  4. Building Digital UK (1919)
  5. FTTC (1882)
  6. Openreach (1822)
  7. Business (1680)
  8. Mobile Broadband (1469)
  9. Statistics (1406)
  10. FTTH (1364)
  11. 4G (1271)
  12. Fibre Optic (1167)
  13. Virgin Media (1159)
  14. Wireless Internet (1153)
  15. Ofcom Regulation (1139)
  16. Vodafone (836)
  17. EE (830)
  18. TalkTalk (761)
  19. 5G (760)
  20. Sky Broadband (744)
Helpful ISP Guides and Tips

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