ASA Adjudication on British Telecommunications plc
British Telecommunications plc
81 Newgate Street
25 August 2010
Television, National press, Regional press, Radio
Computers and telecommunications
Number of complaints:
A TV ad, three radio ads and a national press ad, for BT Total Broadband.
a. The TV ad showed an estate agent showing a man round a property while the man discussed the property with his partner who was at home on her PC. The estate agent attempted to access a webpage, but had to wait for the page to load, while the man’s partner loaded the website instantly and then visited different pages on the website very quickly. The estate agent apologised to the man saying “6 o’clock. Half the world’s online”. A voice-over then stated “BT is rolling out up to 20 meg speeds to give you a consistently faster broadband throughout the day even at peak times”. On-screen text stated “… Comparison to BT’s up to 8Mb service …”.
b., c., & d. The three radio ads all included a voice-over that stated “Luckily BT is rolling out up to 20 meg speeds to give you a consistently faster broadband even at peak times. Switch to BT Total Broadband. 20 meg download speeds compared to 8 meg broadband; speeds vary; BT line required; subject to 12 month term, area and availability”.
e. The national press ad was headlined “CONSISTENTLY FASTER BROADBAND even at peak times”. Text stated “When there are loads of people online the internet can grind to a halt, frustratingly this is often when you need it most. Luckily BT is rolling out broadband speeds of up to 20Mb, to deliver consistently faster broadband even at peak times compared with the ADSL* industry average (dependent on line and location**). Small print stated “**Speeds quoted are download speeds. Factors affecting speed include local availability, distance from the exchange and internal home wiring. We’ll offer the best speed available on your line. Please visit www.bt.com/superfastbroadband to see what speed we can provide on your line. *Industry average = Epitiro data comparing 11 major ADSL providers and Isposure data providing the average speed of 1 other major provider. Data compiled 25th to 31st July 2009 …”.
The ASA received 17 complaints.
1. Eleven viewers and one listener challenged whether the claim "BT is rolling out up to 20 meg speeds to give you a consistently faster broadband even at peak times" in ads (a), (b), (c) and (d), which compared the new service with BTs existing 8 Mbit/s service, was misleading and could be substantiated.
2. Four viewers objected that the visual comparison of speed in ad (a) was misleading, because they believed surfing a typical website would not be any faster with a 20 Mbit/s service than it would with an 8 Mbit/s service.
3. Three viewers objected that ad (a) was misleading, because the speed of the visual demonstration on the woman's laptop was, they believed, faster than that which could be achieved ordinarily at any connection speed.
4. British Sky Broadcasting (Sky) challenged whether the claim "BT is rolling out up to 20 meg speeds to give you a consistently faster broadband even at peak times" was misleading in the context of the industry comparison claim in ad (e) because BT operated a traffic management policy.
Sky and TalkTalk challenged whether ad (e) was misleading because:
5. it did not make clear on-going restrictions on the availability of the 20 Mbit/s service; and
6. they understood the figures used as the basis of the claim "consistently faster broadband speeds compared with the ADSL* industry average" were not independently obtained and were not suitable for an industry wide comparison because they could have been affected by the particular telephone lines selected.
7. Virgin and TalkTalk challenged whether the claim "consistently faster broadband speeds compared with the ADSL* industry average" in ad (e) was misleading, because they believed it was contradicted by a recent independent survey conducted by Ofcom, the results of which were more robust as they were complied over a much longer period of time.
CAP Code (Edition 11)
BCAP TV Code
BCAP Radio Code
1. British Telecommunications plc (BT) said the claim was used in a self-referential context in ads (a), (b), (c) and (d). They pointed out that, at the time the ads appeared, they had begun rolling out their new up to 20 Mbit/s service across the country. BT said they wanted to highlight that newly enabled telephone exchanges were able to provide higher speeds in those areas, which naturally resulted in customers web browsing speeds being consistently faster than on their previous service. They said the ads focused on web browsing to reflect the service's improved HTTP download speeds; they did not expressly claim that BT Total Broadband would be quicker for all online activities.
BT assured the ASA that the speeds achieved using their new 20 Mbit/s service were quicker than those achieved using the 8 Mbit/s service. They provided graphs that were produced using Isposure data and measured the download speeds for both services during March and April 2010. They said the data related to Transfer Control Protocol (TCP) traffic rather than HTTP traffic, which, they maintained, provided a more accurate view of raw line speed achieved by users. The graphs showed the speed was always higher for the 20 Mbit/s service than was the case on BT's 8 Mbit/s service. BT therefore maintained that they had substantiated the claim.
Clearcast said they had been assured that BT were rolling out 20 Mbit/s speeds and had received evidence to show that the speeds had been tested over 24-hour periods. On that basis, they were satisfied that the service was as advertised. They provided the graphs they had received from BT in support of the claim, which related to HTTP traffic speeds for the period 5 to 11 June 2009.
The RACC sent copies of the data BT had provided to them, which consisted of the same graph that had been submitted to Clearcast and a document that explained the background to that data. It stated that the comparison was based on one BT line in an area of London, for each of their two services, and on 12 competitors lines in various different areas.
2. & 3. BT said the depiction in ad (a) was not intended as an actual comparison of the faster speeds achievable on their new 20 Mbit/s service. They said it was merely a means of visualising a typical frustrating moment that viewers could relate to. BT said that, nevertheless, their new service was naturally faster than the existing one. They said the demonstration was more generic, between an area with a consistently fast connection and another area without. They believed that it was reasonable to feature a shortened sequence than would actually occur when using the service due to the nature of TV as an advertising medium. However, BT maintained that, irrespective of their intention in creating the ad, the typical HTTP website loading times shown were not unrealistic.
Clearcast said they were assured by BT that the representation of speed in the ad was accurate and took into account the evidence they had received, which compared the speeds of either service. They also said there had been some confusion with BT's response on Points 2 and 3. BT had stated initially that the visual demonstration was not intended to be representative of actual internet speeds; however, they had since confirmed that the refresh speed shown in the ad was what a customer could expect to achieve with a 20 Mbit/s speed.
4. BT said Sky's assertion was incorrect, but acknowledged that that was due in part to the fact that they had not amended part of their terms and conditions of service to reflect changes to their service by the time the ads appeared. BT nevertheless maintained that such a slight discrepancy did not result in viewers being materially misled. They pointed out that they did not apply the video streaming traffic management cap of 896 Kb to BT Total Broadband Option 1 customers who were on their new ADSL2+ network, which was the service promoted in all of the ads. They said the terms and conditions would be amended to make that clear. BT pointed out that the Epitiro UK Ltd (Epitiro) report used as the basis of the claim demonstrated that at peak times, for HTTP traffic, it was consistently faster than the industry average. They said the policy was applied to heavy users on the new network but was only used as a last resort and such users accounted for less than 1 per cent of their customer base. They added that speed restrictions for persistent heavy users were in place to prevent other normal users experience suffering as a result. BT pointed out that customers using the service in line with the terms and conditions referenced in the ad would not be subject to such measures.
5. BT said their up to 20 Mbit/s service was currently available to around 45 per cent of UK households rising to 55 per cent by March 2010 and 75 per cent in 2011. They maintained that the term "rolling out" indicated clearly to consumers that BT was in the process of providing new higher speeds by updating its exchanges. They also maintained that the disclaimer ... dependent on line and location**" prominently qualified the claim, making it clear that access to the 20 Mbit/s service was not guaranteed for all customers. BT said the small print explained some of the factors that might affect the availability of broadband speeds and, in accordance with Ofcom's Broadband Speed Code of Practice, the footnote also included a link to BT's speed checker, which allowed consumers to see what speed they were likely to be able to achieve. They pointed out that further checks were also made before a customer placed an order but said they were willing to amend their future marketing material to state "enabled areas only" or "selected areas only".
6. & 7. BT said the data used to support the industry comparison claim in ad (e) was independently compiled by Epitiro. They submitted a copy of Epitiros report, which they pointed out stated ... neither BT, nor any of the ISPs, influenced the locations Epitiro selected/selects for monitoring of ISP services". BT said even if they had wished to influence the results, it would not have been possible. They said the report assessed BT and all other major ISPs HTTP download speeds between July and August 2009, with the aim of testing whether BT's 20 Mbit/s service was consistently faster than the ADSL industry average. For Sky, the only major ISP that could not be measured using Epitiros ISP-I measurement as they did not subscribe to it, they used a second stream of data from Epitiros Isposure, which was a database of broadband key performance indicators (KPIs). They maintained that, as with the ISP-I data, Isposure data could not be manipulated or interfered with by ISPs. BT said the technology was downloaded by consumers to their own computer via the Isposure website and stated that that ensured that the data gathered was sufficiently random.
BT accepted that the Epitiro study did not normalise for line length, which would ensure that the data did not feature any inaccuracies derived from the significant variations in speeds achieved by lines of different distances from the telephone exchange. However, they maintained that, because a specific exchange served the test locations, it was not necessary as all of the lines used were of the same length. They said the industry average was made up of all ISPs services at the Epitiro sites; the line length would not favour one ISP over another because they all served the same location.
BT said that Epitiros methodology was designed to ensure the independence of the data collected and to thereby protect any results from being prone to manipulation by ISPs. For instance, they pointed out that individual ISPs could not manipulate the particular telephone lines selected for the tests. BT pointed to the relevant section of the report, which stated "ISP-I performs an independent mystery shopper analysis by testing ISP services as delivered to Epitiro lines we hire as consumers. Epitiro arbitrarily and secretly hires monitoring locations from around the UK, then subscribes to a number of ISP services at these sites anonymously as a typical consumer (i.e. Sam Brown from 123 East Street). From these locations Epitiro hardware agents connect to the internet services, as a typical consumer, and perform a number of typical tasks such as web surfing, emailing and downloading content".
BT said the data for the Ofcom report referred to by Virgin and Talk Talk was gathered between November 2008 and April 2009; it was published in July 2009, two months after its completion. As far as BTs network was concerned, the report was already out-of-date when it was published because their network had changed considerably. The Ofcom report did not include data on their ADSL2+ network of up to 20 Mbit/s speeds. The performance of that network formed the basis of Epitiros more recent report and the comparative claim in the ad. They said the Ofcom report clearly set out its limitations from the outset, stating ... the broadband market continues to evolve rapidly and the speeds and general performance results set out in this report are therefore liable to change".
BT said the data used to support the claim was initially gathered between 25 and 31 July 2009. However, they pointed out that they had been in the process of gathering more data and knew from research that the claim was not going to be contradicted by further tests. The dates were included in the small print of the ad to distinguish the data from the Ofcom report. They accepted that one weeks worth of data might be disproportionate to the claim in the ad; however, the full Epitiro report was extensive and thorough and substantiated the claim on an ongoing basis; the ISP-I data was gathered from the end of July until late September and demonstrated BT could substantiate the claim, as did the weekly Isposure data gathered from 25 July to 29 August.
They said the Ofcom study considered eight major ISPs, plus a number of smaller ISPs; that represented over 90 per cent or subscribers in total whereas Epitiros data represented 98 per cent of subscribers. Ofcoms test duration ranged from one to six months and Epitiros ranged from five weeks to five months. Ofcoms report was based on 1,600 test agents, averaging 200 agents per major ISP, in Epitiros study, the BT 20 Mbit/s service alone had an average in excess of over 500 test agents. BT said it was not possible to test every line; neither study did and any study could be affected by issues such as the consumer's distance from the exchange. However, the claim was not misleading and could be substantiated by the Epitiro study.
BT said the data used by Epitiro was derived from a number of sources, including Isposure and ISP-I. They said those technologies were developed and owned by Epitiro for the purposes of gathering data on performance metrics, which was then provided to ISPs for various internal and marketing purposes. BT pointed out that 12 major ISPs were included in the industry average figures used in ad (e), along with over 50 smaller ISPs, and pointed out that the industry average was comprised of speed tests from over 11,400 agents, including 600 to calculate BT's average HTTP traffic speed. They said all of the tests in the report, for both the ISP-I and Isposure data, were carried out using HTTP traffic from normal websites. The websites included, for example www.bbc.com and www.myspace.com, and the results illustrated that there was an improvement in everyday internet browsing with BT's up to 20 Mbit/s service. They said over 60 per cent of BT customers' internet use was HTTP based traffic; they believed other ISPs would also consider HTTP to be their predominant traffic and, on that basis, research was carried out to establish whether BT was consistently faster than the industry ADSL average. BT said they did not believe consumers would be misled by that and pointed out that the ad did not suggest that BT Total Broadband would be quicker, for example, for downloading certain types of files.
Epitiro said all of their clients were contractually bound to request permission prior to making reference to the companys data. They said BT had made such a request in August 2009 and Epitiro had checked their databases were able to substantiate the claim related to the timeframes specified in the ad. They said the report BT commissioned from Epitiro in October 2009 was not the first time BT had seen the evidence; it was complied in response to the complaints the ASA had received and BT requested that Epitiro consolidate all data obtained so far into one report to allow the ASA to consider the substantiation for the claim. They confirmed that BT had the relevant data to support the claim before the ad was published and were in regular contact with Epitiro regarding collating data to support the claim on an ongoing basis.
The ASA noted the broadcast ads, (a), (b), (c) and (d), compared the performance of BT's new 20 Mbit/s service and their existing 8 Mbit/s service. We considered the claim ... consistently faster broadband even at peak times" when compared to BT's existing 8 Mbit/s service was categorical and therefore likely to lead consumers to expect that BT could demonstrate that that was the case in all instances.
We noted BT had initially based their claim on HTTP traffic speed data gathered before the ads appeared; we understood that data related to the accessing and downloading of information, most commonly found on websites. We understood, however, that HTTP traffic did not account for other common internet traffic, such as streaming media and file transfer protocol. We noted the ads referred to "broadband" and did not make clear that the claim related to web browsing only. Consequently, we considered that many viewers and listeners were unlikely to be aware of the limited scope of the claim and would expect that BT could demonstrate that, in general, their broadband service was faster at peak times.
We also noted the information provided by BT in relation to TCP traffic speeds, which BT said provided a more accurate view of users speeds than HTTP data. We noted, however, the data was gathered on dates after the ads were originally broadcast. Because it was not held at the time the ads were broadcast, we did not consider it further.
Because we had not seen sufficient evidence to support the claim that BTs new broadband service was consistently faster than its existing 8 Mbit/s service even at peak times, we concluded that the ad was likely to mislead.
On this point, ads (a), (b), (c) and (d) breached CAP (Broadcast) TV Advertising Standards Code rules 5.1.1 (Misleading advertising) and 5.2.1 (Evidence) and CAP (Broadcast) Radio Advertising Standards Code section 2 rule 3.1 (Misleadingness).
2. & 3. Upheld
We noted ad (a) featured a visual comparison of the mans partner accessing a webpage very quickly when using BT's new 20 Mbit/s service, when, at the same time, the estate agent was shown to have difficulties in performing the same function when using BT's existing service. We noted the complainants concern that, irrespective of the difference in performance between the two services, a 20 Mbit/s connection would not be able to access a typical website any faster than an 8 Mbit/s service. We understood speed of access to websites related to other factors, such as the speed of the connecting server and other traffic accessing a website along with the speed of the users connection. Although we noted a 20 Mbit/s connection was likely to have a greater potential speed in general, we understood that, in most cases, there would not be an appreciable difference between the downloading times of both services for a typical website.
We also noted the mans partner accessed the website and navigated to several different pages instantaneously. We understood however that a typical website, even loaded with a very high speed connection, would still require a small amount of time to load and appear on the users screen.
Although we noted BT's argument that the sequence was not intended as an actual visual comparison we considered that, in an ad focused on the benefits of a faster service, consumers would expect any comparative demonstration of that feature to be representative of the benefits available to them. Because we understood that that was not the case, we concluded that the ad was likely to mislead.
On points 2 and 3, ad (a) breached CAP (Broadcast) TV Advertising Standards Code rules 5.1.1 (Misleading advertising), 5.2.1 (Evidence) and 5.4.6 (Comparative Advertising).
We noted BT accepted that some users speeds were restricted at peak times in instances of excessively heavy use. We considered, however, that the use of the claim ... consistently faster broadband even at peak times" compared to the industry average was categorical and, therefore, likely to lead consumers to expect that BT could demonstrate that that was the case in all instances. Because those users affected by the traffic management policy were unlikely to achieve speeds faster than the industry average cited by BT, we concluded that the ad was likely to mislead.
On this point, ad (e) breached CAP Code clauses 3.1 (Substantiation) and 7.1 (Truthfulness).
The ASA acknowledge BT were willing to amend their future marketing material. We noted the ad included the text "rolling out ...", which, we considered, was likely to be interpreted to mean that the new 20 Mbit/s service was not currently available to all households but soon would be. We noted small print ... dependent on line and location**" but considered it was likely to be interpreted as a reference to the broadband speeds that could be achieved on each connection, rather than the availability of the service.
We noted BT's new service was available to fewer than half of all households and the roll-out increasing that figure to 75 per cent was anticipated to take around two years. We noted that a significant proportion of the population could not get the service at the time the ad appeared and therefore considered the ad should have made that clear. Because it did not, we concluded that it was likely to mislead.
On this point, the ad (e) breached CAP Code clauses 7.1 and 7.2 (Truthfulness).
We noted Sky and Talk Talks objection that the industry performance data was not independently obtained and was not suitable for an industry wide comparison but noted it had been gathered by Epitiro, an independent provider of industry data to ISPs. While it had in fact been independently obtained, we were concerned that the data was potentially subject to inaccuracies based on line length. We noted Epitiro had not normalised for line length as part of their methodology. We understood that speeds of ADSL services were determined primarily by the distance between the user and the telephone exchange. We also understood that the speeds achieved by users varied significantly as a result of this. We therefore considered that even a sizeable and random sample of lines might not be sufficient control for the effect of line length.
Because we had not seen sufficient evidence to support the claim that BT's new broadband service was consistently faster than the ADSL industry average even at peak times, we concluded that the ad was likely to mislead.
On this point, ad (e) breached CAP Code clauses 3.1 (Substantiation), 7.1 and 19.1 (Other comparisons).
7. Not upheld
We noted Virgin and TalkTalks objection that BT's claim was contradicted by Ofcom's research. However, Ofcom's research did not specifically take into account BT's 20 Mbit/s service; we therefore considered it was not relevant to the claim in the ad, because it was not directly comparable. We concluded that the ad was unlikely to mislead on this point, notwithstanding the reasons for upholding in Point 6 (above).
On this point, we investigated ad (e) under CAP Code clauses 3.1 (Substantiation), 7.1 and 19.1 (Other comparisons) but did not find it in breach.
The ads must not appear again in their current form. We reminded BT to ensure they held robust documentary evidence to prove all claims capable of objective substantiation. On Point 5, we also told them to ensure significant restrictions were made clear in future.
Adjudication of the ASA Council (Broadcast)
Adjudication of the ASA Council (Non-broadcast)