Broadband News

CAP rules that at least 50% of customers will need to get advertised broadband speeds

We now have an answer on broadband speeds in adverts, ISP websites, package listings etc they will need to feature the median download speed that customers receive during the peak time period, with the peak period defined as 8pm to 10pm.

“There are a lot of factors that affect the broadband speed a customer is going to get in their own home; from technology to geography, to how a household uses broadband. While we know these factors mean some people will get significantly slower speeds than others; when it comes to broadband ads, our new standards will give consumers a better understanding of the broadband speeds offered by different providers when deciding to switch providers. We continually review our standards to make sure they reflect consumers’ experiences, the technology available and the evidence base to make sure our standards are in the right place. Following extensive research and consultation, we hope our new standards will improve customer confidence in future ads.

Director of the Committees of Advertising Practice, Shahriar Coupal

We doubt anyone is going to be opening the celebration champagne over this news, but are sure some will celebrate this news as the result of years of campaigning but there are not many that can claim our position as innovators in letting the public test the speed of their broadband connections for over 15 years and how things have changed since back in 2002 people were worrying about differences in performance of 60 Kbps between 1st and 10th place in our monthly round up.

The other options such as a range of speeds were considered as a solution to the speed problem by CAP but it was felt a range might confuse the public who mistake this for the much more important personalised speed range given once they start enquiring about broadband with the various providers. Peak time speeds have been adopted for the simple fact that since its peak time this is also when most people want to use their broadband and with the range in performance that we see sometimes we would agree with CAP that quoting a 24 hour average has potential to mislead, particularly given the time its taken to arrive at this decision point.

There is one type of advert aimed at consumers where the median download speed can be missing and that is your lifestyle adverts that just like today omit minor technical details like how fast the service is. Also no new requirements on upload speeds needing to be mentioned.

Nothing is going to change immediately as providers have until 23rd May 2018 to comply with the new rules and importantly this only applies to residential broadband services. The residential aspect is interesting as moving forward it means some providers will still be selling up to 38 Mbps VDSL2 services, when others will be selling 30 Mbps average speed services.

In adverts you will most likely not see the word median appearing but the word average and the wording of the new guidance is such that if a provider wants to they can do things like talk about the speed that 75% or some other figure achieve, in other words so long as the figure is what 50% or more achieve at peak times.

So what will this mean for adverts, what figures are we going to see? Hard to say for sure, as while we do track peak time speeds, these are not split into specific products and the 8pm to 10pm window, we normally work on a 4 or 6 hour wide window for peak time, but with some rejigging of the analysis software we hope to be able to give better guideline figures. Adjusting our methodology to give some backup figures compared to what the providers release is very important, as without any independent verification and analysis it would be relatively easy for a broadband provider to game the system to gain a 2 to 3 Mbps advantage over a competitor.

  • Peak time ADSL technology speeds are such that we expect adverts to talk of 6 to 9 Mbps download speeds.
  • VDSL2/FTTC up to 38 Mbps products are likely to drop into the 24 Mbps to 30 Mbps region
  • VDSL2/FTTC up to 52 Mbps products are only sold in volume by BT Consumer and with speed boosts still ongoing, we don't feel able to predict until another month or two of data has been seen.
  • VDSL2/FTTC up to 76 Mbps products will probably end up talking about 45 to 55 Mbps average speeds.
  • Cable broadband, while we have average speeds of 126 Mbps for the Virgin Media 300 Mbps service, we expect Virgin Media to still advertise speeds well above this, if we have to place a bet the 300 Mbps tier will shift to a 200 Mbps average. This is down to a mixture of factors including things like measuring just between router and a location on the providers own network and improvements to their access network.
  • Full Fibre/FTTP/FTTH/FTTB providers while not having any of the variations due to connection speed still have the shared nature of consumer broadband to contend with.

We do have some worries about what may happen to the market, and these are things like providers filtering customers so that they only take on those with lines that are going to be favourable to their median results, this filtering may also extend to profiling customers who are likely to be very heavy data users. The filtering may be just steering potential customers to alternative products, but if customers are refused service they find that their choice of providers packages is restricted, this will need to be something Ofcom watches out for. Another oddity is how the average speed is arrived at and how does CAP, the public and Ofcom know that a provider is testing a representative sample of all the line lengths and areas of the UK e.g. if a provider is known to perform poorly in one town at peak times avoiding testing in those areas will be a big temptation.

There is a second string to the guidance from CAP it has a response to its review of 'fibre' in broadband advertising and it is that the way the word 'fibre' is used in advertising can largely remain.

The ASA has also today announced the outcome of its review of “fibre” claims in broadband advertising. Having considered all of the evidence provided during the review, it has concluded that it is not materially misleading to describe broadband services that use fibre-optic cables for only part of the connection to consumers’ homes as “fibre broadband”.

The ASA has published research which it commissioned as part of its review, carried out by the agency Define, which found that: “fibre” is not one of the priorities identified by consumers when choosing a broadband package; that consumers did not notice “fibre” claims in ads; that consumers, when probed, saw it as a shorthand buzzword to describe modern, fast broadband; and that consumers did not believe they would change their previous decisions after the differences between those services and broadband services that use fibre optic cables all the way to the home were explained to them.

However, in recognition of the performance differences between different types of broadband service, including between ‘part-fibre’ and ‘full-fibre’ services, the ASA advises that:

  • as has always been the case, ads should not describe non-fibre services as “fibre”;
  • ads should not state or imply a service is the most technologically advanced on the market if it is a part-fibre service;
  • ads should make performance claims for fibre services (part- or full-) that are appropriate for the type of technology delivering that service, and should hold evidence to substantiate the specific claims made; and
  • specifically, ads should refer to speed in a manner that is appropriate for the technology, including by having due regard to CAP’s new guidance on numerical speed claims.
CAP guidance on 'fibre' in broadband adverts

We have included the bulk of this section from the CAP press release to ensure no nuances are lost and we are certain that the passion felt over the fibre issue is going to mean some people are upset. While we understand why some are likely to be upset the research seems pretty clear that once you get outside the broadband industry bubble that so many of us live and work in the public perception of all this can be very different.

I'm delighted to see that CAP is finally changing the way broadband speeds are advertised. Headline "Up to" speeds that only need to be available to 10% of consumers are incredibly misleading - customers need clear, concise and accurate information in order to make an informed choice. We have been fighting for this for some time now, and it's a great victory for consumers. The clarification on the way fibre services should be advertised is a welcome step in the right direction, and I hope the ASA will keep the matter under review.

Minister for Digital Matt Hancock on CAP speed guidance

ISPA supports today’s change to rules governing the advertising of broadband speeds as an important way of providing consumers with clear and accurate information. The new rules, alongside existing steps of providing speed information at point of sale, the ability to exit a contract and switching, mean consumers have a number of tools at their disposal to make an informed choice about which ISP or broadband service is right for them.

Commenting on CAP guidance, ISPA Chair Andrew Glover

In final summary, your existing services are not going to be changing at all. If you are getting 36 Mbps today at peak times from our speed tester and you switch provider and their technology and performance is the same as your current provider then there will no reason why you won't get the same 36 Mbps again, even though the adverts may change from up to 38 Mbps to an average of 30 Mbps after May 2018. A caveat we don't believe we have mentioned in the news before, is that if you switch between partial fibre providers (VDSL2/FTTC technology) that the Dynamic Line Management systems may reset and some speed improvements may vanish for a few weeks - the recent upgrades for TalkTalk and PlusNet customers to increase upload speeds have triggered this quirk for some but generally the lost speed does come back in a couple of weeks.

Comments

@thinkbroadband About time

  • @banger_x
  • comment via twitter
  • 23 days ago

By May 2018 VM should be in a position to have to change nothing. Their test results are bizarrely poor to TBB but on SamKnows metrics and to speedtest.net are okay. They're very close to median 350Mb on that tier if not already there. The question is whether or not they will continue to make the investments in capacity and ensure it can be done.

  • CarlThomas
  • 23 days ago

What isn't going to happen is for them to reduce the speeds they advertise. That's pretty much the only thing VM have, download speeds. The marketing people will be jumping up and down and marketing inevitably drives things.

There are ways for VM to reduce the load on their network too.

  • CarlThomas
  • 23 days ago

What isn't going to happen is for them to reduce the speeds they advertise. That's pretty much the only thing VM have, download speeds. The marketing people will be jumping up and down and marketing inevitably drives things.

There are ways for VM to reduce the load on their network too.

  • CarlThomas
  • 23 days ago

@thinkbroadband There is extraordinary room for this to change the behaviour of providers, especially xDSL provider… https://t.co/V3eTUvdm2C

  • @ultrafastcarl
  • comment via twitter
  • 23 days ago

Maybe this is why BT have upgraded a load of Infinity up to 55Mbps users to an 80/20 service. They may be the only ISP that is able to advertise median speeds near to what the product offering is advertised as because they will have a number of users who are above the product speed.

  • ian72
  • 22 days ago

Biggest concern is risk of further reducing choice of users with long lines off market A exchanges who already have to pay a surcharge to many operators due to the lack of competition.

  • brianhe
  • 22 days ago

On allowing "Fibre broadband": Seems like a sound, evidence-based decision. I think they're right about "fibre" now being used as a shorthand buzzword.

On speeds: Using a median is a choice they can make, but /whatever/ number they allow is a poor substitute for a personalised estimate. If your line is longer than the median, then the speed is less than the median, and you'll still complain. This just means fewer people complain ... but not zero.
...

  • WWWombat
  • 22 days ago

The biggest negative is that the advertised speed (still heavily qualified, with reference to estimates) is now less associated with any kind of technology level. Which will make people more dependent on those shorthand buzzwords.

Where we've gone here is to be less precise. Less accurate.

Mat Hancock tries to take it on the chin.

  • WWWombat
  • 22 days ago

This is good news hopefully this will encourage more money being spent on getting nodes closer to people on long lines . It would be great to see lets say a 300 meter cap on all line lengths if BT wish to keep using aging copper/ali drop cables

  • 2doorsbob
  • 22 days ago

@2doorsbob - that won't happen. What will happen is more of what is already happening in Market A areas, ISPs will choose not to provision any broadband at all in areas that they can't meet their advertised speeds.

  • sheephouse
  • 22 days ago

Using average speeds in advertising would work if all connections were FTTP and speed loss was due to contention which is something ISPs can fix. Any move to using advertised speeds for DSL services will just encourage ISPs to reject customers were the speed is expected to be low, in much the same way as schools are reported to discourage some children from taking exams that they might fail.

Should the word fibre just be used for FTTP connections that don't suffer from DSL distance problems, it appears not.

  • Michael_Chare
  • 22 days ago

We've had the product twisting on FTTC ever since it was introduced, with BT Infinity not supplied to anybody with a BT Wholesale estimate below 15Mbps connection speed.

Any lower than that can only have "Faster Broadband with Fibre".

Plusnet do similar with their 76/19 product. If someone wants to upgrade from 38/1.9 to it because they would get significantly higher upstream and accept that they wouldn't get any downstream improvement, they are told no. (Whether this will change as they upgrade 40/2s to 40/10 Openreach products we will have to wait and see).

  • uniquename
  • 22 days ago

The obvious way to game this for ISPs using xDSL services is to produce two packages with a fairly nominal price difference, call them something meaningless like hyper and turbo. Thes advertise the high speed "hyper" service. When it comes to ordering the product, if the supplier estimate is that is can't meet some threshold required to hit their desired median value for "hyper" then just offer them "turbo", even if the settings are exactly the same.

For the sub-SF crew, then I expect yet another product (and BT does have a low speed VDSL offering not well advertised).

  • TheEulerID
  • 21 days ago

I should also add I will be interested to see how the ASA are going to police this and what evidence will be demanded as to actual throughput figures achievable at peak times.u There are vast numbers of variables, not just at the ISP with contention ratios, but is it single thread, multiple thread? What about problems with customer configs - poor WiFi performance, poor wiring and so on.

Sync speeds on xDSL (& raw DOCSIS speeds) are easily monitored. Actual peak throughput achievable, quite another matter.

  • TheEulerID
  • 21 days ago

Well I'm feeling much better, instead of getting 1/9 of the advertised speed I will be getting 1/6. :-|

  • burble
  • 21 days ago

Wi-Fi will be totally ignored, i.e. testing will be between router and ISP network and massively multi-thread

  • andrew
  • thinkbroadband staff
  • 21 days ago

You might think that after 10-15 years of 'distance from exchange' or 'distance from cab', there would be a more intuitive way for consumers to estimate their likely min / med / max speeds.

The BTW line test is hardly comprehensive to the average punter. Too much confusing data of impacted speeds etc.

Then ISPs could then give estimates on a per-address basis. Of course that's not in the interests of ISPs, and UK consumers are too used to sucking up what is served.

  • camieabz
  • 21 days ago

@Camieabz Try ordering a broadband package most do give estimates once you supply address details, a small thing called the Ofcom Broadband Speeds Code of Practice.

Maybe you are ordering from providers who are not signed up to the code.

  • andrew
  • thinkbroadband staff
  • 21 days ago

That's OK, but why shouldn't it be higher, like 75%. They all do a check and tell you what your speed is likely to be once you have it, so I don't see why it shouldn't be higher. I don't like any of the ways that Broadband is advertised in any case. A Bus driver doesn't say, I will take you roughly to where you want to go, or if you buy a can of beans, you expect it to reasonably full, excepting for the infamous "Contents may settle during transit" exclusion, which in most cases is valid.

DOes that make sense, or am I just expecting too much from a broadband advertisement?

  • TrevorSP
  • 21 days ago

Plusnet charge more at small exchanges, so it could well be that they have fewer customers at these exchanges. Are the average telephone cable lengths at these small exchanges longer than those at larger urban exchanges, leading to average speeds being slower?

  • Michael_Chare
  • 21 days ago

Line lengths are generally longer in rural areas, you can tell if you compare coverage e.g. urban areas the fibre based figure is a LOT closer to the superfast figures.

So yes a danger rural areas may be discouraged from buying some services.

  • andrew
  • thinkbroadband staff
  • 21 days ago

Broadband adverts need to know where you live and could then offer a personalised estimate and that would be a lot better.

There are questions around adverts that cover things like 'new mega service now available in your area', should they feature a more localised speed estimate or can they rely on a national median.

  • andrew
  • thinkbroadband staff
  • 21 days ago

Andrew, how localised can an advert be? If I lived just over a mile up the road my 4MB service would be 80MB.

  • burble
  • 21 days ago

Sarcasm doesn't translate too well online :)

  • CarlThomas
  • 21 days ago

How local, well that's for ASA to CAP to determine, but it is a real question particularly with peak time speeds as this can vary widely from one town to another.

The anger around not getting the advertised average is probably going to be more obvious than the previous anger at the 10% rule. Realising you are not in the top 10 people can sort of accept but knowing you are below average will irk people more.

  • andrew
  • thinkbroadband staff
  • 20 days ago

Post a comment

Login Register