ASA Adjudication on Costa Ltd

Ad

A poster, which also appeared on a telephone box, and a press ad for Costa coffee.

a. The poster said "7 OUT OF 10 COFFEE LOVERS PREFER COSTA". Smaller print below said "In head-to-head taste tests, 7 out of 10 coffee lovers preferred Costa cappuccino to Starbucks". A cup of coffee carrying Costa branding was shown. Text below stated "WE MAKE IT BETTER". Small print at the bottom of the poster said “Source … In blind head-to-head taste tests between Costa’s cappuccino and a cappuccino from Starbucks or Caffe Nero 70% of respondents who identified themselves as ‘Coffee Lovers’ preferred Costa cappuccino. Total sample size of coffee lovers = 174. For more information please visit www.costa.co.uk”.

b. The press ad was headlined "STARBUCKS DRINKERS PREFER COSTA In head-to-head taste tests, 7 out of 10 coffee lovers preferred Costa cappuccino to Starbucks". A cup of coffee carrying Costa branding was shown. Smaller print below said "WE MAKE IT BETTER". Small print at the bottom of the page said "Source: Independent survey ... November and December 2008. In blind head-to-head taste tests between Costa's cappuccino and a cappuccino from Starbucks or Caffe Nero 70% of respondents who identified themselves as Coffee Lovers preferred Costa cappuccino. Total sample size of coffee lovers = 174. For more information please visit www.costa.co.uk".

Issue

Starbucks challenged whether:

1. the headline claim "7 OUT OF 10 COFFEE LOVERS PREFER COSTA" in ad (a) was misleading, because they believed it implied a preference for all Costa products over those of their competitors when it referred only to cappuccinos. They considered that the qualifying claim "... 7 out of 10 coffee lovers preferred Costa cappuccino to Starbucks" was insufficiently prominent and contradicted rather than qualified the headline claim;

2. the small print in ad (a) was sufficiently legible;

3. the headline claim "STARBUCKS DRINKERS PREFER COSTA" in ad (b) was misleading, because they believed it implied a preference for all Costa products over those of Starbucks' when it referred only to cappuccinos;

4. the headline claim "STARBUCKS DRINKERS PREFER COSTA" in ad (b) was misleading and could be substantiated, because they believed consumers would understand it to refer to coffee drinkers who cited Starbucks as their preferred outlet, which they understood related to a different sub-set of survey results than those on which the qualifying claim "... 7 out of 10 coffee lovers preferred Costa cappuccino to Starbucks" was based. They therefore did not believe the results of the survey would support the headline claim;

5. the ads, and in particular the headline claims, were misleading, because they believed the survey methodology on which the claims were based was flawed.

CAP Code (Edition 11)

Response

Costa said their agency consulted the CAP Copy Advice team in advance of publication, although some modifications were made before the final versions of the ads were published. They said the survey referred to in ads (a) and (b) was carried out by a leading blind-taste agency with over ten years experience of conducting taste test comparisons. Costa provided a copy of the survey and said they followed the advice of the CAP Copy Advice team by making a copy available on their website so that consumers could verify for themselves the comparisons made in the ads.

1. & 3. Costa disagreed with Starbucks interpretation of the complaint, that the claims "7 OUT OF 10 COFFEE LOVERS PREFER COSTA" and "STARBUCKS DRINKERS PREFER COSTA" implied a preference for all Costa products over those of their competitors. They believed the headline claims on their own were meaningless and were therefore incapable of substantiation. However, when the claims were read in conjunction with the qualifying text, they clearly referred to cappuccinos, rather than Costa products generally. Costa said the target audience of the poster and press ad, coffee lovers, would identify the image of the cappuccino cup and would therefore understand that the claims related to a preference for cappuccinos.

2. Costa said the small print was clearly visible and could be read at a distance of between four and five metres. Costa explained that the posters only featured on ground-level advertising spaces, such as at bus stops, and they considered that the small print was therefore visible for a normal-sighted person at a reasonable distance.

4. Costa reiterated that, as a stand alone claim, the headline "STARBUCKS DRINKERS PREFER COSTA" was meaningless and had to be read in conjunction with the body text. They said the survey showed 67% of respondents who identified themselves as Starbucks drinkers preferred Costas cappuccino. Costa argued that it was clear from the small print that the sample size of coffee lovers related to taste tests involving both Cafe Nero and Starbucks. They said the information given to consumers was accurate, and they did not believe the ad implied that all 174 respondents had identified themselves as Starbuck's customers. They pointed out that the ad directed consumers to their website, where full details of the tests were provided.

5. Costa pointed out that the methodology of the survey had been detailed in the explanatory notes on their website. They reiterated that both ads clearly referred to one product, cappuccinos, rather than each companys full product range.

The test survey consisted of 334 participants, 174 of whom identified themselves as coffee lovers. The results of the survey were statistically significant at the 95% confidence level. The tests were carried out in three locations in the UK (Glasgow, High Wycombe and Sheffield). Each location was chosen to ensure that the distance between the testing facility and each coffee retailer was as close as possible to ensure that cappuccinos arrived at the optimal serving temperature.

Participants were asked to blind test a total of two cappuccinos; they were not told which brand of products they would taste but were asked to identify their preferred coffee retailer, if any. The tests were all single blind; those conducting the tests knew which product was being tested but the participants did not. Costa pointed out that this information was given on their website. Each cappuccino container was covered with an unbranded sleeve and lids were removed to ensure that the identity of the brand was kept anonymous. They sent us a photograph showing how the sleeves covered both the Costa and Starbucks branding.

Costa explained that the order of tasting was rotated across the sessions. They supplied a summary sheet setting out the breakdown of the number of participants who tasted each product first. It showed that, of the 166 head-to-head taste tests carried out between Costa and Starbucks, 86 participants tried Costa first and 80 tried Starbucks first. They did not believe the difference of six participants had an effect on the results of the survey.

Costa said each participant was required to complete a written survey to ensure that their responses were not influenced by the individuals conducting the testing and provided us with a copy. Participants were asked to answer the same questions in the same order for each taste test and were only asked questions about the comparative quality of the two cappuccinos after the taste test was concluded. Costa said the results were statistically significant at the 95% confidence level.

Costa reiterated that the CAP Copy Advice team had reviewed the survey and was satisfied that it substantiated the claims in the ads.

Assessment

The ASA understood that the CAP Copy Advice service had offered guidance to Costa on versions of the ads that featured different headline and body copy claims to ads (a) and (b).

1. Not upheld

We noted Starbucks' concern that the headline claim, "7 OUT OF 10 COFFEE LOVERS PREFER COSTA", implied a preference for all Costa products over those of their competitors. Whilst we acknowledged that it was unclear from the headline claim which competitor the comparison was being made with, we considered that the specific reference to coffee and the image of the coffee cup would be understood by consumers to imply a particular preference for Costa's coffee products only, rather than all of their products.

We understood that the CAP Copy Advice team had advised that it would be acceptable for the headline claim not to refer to the fact that the preference related to cappuccinos specifically, as long as that was made clear in qualifying text in the body of the ads. We noted that the poster did contain qualifying text, as advised, that stated "In head-to-head taste tests, 7 out of 10 coffee lovers preferred Costa cappuccino to Starbucks". We considered that that qualifier made clear the specific competitor and coffee product on which the headline claim was based, and that it was sufficiently prominent to capture consumers' attention. We therefore concluded that on this point ad (a) was not misleading.

On this point, we investigated ad (a) under CAP Code clauses 7.1 (Truthfulness) and 18.1 (Comparisons with identified competitors and/or their products) but did not find it in breach.

2. Not upheld

We noted that the small print was clear and of a suitable size for it to be read by a normal-sighted person from a reasonable distance. Notwithstanding our concern, discussed in point (5) below, that the information in the small print was inaccurate, we concluded that the small print was sufficiently legible.

On this point we investigated ad (a) under CAP Code clauses 7.1 and 7.2 (Truthfulness) but did not find it in breach.

3. Not upheld

We noted that the headline claim "STARBUCKS DRINKERS PREFER COSTA" was immediately followed by qualifying text that stated "In head-to-head taste tests, 7 out of 10 coffee lovers preferred Costa cappuccino to Starbucks". We considered that the qualifying text was sufficiently prominent to capture consumers' attention, and that it was therefore clear that the headline claim referred to a preference for Costa cappuccino only, rather than for all Costa products. We concluded that on this point ad (b) was not misleading.

On this point we investigated ad (b) under CAP Code clauses 7.1 (Truthfulness) and 18.1 (Comparisons with identified competitors and/or their products) but did not find it in breach.

4. Not upheld

The ASA took expert advice.

We noted Starbucks' concern that the headline and qualifying claims were based on two different sets of results from the survey; those relating to participants who identified themselves as Starbucks drinkers and those relating to participants who identified themselves as 'coffee lovers'. We understood from the survey results that the Costa-Starbucks taste tests were based on a sample size of 166 people, and that of the 57 participants who identified themselves as Starbucks drinkers, 67% said they preferred Costa's cappuccino. We therefore considered that the survey results supported the headline claim that "Starbucks drinkers prefer Costa". We noted that our expert had found that the results of the survey were statistically significant and that the sample size was adequate.

We understood that the qualifying claim, "7 out of 10 coffee lovers prefer Costa cappuccino to Starbucks", was based on a different set of results from the survey; of the 81 participants who identified themselves as coffee lovers in the taste tests, 72% preferred Costas cappuccino. However, we considered that, in the absence of any explanation to the contrary, the impression created by the qualifying claim was that 7 out of 10 Starbucks drinkers preferred Costas cappuccino, when that was not the case. We considered that the relationship between the headline and qualifying claims in the ad was unclear, and we were therefore concerned that the presentation of the two claims could be confusing for consumers.

We noted that the headline claim was linked to small print, which explained that the survey was based on a sample size of 174 coffee lovers, 70% of whom had preferred Costa's cappuccino. We understood, however, that the small print referred to combined sample sizes and results for Costa-Starbucks and Costa-Caffe Nero taste tests. We were concerned that the smallprint could give a misleading impression about the strength of results on which the main preference claim for the Costa-Starbucks taste tests were based, and we therefore considered that the correct sample sizes of 57 and 81 participants should have been stated in the small print, in order to make clear for consumers the basis on which the headline and qualifying claims were made.

Nevertheless, because the results of the survey did support the claim that "Starbucks drinkers prefer Costa" we concluded that the headline claim was not misleading.

On this point, we investigated ad (b) under CAP Code clauses 3.1 (Substantiation), 7.1 (Truthfulness) and 18.1 (Comparisons with identified competitors) but did not find it in breach.

5. Not upheld

We understood that Starbucks believed the survey methodology was flawed for a number of reasons: it was based only on cappuccinos which did not accurately reflect the product ranges of the companies involved; it was not clear whether the tests were double blind to prevent tasters from identifying the tested brand; it was not clear if the tasters knew in advance which products they would be tasting; it was unclear whether equal numbers of tasters tasted the same brands first; it was not clear whether the questions asked of tasters would have influenced their answers; the sample size was not sufficiently large to give a reliable result; and whether the chosen location influenced results.

We noted our expert's view that the survey methodology did not contain design errors, that the results were statistically significant and the sample size was adequate. We also noted that our expert had concerns that there were weaknesses in the amount of reported detail on the methods used, and that consequently some of the information provided about the methodology had to be taken on trust.

We noted that the ads stated that the claims related to a preference for cappuccinos, and we did not consider that the survey was flawed on the basis of only selecting one product for comparison. We noted the products were anonymised prior to testing and participants were not told the brands of the products they would be testing. We understood that double blinding would have eliminated the possibility of bias in the results, but we considered that, while it would not totally eliminate the possibility of bias, the fact that participants had to complete a written survey would reduce the possibility that respondents were influenced by the testers. We therefore considered that single blind tests were acceptable. We considered that consumers would understand taste to be a subjective measure, and that the survey methodology as reported was suitable to support a claim based on the subjective impressions of the participants.

We noted that the small print in ads (a) and (b) referred to combined results and sample sizes for the Costa-Starbucks and Costa-Caffe Nero taste tests, and for the reasons described in point (4) above we considered that the correct sample sizes of 81 'coffee lovers' and 57 'Starbucks drinkers' on which the claims in the ads were based should have been stated in the small print. However, because we considered that the survey methodology was suitable to support the claims made in ads (a) and (b), we concluded that on this point the ads were not misleading.

On this point, we investigated ads (a) and (b) under CAP Code clauses 3.1 (Substantiation), 7.1 (Truthfulness), 18.1 and 18.3 (Comparisons with identified competitors and/or their products) but did not find them in breach.

Action

No further action necessary.

Adjudication of the ASA Council (Non-broadcast)