Skip Navigation


ASA publishes research highlighting low understanding of broadband speed
Thursday 17 November 2016 00:10:58 by Andrew Ferguson

Following on from the roaring success of the inclusive pricing rule changes, the UK it appears is one step closer to a big change in how broadband speeds are advertised.

The research conducted by GfK for the ASA, looked at different ways of expressing the speed of broadband services that are advertised, and the four key points the ASA has chosen to highlight from the research are:

  • Speed is an important factor for a significant proportion of consumers who are making decisions between providers
  • Levels of knowledge and understanding of broadband speeds vary, but it is low overall with many not knowing what speed they need to carry out daily online tasks
  • Most understand that the higher the number in the ad, the higher the speed of the service, but many are unclear on what this means for them and what speed they would likely achieve
  • Crucially, the research shows that despite that uncertainty most consumers believe they are likely to receive a speed at or close to the headline speed claim when, for many, that is not likely to be the case

The research underpins the calls by the ASA for changes to how broadband speeds are advertised and now the Committees of Advertising Practice who maintain the various codes used in advertising has announced a review that will report in Spring 2017.

"CAP welcomes the ASA's research and we'll now begin the process of updating our guidance and publish a response next spring.

The research provides good insights into consumers' understanding of broadband speed claims, but it doesn't identify an obvious alternative way to communicate speeds that would be suitable to everybody's needs. It also tells us that consumers believe that advertising can only do so much, which underpins the importance of detailed broadband speed information being provided elsewhere.

CAP will take these findings and other information into account in its review of the guidance to ensure that broadband providers aren't over-promising on their speed claims.

Director of CAP, Shahriar Coupal

So it looks likely that changes will take place, whether it is an average speed or some other percentile figure is not clear and won't be until Spring 2017. With customers saying speed is important any changes in how broadband speeds are advertised needs to be accompanied by a lot of education so the majority of the public is aware that things have changed.

With our speed testing we already have a very good clue as to the speeds providers would end up using in advertising and in line with most adverts we have removed upload speeds to simplify the tables (requiring advertising to mention upload speeds in the body is something that is well overdue).

Download Speed ranges for various types of broadband connection and providers
Ordered by mean average speed and derived from thinkbroadband.com October 2016 speed test results
Provider Bottom 10%
Mbps
Bottom 25%
Mbps
Median Download
Mbps
Mean Download
Mbps
Top 25%
Mbps
Top 10%
Mbps
BT Infinity FTTP/FTTH 16.6 27.5 43.7 56 70.3 107.7
Virgin Media Cable 8.1 21.5 41.6 50.1 65.4 101.5
BT Infinity FTTC/VDSL2 14 22.9 33.5 35 47.1 55.4
Plusnet FTTC/VDSL2 13.7 22.7 33.3 34.5 43.8 59.8
Vodafone FTTC/VDSL2 14.4 20.9 30.5 32.4 38.8 54.2
TalkTalk FTTC/VDSL2 14.2 20.8 27.6 28.8 36.2 40.2
EE FTTC/VDSL2 10.8 18.2 27.6 27.4 35.3 38
Sky FTTC/VDSL2 10.9 17.7 25.7 26 33.1 37.5
Vodafone ADSL2+ 0.6 1.9 6.2 7.4 11.8 15.7
TalkTalk ADSL2+ 1.3 2.8 5.6 6.7 9.8 14.5
Sky ADSL2+ 1 2.3 5.2 6.7 10.1 14.8
PlusNet ADSL/ADSL2+ 0.9 2.4 5.4 6.6 10 14.8
EE ADSL/ADSL2+ 0.7 2.2 5 6.3 9.4 14.2
BT ADSL/ADSL2+ 0.7 1.9 4.7 6.2 9.1 15
Rural ADSL/ADSL2+ 0.7 1.5 3.6 4.6 6.6 9.9

Identifying precise products users are on can be difficult and the popularity of the up to 76 Mbps services varies greatly from provider to provider e.g. EE has only around 15% of its customers on what looks to be an up to 76 Mbps service, but BT Consumer has 28%, and Vodafone goes even higher at 30%. Given this imprecision it is still possible to arrive at a set of likely average speeds that could be the new advertising speed, up to 38 Mbps FTTC services will range from 26 Mbps to 30 Mbps in adverts and ADSL/ADSL2+ would be at around 5 to 6 Mbps. The up to 52 Mbps VDSL2 service looks likely to be around 40 Mbps and the up to 76 Mbps product maybe 50 Mbps if average speeds are used.

The problem of using the average speed in advertising is highlighted though by the difference in speeds between rural ADSL/ADSL2+ connections and normal ADSL2+ connections, where a new average rule would require a figure around 3.5 to 4.5 Mbps. So any national TV adverts are still giving skewed picture for those users, the difference between rural FTTC connections and the national picture is different, as rural areas appear to be getting a better average by 0.5 to 2 Mbps (varies from month to month) compared to the national average.

Virgin Media stands to do well potentially from the new changes, and while their over provisioning ensures they meet the 10% rule, there are plenty of people who suffer due to local congestion who do not believe the advertising hype already, so we would expect speeds advertised to still change for Virgin Media and an average figure may affect the 200 Mbps and 300 Mbps tiers much more than entry level 50 Mbps service, highlighting various network configuration issues.

So what is the answer? Well as we believe that no matter what figure you show in advertising it never going to be right due to the wide variations in how technology performs in different locations, and even on fixed connection speed services local capacity and regional/national congestion issues can prey heavily on the consumer experience. So the answer is we believe is for advertising to highlight that services will offer a range of speeds so people are aware upfront, and to encourage people check for the personal estimate at their property. With the Ofcom Broadband Speeds Code of Practice a speed estimate/range should be supplied as part of the sign-up process, so in theory most people should have some idea beyond the existing advertised figure, but we'd like to see providers make these checks easier to access, i.e. to be available both as part of the order journey but also in a more visible manner that does not opt you into marketing activities.

The switch to inclusive broadband pricing has at face value being a success, but after years of people being used to extra charges appearing, there is a real risk that those not aware that pricing includes line rental will think broadband prices have short up by just short of £20 since the summer, and a drop to average speeds in advertising may make people think a double whammy has occurred with connections slowing down. A more realistic scenario is a person who is enjoying 36 Mbps from VDSL2 now and sees an enticing advert but since the speed advertised is lower than what they get now they believe their speeds will reduce if they switch and thus it will dampen the switching market and stifle competition between the retail providers and make it very difficult for a new entrant like Vodafone to expand.

The ultimate answer of course is to not mention speed at all in an advert.

The research outlines several options going forward, and given the pressure for change we suspect that staying as we are is off the table.

  • Specify a minimum speed, alas as the research highlights the public will see this as a guarantee and therefore difficult to implement, since consumers can experience zero speed from their broadband if a provider is experiencing very high loads.
  • A speed range, perhaps bottom 25% to top 25%
  • An average speed, but as the research suggests the public take the average to mean half are faster and half are slower, i.e. the median, but generally when providers talk about average speed they use the mean.

There is one more option, adopt business leased line pricing, a minimum guaranteed throughput with a higher burst speed, but how attractive does a 10 Mbps throughput with burst to 175 Mbps service sound?

Update 9:45am The full report and detail of the questions asked of the 84 people across 14 groups are now online on the ASA website.

Comments

Posted by leexgx 20 days ago
but the thing is Virgin media is selling you a 70/150/200/300mb service as its profile limited not distance limited(slightly ignoring congestion for small set of streets and wifi) as ADSL and VDSL\g.Fast is

avg advertising should not apply to FTTP/real fiber and DOCSIS(virgin) But they do need to fix congestion issues faster (with Virgin the FTTN nodes need more cards or split cab when there is congestion instead of ignoring it for a year or 2 or when they add faster DOCSIS technology like 3.1 soon)
Posted by baby_frogmella 19 days ago
ASA again sticking its oar in unnecessarily. If Joe Public has difficulty understanding the words 'Up to 24 mbps' or 'Up to 80 mbps' in the ISPs marketing blurb, then they really need to retake their GCSE/O Level in English. The day will come when ASA will force ISPs to remove any speed figures in their ads and just call it 'broadband internet' in order to stop confusing poor old Joe Public.
Posted by ian72 19 days ago
This is still mixing sync speeds and throughput. The estimates at sign up are based around sync speed expectation. To do throughput you need a standard way of measuring it - so does the ASA/government need to create a standard speed tester for all ISPs (or use the TBB one)? Congestion is variable so what is the advert - is it for what people will see most of the time or is it what they will see at "peak" times or is it the minimum they might see? I still don't see a simple easy way to describe all the variables.
Posted by ian72 19 days ago
@lleexgx - it must apply to Virgin/FTTP/etc. FTTP is going to be different across ISPs as it isn't just the connection to the ISP but how much bandwidth they have in their own networks to cope with demand. FTTP is as open to congestion in various parts of the network as any other technology.
Posted by baby_frogmella 19 days ago
@ian72
Its simple, you just ask the ISP to put the small print at the bottom of their ads, such as

1) Speeds quoted are the maximum speed the customers modem can connect at based on the customers line length to nearest exchange or fibre cabinet.

2) Actual throughput will be slightly less than the connection speeds due to overheads. In ideal line & network conditions the actual throughput will be no less than 80% of modem connection speed.

3) Actual throughput can vary due to line conditions, network performance at peak times, and any fault outside of our control.
Posted by baby_frogmella 19 days ago
(cont'd)
4) The customer is responsible for ensuring that their internal home wiring is optimised for our services. We can offer a Engineer visit for a one off fee of 49.99 where the Engineer can inspect and rectify (where necessary) any issues.

5) Customers can check their current internet speeds by going to www.xxxx.co.uk

Get the above t&c's (and many others) rubber stamped by the Plain English Campaign and then the public has no reason whatsoever to whinge about poor speeds.
Posted by nervous 19 days ago
@baby_frogmella

I tend to agree with you on this point. We shoud stop trying to wrap people in "bubblewrap" trying desperatley to remove any need for them to educate themselves to any danger whether it be financial/physical/mental, isnt that how people learn and how we got to be the dominant species?
Peope will only get more ignorant the more you lead them by the hand.
Posted by nervous 19 days ago
Here is an example of computer automation designed to make an airline pilots job easier but eventually made them ignorant and cost the lives of passengers. http://www.slate.com/blogs/the_eye/2015/06/25/air_france_flight_447_and_the_safety_paradox_of_airline_automation_on_99.html
Posted by nervous 19 days ago
A question,

Why would anyone want an internet connected kettle/washing machine/heating/etc?
Posted by Blackmamba 19 days ago
Hi Broadband Watchers.
The new advertised speeds show up to say on ISP ZEN shows 17 meg on a fixed address where BT shows 15 meg it is possible for both to be correct. This can only be proved untill the line is provided. This must come under the heading buyer beware. The up to results may change over time + or -
Posted by JHo1 19 days ago
I've been wondering that too. I mean you still have to get up to put water in, or bread so what's the advantage? I can only imagine it appeals to a special variety of idiot.

John
Posted by AndrueC 19 days ago
"Speed is an important factor" - what a shame that choosing your provider
doesn't have an affect on speed for half the population.

"most consumers believe" - in almost anything and always assume the best outcome. This is
why snake oil sales remain high.
Posted by AndrueC 19 days ago
And those who base their purchasing decisions primarily on advertising are the biggest fools of the lot. At least do some research before signing up. In my opinion even paying attention to adverts is a bad idea. The best option is to avoid the damn' things as much as possible.

https://en.wikipedia.org/wiki/The_Space_Merchants

That was written in 1952. I took it to heart when I read it in the 1980s.
Posted by Gadget 19 days ago
Not helped by programs like the recent Watchdog item which took speed tests over internal WiFi at what appeared to the be limit of the usable range without any attempt to explain and only suggested new router without even mentioning WiFi extenders or using wired connections where possible all of which are nothing to do the ISP (I'd even include the router spec in that as most new supply these day use halfway decent kit and usually there is nothing other than cast stopping the user upgrading to top-of-the-range one).
Posted by ian72 19 days ago
There are 2 issues for me. First is comparing network congestion between providers. If a provider is good when you first sign up then over-contends the network to reduce costs then I should be able to get out of contract.
The second is line quality. We know that lines deteriorate but there is a fine line between reasonable reduction in speed and actual line fault - and ISPs/OR don't necessarily accept where lines are faulty.
The advertising is gumph and ISPs that suggest they can run a wholesale line faster than another ISP should have the book thrown at them
Posted by ian72 19 days ago
Cont. Speed advertising that is based on physical line capability is pointless when the ISPs are reselling the same OR product. Speed under the ISPs control is what needs to be understood and they need to be able to explain how they ensure their networks are not running too hot or what level of slow down is considered to be reasonable.
Posted by AndrueC 19 days ago
Aye. We even get ISPs like BT and Sky proclaiming that their service is good because of their router's wifi capabilities. Marketroids have been given too much free-rein in the broadband market for too long.

They should be forced to stick to the facts and if there aren't enough facts they should admit it. Make the user do the research.
Posted by Blackmamba 19 days ago
Hi Broadband Watchers.
I would think the (up to speed ) is advertised to the master socket that the ISP can provide over its network at that given time using its own equipment thus every ISP is different. The difference in speed will not be very large over time as technology changes at the post code area so keep checking what service are available on TBB maps.
Posted by andrew (Favicon staff member) 19 days ago
Oddly Blackmamba when you compare FTTC in the same location from different providers generally very little variation, since all sharing the same VDSL2 hardware.

Peak time congestion/load issues are another matter, but our monthly results show generally not an issue on VDSL2, variations between providers explained by product choice.

See https://twitter.com/thinkbroadband/status/798934975391428608 for graphic illustrations
Posted by Michael_Chare 19 days ago
@baby_frogmella. If you look at the authors 1st names what else would you expect?

Posted by jumpmum 19 days ago
What gets me is the sample size, 14 types of user with 6 in each group is hardly statistically significant. Reminds me of adverts with 75% of 60 women recommend xxxx hair shampoo!
Posted by Michael_Chare 19 days ago
Due to the nature of DSL technology connection speeds depend on the location of the customer. So if it is not possible to order a broadband service without being told the expected connection speed, there really is not much of an issue.
Posted by AndrueC 18 days ago
I suppose the issue is that a lot of people go by adverts. I've also found that quite a few think that 'up to' means 'your connection will sometimes run at up to..' rather than the correct contextual meaning of 'your connection will be fixed at some speed up to..'.

And of course no-one ever discusses contention :-/
Posted by WWWombat 18 days ago
I think its time I gave up on having an opinion on this topic.

Every survey seems to come back with one answer: most people are too thick too understand anything about broadband speeds. Education campaigns, and sites like this, seem to do no more than scratch the surface. And, of course, the technology only ever gets more complicated.

We've got to the point that the only viable advert is something like: "We sell broadband. Come and have a look at us".
Posted by Kr1s69 18 days ago
The advertising is trying to be too simple.

For DSL based services just allow advertising at these line speed e.g. 40/55/80 or 8/24 for ADSL and say subject to line length.

All this percentile stuff is silly. We could end up with providers not wanting to accept long lines if it reduces their average or median service.
Posted by burble 18 days ago
I've yet to see any proposal that can overcome the problem of people who don't understand how broadband speeds work. I've spoken to someone who thought changing provider was going to give him a 10x speed increase, he was confused when I pointed out he wouldn't.
Posted by leexgx 18 days ago
i agree with this people just dont understand what they are buying

got a customer right now who is on 1.2mb ADSL but she is on post office broadband, one of the only providers that does not offer FTTC so she is stuck with ADSL for 18 months or change to another ISP and dispute the poor speed miss selling (i ended up setting the laptop up using my 3 4G phone)
Posted by leexgx 18 days ago
but one note ISPs are not always giving the customer there estimated speeds post office CS person did not tell her estimated broadband speed (not that she would of understood it any way) she would of been better of with virgin (already wired) or plus net as she was thinking of going with them (as its only £5 more for FTTC)
Posted by leexgx 18 days ago
(my brother worked for sky retentions for short time and things he was coming out was stupidly untrue like sky has its own fiber network, it taken me like 3 days to get him corrected)

but my original post was that Current advertising is based on the actual Connection speed, not congestion as to why Virgin (DOCSIS or FTTP DOCSIS) and real fiber to the house is not applicable to these changes as majority of people do get the speed profile they are paying for on virgin (most realistically only need the lowest speed virgin 70mb due to limitations of wifi)
Posted by leexgx 18 days ago
but there should be things in place to get virgin to fix Street level congestion at the FTTN node faster (if not have fines in place if they are not corrected within 3 months) as there is no excuses as virgin knows the utilisation of every FTTN node in the UK if its past 50% on the upstream on the Coxa side it should get planned for upgrad as tends to fall apart past 70% in a see of high pings and packet loss, if they don't do it they just move to sky (TV/Phone/broadband) as they do lose customers over this when you cant stream video due to packet loss
Posted by brianhe 16 days ago
The speeds quoted should also be the user data rate, not the line sync speed. This particularly affects BT adsl1.
Posted by andrew (Favicon staff member) 16 days ago
The speeds quoted are so vaguely worded in the what you are meant to say in an advert guidelines, that its never clear whether line sync, max IP throughput or user experience is meant.

Posted by Blackmamba 15 days ago
Hi Broadband Watchers.
As a lot of advertising is done on TV (BT Infinity) and have chosen a post Code that is under their speed target for Infinity. and found they state the correct band 13-15 on the address. The customer is showing an average of 13.3 TBB results over the last six months they are not with BT so it look like BT is advertising the correct information today. The results are taken off BT/ Openreach checker results.
Posted by andrew (Favicon staff member) 15 days ago
@blackmamba I think you've confused yourself and others again.

This ASA thing is NOT about the estimates at point of sale, but the figures featured in the actual TV advert.
Posted by pfvincent 15 days ago
I fully agree with Andrew that “requiring advertising to mention upload speeds in the body is something that is well overdue”, but feel that removing upload speeds to simplify your tables is a mistake. When I first got broadband from Demon many years ago upload was quoted as 10% of download by virtually all ISPs, and cloud computing and social media hadn’t been invented, so it wasn’t much of an issue. Many ISPs now fail to quote upload at all even though most people will be using cloud computing and/or social media unless all they do is stream TV!
Posted by pfvincent 15 days ago
I cannot find what my nominal upstream is since Virgin gave me a “free” upgrade to 70 Mbps (and then increased prices quoting their speeds as justification), but a test I’ve just run shows download as 110% of the nominal, but upload as only 7% of the nominal download. I think that this 5Mbps upload speed is virtually unchanged since the “free” upgrade. If I decide to change ISP, I should certainly want to know what upload speed I would be getting, as well as the download speed.
Posted by andrew (Favicon staff member) 15 days ago
Note including upload in this article was because of adding extra download info. Our monthly speed test roundup does include upload speeds, as do our package listings (unlike some providers own web pages).

On the VM 70 Mbps service, its complicated as they had an aim of 10% of download, but roll-out of that has not completed.

NOTE: There is no generic mathematical formula for download versus upload.
You must be logged in to post comments. Click here to login.