» ISP News » 

UPDATE ASA Slap BT Broadband Checker for Misleading Speed Estimate

Wednesday, May 7th, 2014 (7:52 am) - Score 1,449
bt uk isp logo

The Advertising Standards Authority has triggered some confusion after it ruled that an Internet download speed estimate provided by the superfast broadband BTInfinity (FTTC) “availability checker” on BT’s consumer website (bt.com) was “misleading” because the complainant could not receive the speed promoted. The ruling could also have repercussions for other ISPs.

As most ISPreview.co.uk readers know the act of estimating a customer’s potential broadband speed over a copper telephone line is somewhat of a guesstimate, which doesn’t always reflect reality. ISPs can make an estimate based on the length of the copper line and other factors, although these can easily be undone by other problems on the line such as damaged infrastructure, poor home wiring, external electromagnetic interference and a whole host of other things (many of which may not be within the ISPs ability to control). Accuracy of such estimates is further eroded when a postcode check is the only option.

Suffice to say that there usually has to be an acceptance of the fact that the estimate given may not be entirely reliable and your actual speeds could vary, which is a problem inherent to the various underlying ADSL and VDSL technologies being employed (the signal degrades with distance). In addition, Ofcom’s 2010 Voluntary Broadband Speed Code requires that member ISPs provide customers with an “estimated access line speed“, so it’s not something they can simply stop doing.

Ofcom’s code also means that any customer who suffers a significant fall in performance, below their original estimate, should within the first three months of a new service be able to exit their contract penalty free.

As a result the ASA’s latest ruling risks causing confusion in the market after one of BT’s potential customers challenged whether the “23Mb-33Mb Download speed range” estimated for their telephone line was “misleading and could be substantiated“. An example of the outcome style is given below, albeit with a different customers line.


In response BT said that they were “unable to provide detailed analysis of non-BT customers’ lines because that was dealt with by Openreach and that due to their organisational structure, they did not have access to data held by Openreach“. Meanwhile BTOpenreach explained how their estimates worked.

Openreach on Broadband Speed Estimates

Openreach explained that to estimate line speeds, they carried out detailed and large scale statistical analysis of possible speeds and the quoted values seen by the complainant were typical of what could be achieved by the vast majority of superfast broadband lines. They said the speed estimate ranged from the 80th to 20th percentile for similar phone lines, therefore, 80% of end-users could achieve the quoted speeds. They said they had checked the complainant’s line and confirmed that it lay outside the statistical range, and due to a variety of reasons, the complainant would be unable to achieve the quoted speeds.

However the ASA has effectively ruled that estimates are not good enough if the customer cannot receive what is predicted and they told BT to “ensure their availability checker provided accurate information“, which could have significant ramifications.

ASA Ruling (REF: A14-257752)

The ASA understood that the Ofcom report related to consumers in general. We also understood that speeds could be checked via BT’s ‘availability checker’ by a specific phone number or by house number and post code and therefore considered that consumers would expect that any resulting download speed claims would be accurate for their address. We considered that the download speeds quoted on the BT website would be a material consideration for consumers when deciding whether or not to take up BT’s superfast fibre optic BT Infinity product.

Because the website included a download claim related to a specific address which was not available to that consumer, we concluded the ad was misleading.

The outcome is an unusual one and there’s also a distinct lack of background detail concerning why the estimated speed was not available to the customer in the first place, which strikes us as being a crucial question that has not been fully explained.

At the time of writing BT’s availability checker still exists and we have hailed the operator in the hope of finding out what changes they have had to make in order to satisfy the ASA’s ruling. On the flip side we expect that some pure fibre optic providers (and possibly Virgin Media too), which are able to offer fairly reliable predictions about performance due to their stronger infrastructure, will be very pleased.

UPDATE 8th May 2014

BT’s consumer division has now given us a reply, although it’s a bit cagey on the subject of how they intend to rectify the ASA’s concern.

A Spokesperson for BT Consumer said:

Ofcom has defined rules for all ISPs regarding how they quote broadband speeds. BT has taken care to abide by those rules and regularly reviews communications to ensure we offer customers clarity. We currently give customers the range of speeds they can expect and make it clear that this is an estimate.

Despite this, we note the ASA decision relating to the presentation of the speed results and we will work with them to ensure that they are also happy with the way we describe the results from our Infinity speed checker. Speed estimate data is supplied to all ISPs by Openreach, in line with Ofcom’s Code of Practice. However, it’s important to note that speeds can be affected by factors outside our control, such as internal wiring and the modem or PC being used by the customer.”

The ASA’s ruling was extremely vague and this will no doubt make life harder for BT in examining how best to adjust their results. From our perspective the solution would be to offer a more reflective /wider range of estimated performance, while also doing more to clearly explain the caveats of that estimate.

Add to Diigo
Tags: ,
Mark Jackson
By Mark Jackson
Mark is a professional technology writer, IT consultant and computer engineer from Dorset (England), he is also the founder of ISPreview since 1999 and enjoys analysing the latest telecoms and broadband developments. Find me on Twitter, , Facebook and Linkedin.
Leave a Comment
26 Responses
  1. Chris Conder

    Its about time too! The BT ‘checker’ is so flawed it’s useless. It has muddied the waters for far too long. Glad that ASA is starting to realise this and hope they investigate further instead of believing the hype.
    Bring on the fibre. Moral and optic.

  2. Andy

    BT said that they were “unable to provide detailed analysis of non-BT customers’ lines because that was dealt with by Openreach and that due to their organisational structure, they did not have access to data held by Openreach“

    They just LOVE hiding behind Openreach.

    • Gadget

      Andy, if you were Sky, Talktalk or any other ISP customer of Openreach would you be happy if Openreach gave your information to a competitor? As it stands Openreach give the same information to every ISP, but the ISP holds data on its own customers which it can use.

  3. TheFacts

    Gives correct numbers for me.

  4. MikeW

    Interesting to see the 80/20 explanation.

    I guess the only solution will be to change the lower bound from the 20th %ile to the zero %ile – which will make it “up to”.

    I know ASA say they want it to be more accurate, but we know the net result will be to make results less accurate combined with a more accurate get-out clause.

    • I think the point is less dense than that.
      If you supply a ‘tool’ that claims to do something – it should do that thing in an accurate and predictable manner.

      Hammers need to be capable of delivering their function and to advertise a hammer that is actually a heavy screwdriver is misleading.

    • gerarda

      Dropping the bottom 20% does not explain why BT’s download range for our village is 1-3.5 when the reality is 0.0-0.7, except if all lines over a certain length are bundled together as “similar lines”

  5. Adrian

    “If you supply a ‘tool’ that claims to do something – it should do that thing in an accurate and predictable manner.” so you just need to make the “tool” say “this provides a range based on 20th and 80th percentiles of similar lines”, and bingo you have a tool that does exactly what it says it does, and an advert that is not misleading.

    We see that lines are normally within the ranges estimated, and often faster than the upper limit given.

    However, ISPs contracts with BT for FTTC allow the line to be cancelled and put back to DSL or whatever at no charge to the ISP if the line fails to meet the forecast (i.e. minimum) at install, so there are already ways to address this if the estimated is inaccurate for any reason on an individual line.

    To address the advertising, the advert just needs to say what the estimate meands, i.e. that it is in fact an “estimate”.

    This is another occasion where I disagree with the ASA.

    • DTMark

      In the event that the estimated speed is not met..

      And the customer elects to cancel outright and go to cable or 4G instead..

      – Does the customer get a 100% refund?
      – What does the ISP get billed?

  6. PhilT

    Shall I see what the ASA make of b4rn.org.uk/services ?

    No “up to”. “two way video without freezing”. Both fall foul of previous rulings.

    Claiming 1000 MBits/s symmetrical – that can’t be substantiated, especially the upload.

    • No clue

      There is no “UP TO” mentioned in BTs results shown in the news story either.

      An ISP only has to demonstrate 10% of its lines are capable of the speed advertised, i suspect 10% of BR4Ns lines can do 1000Mb if thats what people order and pay for.

      10% of BT lines capable of 75-80Mb though……. I don’t think so.

    • TheFacts

      Can you get 1G transfer over a 1G link? What about the overheads?

    • Raindrops

      You can not get 80Mb sync over FTTC if you include overheads either.

      Then we have the constant BS news items submitted by BT employees talking about tech like G.Fast FTTdp which quote speeds and the product does not even exist.

      I suggest instead crying you run along and tell your precious BT to sort itself out.

  7. Phil

    My FTTC went live last february as bt checker say I can get estimated 79.9 / 20.0 but soon after sign up with plusnet, a month later it now changed to 77.0/19.9 and now after ip profile with both ends bt and plusnet I getting 72/17. So, it should be 72/17 on bt checker because of cabinet filled up, cross talk and ip profile are involved too from DLM.

  8. Given it is an estimate, I find the ASA ruling laughable, if it were not so serious (their inability to understand).

    OFCOM clearly understands what an estimate is, and why only an estimate can be given.

    As for Phil (79.9/20 down to 72/17) – perhaps there also needs to be some wording to cover the fact that changes over time can impact any estimate. I see nothing too bad with his figures – perhaps Phil can comment on whether he’d have decided “not to bother” if the initial estimate had been 70/15, or if he went ahead, would he have considered anything above the estimate to be a bonus?

    For me (with estimate of 33.5 Mbps), the estimate does make a difference, because some ISPs will offer a lower speed but impose a usage limit too, as if the two should be “linked” in some way, and yes, the simple fact is that I won’t pay for an “up to 80” service which can cost quite a bit more than the “up to 40”.

    • DTMark

      It matters most especially when the estimate is on the threshold of what the customer is prepared to accept. ADSL is one example – less than 6 Meg down and it’s practically useless for modern applications like streaming.

      Another example: we have 4G at 25/20 and apparently will have access to VDSL eventually. Estimate for line length is 18/3. If I were prepared to accept that degree of downgrade I’d want to see all of that. If I put up with the slow downstream, less than 3 upstream – forget it, it will be useless to me.

      It isn’t the case that the estimate can’t be achieved, it could. Swapping out pairs or line bonding for instance. It is not “impossible”.

      I can’t really see why anyone accepts that with most things, the usual analogy being the “pint” of beer, the customer should get what they pay for but with broadband it’s somehow different and it’s acceptable to supply “whatever” at minimal effort, but the customer is expected to pay for that for 12 months or indeed pay anything at all if it falls short. Surely the ISP should pay the customer for wasting their time?

      No, I didn’t get an “estimate” with 4G. But then it doesn’t come with a minimum commitment, either and that tech has issues that fixed line should not have – for instance “natural” contention.

      Buyer beware of xDSL tech – perhaps. Though sadly, it may be the customer’s only option. And I don’t think a “get out of jail free” card with the ISP wringing their proverbial hands is appropriate or acceptable.

    • FibreFred


      How you or anyone can use a “pint of beer” analogy to compare to (x)DSL is beyond me.

      Possibly its used because it makes no sense at all and somehow proves something when it actually proves nothing at all.

      Unless of course your pub serves pints in whatever receptacle they have available and it may not always be a pint , it could be a half or even just a cup but that is just a physical limitation of what they have available and will charge you a pint regardless of what they can fit in the receptacle

    • gerarda

      @fibrefred your argument would stand if the estimates were reasonably accurate but to be out by a factor of several hundred percent is not acceptable nor is it acceptable to offer a service which is premium priced but delivers only the same as the standard price.

    • No clue

      Were the BT fan boys the other day not blabbering on about how accurate BTs checker is?

      Seems even more funny now.

    • DTMark

      I don’t see the problem with the analogy. The contract is “a pint of beer for fixed price £x”.

      Not “some quantity of beer for fixed price £x”.

      From what I have seen the Wholesale checker is reasonably accurate until it comes to failing or aluminium circuits when it’s useless. The latter is easily capable of halving the estimated performance.

      The idea that the ISP can then wash their hands of the situation and expect the customer to pay them for 12 or 18 months despite only having delivered half of the estimate seems laughable – as I said, it ought to be the ISP paying the customer in such scenarios for wasting their time.

      Estimates can only ever be estimates, that’s the failing with trying to do broadband over such ancient mixed metal line plant.

      Just so long as the customer can get a full refund, then the problem is mitigated somewhat.

    • Raindrops

      “The idea that the ISP can then wash their hands of the situation and expect the customer to pay them for 12 or 18 months despite only having delivered half of the estimate seems laughable”

      BT are worse than that. I clearly a few months back had a line fault, Upload rate dropped from 19Mb to 8-10Mb. Download speed lost around 20Mb. At sign up my estimated upload rate was a FULL 20Mb.

      Upon complaining i like so many others on BTs own forums were told that was within “tolerance” i being stubborn though continually pestered them until an engineer came out, fixed the line problem and then reset the line. Its been back to my normal speeds ever since.

      BT will just get away with whatever they can, they are lousy when it comes to customer service. You have to pester and fight them at every turn, its about time someone put them in their place, ofcom are gutless. So well done to the ASA.

      Ill be watching the checker from now on for false estimates on lines and if any of them are massively different ill certainly follow it up to the ASA, i suggest anyone ordering FTTC which do not get the speed BT “estimate” do the same thing. The only way BT will learn is if they are punished.

    • FibreFred

      “I don’t see the problem with the analogy. The contract is “a pint of beer for fixed price £x”.”

      But that isn’t the same as your broadband contract, you get an estimate of what will be delivered and then the contract states you’ll get “up to” a certain amount.

      For the analogy to work (it simply doesn’t) your contract with the barman would have to go something like:-

      “This drink you are after, I reckon I’ll be able to give you half a pint but you could end up getting a full pint if you are lucky, just depends…”

    • MikeW

      Try a pub in Holland. I reckon that is exactly the contract the barmen work to there…

    • Raindrops

      If BT or any ISP can not give anything near accurate estimates of what speed a customer will get then why do they continue to quote figures. Are they like a pikey selling magic beans which claims it has approximately x amount of beans in the package?

      Maybe the answer is to change things see they have to quote a minimum speed, which for everyone taking FTTC better be at least 25+Mb as that is what the government expect from BTs “superfast fibre” is it not?

      The problem was and i quote”one of BT’s potential customers challenged whether the “23Mb-33Mb Download speed range” estimated for their telephone line was “misleading and could be substantiated“.” So it appears according to what the law wants superfast to be classified as BT tried to sell this customer something that was potentially slower than superfast.

      Can FTTC give everyone 25+Mb or not???????????

  9. No Clue

    “Can FTTC give everyone 25+Mb or not?”

    Nope and it never has been able to despite what the worker trolls of BT peddle.

IMPORTANT: Javascript must be enabled to post (most browsers do this automatically). On mobile devices you may need to load the page in 'Desktop' mode to comment.

Comments RSS Feed

* Your comment might NOT appear immediately (the site cache re-syncs periodically) *
* Comments that break our rules, spam, troll or post via fake IP/proxy servers may be blocked *
Cheapest Superfast ISPs
  • Hyperoptic £19.00 (*22.00)
    Up to 30Mbps, Unlimited
    Gift: None
  • NOW TV £20.00 (*35.99)
    Up to 38Mbps, Unlimited
    Gift: None
  • Vodafone £21.00 (*23.00)
    Up to 38Mbps, Unlimited
    Gift: None
  • TalkTalk £22.50 (*33.50)
    Up to 38Mbps, Unlimited
    Gift: None
  • Plusnet £24.99 (*33.98)
    Up to 38Mbps, Unlimited
    Gift: £50 Cashback
Prices inc. Line Rental | View All
*Javascript must be ON to vote*
The Top 20 Category Tags
  1. BT (2084)
  2. FTTP (1448)
  3. Broadband Delivery UK (1401)
  4. FTTC (1355)
  5. Openreach (1077)
  6. Politics (1066)
  7. Business (971)
  8. Statistics (851)
  9. Fibre Optic (806)
  10. Mobile Broadband (774)
  11. Wireless Internet (720)
  12. Ofcom Regulation (717)
  13. 4G (663)
  14. Virgin Media (653)
  15. FTTH (633)
  16. Sky Broadband (503)
  17. TalkTalk (472)
  18. EE (428)
  19. Security (337)
  20. Vodafone (318)
New Forum Topics
Helpful ISP Guides and Tips

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