Interflora Inc & Anor v Marks and Spencer Plc & Anor

The Claimants (“Interflora”) seek the permission of the Court to adduce in evidence the witness statements of 13 witnesses at the trial of this action scheduled for mid April 2013. Interflora do not accept that they need the Court’s permission, but they have sought it out of an abundance of caution. Having regard to the previous history, that was a sensible approach to adopt. As Interflora rightly anticipated, the First Defendant (“Marks & Spencer”) contends that permission is required and should be refused.

The general background to the matter is set out in a number of previous judgments of mine in this case and in the recent judgment of the Court of Appeal [2012] EWCA Civ 1501, [2013] ETMR 11 at [3]-[12]. I shall not repeat it again. For present purposes it suffices to say that Interflora allege that Marks & Spencer has infringed their United Kingdom and Community trade marks for the word INTERFLORA by ensuring that advertisements for Marks & Spencer’s flowers are displayed when consumers search for “interflora” (and similar terms) using Google.

Interflora allege infringement pursuant to Article 5(1)(a) and 5(2) of what is now European Parliament and Council Directive 2008/95/EC of 22 October 2008 to approximate the laws of the Member States relating to trade marks (codified version) (“the Directive”) and Article 9(1)(a) and (c) of what is now Council Regulation 207/2009/EC of 26 February 2009 on the Community trade mark (codified version) (“the Regulation”). Interflora’s primary case under Article 5(1)(a) of the Directive and Article 9(1)(a) of the Regulation is that Marks & Spencer’s advertisements do not enable reasonably well-informed and reasonably observant internet users, or enable them only with difficulty, to ascertain whether the goods and services referred to in the advertisement originate from Interflora, or an undertaking economically connected to it, or on the contrary emanate from a third party. In particular, Interflora contend that some users wrongly think that there is a trade connection between Marks & Spencer and Interflora, for example that Marks & Spencer is part of the Interflora network.

The Court of Appeal’s judgement concerned an application by Interflora for permission to adduce the evidence of witnesses who had been identified by means of two exercises which Interflora described as “pilot surveys”. Interflora did not intend to rely on the “pilot surveys” themselves, which Interflora accepted were not statistically reliable. Nor did Interflora intend to repeat those surveys on a larger scale.

Lewison LJ described the first exercise in his judgment at [15] as follows:

“… Respondents were recruited on the street and were first asked a number of screening questions. Three of these questions were whether the potential respondents had purchased or looked into the purchase of flowers in the previous two to three years; whether they had undertaken activities on the internet but were not confident about creating a web page; and whether they were in a position to read a computer screen. Some 95 respondents passed the screening test, and were chosen to answer further questions. They were brought into a hall set up with four laptop computers. Each participant was asked to type the word ‘Interflora’ into the search box on the Google homepage. It was not a real Google search, because each computer was set up so as to display one of the search results pages which are the subject of this litigation. The search results page brought up Interflora, both as a natural or organic result and also as a sponsored link. The search results page also brought up an M & S sponsored link, quoted above, together with other sponsored links. Participants were then invited to scroll up and down the page and to tell the interviewer when they were ready to answer questions. Once ready each participant was asked a series of structured questions. These questions were changed during the course of the survey. The questions did not distinguish between natural or organic results on the one hand and sponsored links on the other. The questions were framed generically by reference to ‘results’ or ‘search results’. … In the first survey [question 6] was posed as follows:

‘Thinking specifically about this search result (POINT to MARKS & SPENCER RESULT). What if anything do you think this tells you about any relationship between Interflora and Marks and Spencer?’”

The second exercise was carried out using broadly the same methodology, but question 6 was different.

Interflora wished to adduce evidence from selected participants in the two exercises: in essence, those who had given answers favourable to Interflora’s case and were agreeable to being called as witnesses. Interflora accepted that they needed permission to adduce this evidence as a result of an order I had made on 24 February 2012. Marks & Spencer resisted the grant of permission. I gave Interflora permission to call witnesses identified as a result of the first exercise, but not as a result of the second exercise: [2012] EWHC 1722 (Ch), [2012] FSR 32. I gave Marks & Spencer permission to appeal on the basis that the matter raised an important question of principle as to the correct approach to such applications. Interflora was granted permission to cross-appeal by Lloyd LJ. Marks & Spencer’s appeal was allowed and Interflora’s cross-appeal was dismissed. Shortly before the hearing of the present application the Supreme Court refused Interflora permission to appeal against the Court of Appeal’s decision.

In his judgment, with which Etherton LJ (as he then was) and Hughes LJ agreed, Lewison LJ considered two objections to Interflora’s application raised by Marks & Spencer which he described as the macro objection and the micro objection.

Lewison LJ summarised the macro objection at [24] as follows:

“Mr Hobbs' macro objection is that the court should not permit evidence to be adduced from witnesses selected as a result of a survey unless the survey itself is statistically reliable. It does not matter whether such evidence is technically admissible, because the court has power under CPR 32.1 (2) to exclude evidence that is otherwise admissible. The argument runs thus:

i)

The question at issue is whether the reasonably well-informed and reasonably observant internet user would or would not understand the M & S advertisement to indicate that M & S was part of the Interflora network.

ii)

The reasonably well-informed and reasonably observant internet user is not a real person. He or she is a legal construct. As the Court of Justice made clear, the fact that some internet users may have had difficulty grasping that the service provided by M & S is independent from that of Interflora is not a sufficient basis for a finding that the function of indicating origin has been adversely affected.

iii)

Accordingly, simply to call some internet users to give evidence is not probative of the issue in the case.

iv)

That evidence can only be probative if those who are called can in some way be seen to stand proxy for the legal construct through whose eyes the essential question must be judged.

v)

Unless the survey used for witness collection is itself a reliable survey, the court will have no means of knowing whether the selected witnesses can be treated as reliable proxies for this legal construct. The problem is compounded where, as here, the party calling the witnesses is permitted to select those who give most support to its case.

vi)

Even if the evidence of such witnesses is, in principle, admissible, it is likely to be of such marginal utility and so expensive and time consuming to collect, analyse and deal with in court, that the court ought to exclude it in exercise of its powers under CPR Part 1.4 (2) (h) and CPR Part 32.1.”

Lewison LJ accepted this argument for the reasons he gave at [26]-[146]. His reasoning may be summarised as follows.

First, he started from the proposition that the “average consumer” and the “reasonably well-informed and reasonably observant internet user” (who are the same in the circumstances of this case) were legal constructs like the person skilled in the art (patents) and the informed user (designs): see in particular [13], [37]-[44], [73]. In this connection he made a distinction between trade mark cases and passing off cases: see in particular [33]-[34].

Secondly, he emphasised that the question to be decided was a qualitative assessment: see in particular [34]-[36].

Thirdly, he held that the results thrown up by search engines on the internet fall within the general description of ordinary consumer services in relation to which the judge could make up his or her own mind without the need either for expert evidence or the evidence of consumers: see in particular [46]-[50] and [56]-[59].

Fourthly, he accepted that properly conducted and statistically reliable consumer surveys may be of evidential assistance in some trade mark cases, particularly where the issue is distinctiveness of the mark: see [51]-[55].

Fifthly, having reviewed both the guidelines for survey evidence laid down by Whitford J in Imperial Group plc v Philip Morris Ltd [1984] RPC 293 and the recent practice adopted by the judges of this Division with regard to the admission of survey evidence and “witness collection programmes” at [60]-[72], he posed this question:

“I stress again that what is in issue is the effect on a hypothetical legal construct: the reasonably well-informed and reasonably observant internet user. In the absence of special circumstances, how can calling ten, twenty or thirty witnesses selected from a statistically invalid survey be extrapolated into the effect on that legal construct of the advertisement or sign in issue?”

After citing a passage from the judgment of the Federal Court of Australia in Arnotts Ltd v Trade Practices Commission (1990) 97 ALR 555, he answered this question at [76] as follows:

“The proposal in the present case is not quite as dire as the second possibility discussed in Arnotts because Interflora have disclosed all the answers to the questionnaires. But the remaining objections [mentioned by the Federal Court] still hold good: viz. that the selected witnesses are not (or at least cannot be shown to be) a fair sample of the class of reasonably well-informed and reasonably observant internet users, with the consequence that there is no ground for any extrapolation on a statistical basis, or on the basis of any mathematical or logical probability, of the views of the selected witnesses as representing the effect of the M & S advertisement on the hypothetical reasonably well-informed and reasonably observant internet user. If evidence of this kind cannot form the basis for extrapolation on the basis of any mathematical or logical probability leading to a conclusion about the effect of M & S's advertisement on the hypothetical reasonably well-informed and reasonably observant internet user, then in the absence of special circumstances it cannot be useful. And if it cannot be useful, it should not be allowed to distract the focus of the trial even if it is technically admissible.”

Sixthly, he reviewed a considerable number of previous trade mark and passing off cases at [77]-[134] and concluded at [135] as follows:

“The upshot of this review is that courts have allowed the calling of evidence of the kind that Interflora wishes to call and have considered it, either in conjunction with or in the absence of a statistically valid and reliable survey. But it is generally of little or no value. Sometimes it does no more than confirm the conclusion that the judge would have reached without the evidence. In passing off cases it sometimes has greater effect, but as I have said more than once, passing off raises a different legal question. Unless the court can be confident that the evidence of the selected witnesses can stand proxy for the persons or construct through whose perception the legal question is to be answered it simply represents the evidence of those individuals. In a case in which the witnesses are called in order to amplify the results of a statistically reliable survey their evidence may be probative. But unless the court can extrapolate from their evidence, it is not probative.”

Seventhly, he distinguished the situation with which he was concerned, namely the calling of selected witness obtained by means of a statistically unreliable survey, from other possibilities. Since this passage is important to the arguments in the present case, I shall quote it in full with the particularly relevant points emphasised:

“136.

Mr Silverleaf argued that if we acceded to Mr Hobbs' submissions then evidence from consumers would never be admitted in a case of trade mark infringement in the absence of a statistically valid and reliable survey. I do not think that follows. One of the objections to the witness collection exercise, as Rimer J pointed out in UK Channel Management is that the evidence thus collected is not the spontaneous reaction of members of the public who have been exposed to the allegedly infringing sign or advertisement, but is evidence obtained under artificial conditions by applying artificial stimuli. If there is evidence of consumers who have been confused in the real world, there can be no objection to calling it. He also submitted that the calling of evidence from witnesses identified by means of a witness collection exercise was quite independent of the process by which they had been identified. That may be so in some cases; but in the general run of cases where witnesses have been identified by a tailored series of questions, they will have been led towards a particular mindset which no longer represents the unstimulated evidence of people in the real world.

137.

That is not to say that there can never be evidence called in a case of trade mark infringement. The court may need to be informed of shopping habits; of the market in which certain goods or services are supplied; the means by which goods or services are marketed and so on. In addition I must make it clear, however, that different considerations may come into play where:

i)

Evidence is called consisting of the spontaneous reactions of members of the relevant public to the allegedly infringing sign or advertisement;

ii)

Evidence from consumers is called in order to amplify the results of a reliable survey;

iii)

The goods or services in question are not goods or services supplied to ordinary consumers and are unlikely to be within the judge's experience;

iv)

The issue is whether a registered mark has acquired distinctiveness; or

v)

Where the cause of action is in passing off, which requires a different legal question to be answered.

138.

Outside these kinds of cases there may be others where a judge might think that it would be useful to hear from consumers. I would not wish to rule out the possibility. So I would not accept the proposition that evidence from respondents to a questionnaire can never be called in the absence of a statistically valid and reliable survey. But (apart from those I have mentioned) the cases in which that kind of evidence might be of real use are difficult to imagine. I would not therefore hold that such evidence is inadmissible as a matter of law.”

Eighthly, he held at [139]-[145] that the combined effect of CPR rr. 32.1, 1.4(1) and 1.4(2)(h) required the court to consider whether the likely benefit of the evidence which was sought to be admitted justified the cost of adducing it. Accordingly, he stated that the correct approach was as follows:

“I consider that, even if the evidence is technically admissible, the judge should not let it in unless (a) satisfied that it would be valuable and (b) that the likely utility of the evidence justifies the costs involved.”

Ninthly, he held at [146] that Interflora’s application failed the first part of this test:

“In the present case I do not consider that Interflora has demonstrated that the evidence it wishes to call would be of real value. To put it bluntly, Interflora starts with an unreliable dataset from which it proposes to select the witnesses most favourable to itself.”

Lewison LJ also discussed the procedure which should be followed in future as follows:

“149.

For the future, the standard form of order should be redrafted so as to make it clear that:

i)

A party may conduct a true pilot survey without permission, but at his own risk as to costs;

ii)

No further survey may be conducted or adduced in evidence without the court's permission; and

iii)

No party may adduce evidence from respondents to any survey without the court's permission.

150.

In deciding whether to give permission, the court must evaluate the results of whatever material is placed before it. Only if the court is satisfied that the evidence is likely to be of real value should permission be given. The reliability of the survey is likely to play an important part in that evaluation. Even then the court must be satisfied that the value justifies the cost. As Mr Hobbs said, this requires the court to conduct a cost/benefit analysis. In a case of trade mark infringement in which the issue is one of deception in relation to the provision of ordinary consumer goods or services, these criteria are likely to be satisfied only in a special or unusual case.

151.

If what is sought is permission to carry out a survey, the applicant should provide the court with:

i)

the results of any pilot survey;

ii)

evidence that any further survey will comply with the Whitford guidelines; and

iii)

the cost of carrying out the pilot survey and the estimated cost of carrying out the further survey.

152.

If what is sought is permission to call witnesses who have responded to a survey or other experiment, the applicant should:

i)

provide the court with witness statements from the witnesses proposed to be called;

ii)

demonstrate that their evidence will be of real value in deciding the issues the court has to decide;

iii)

identify the survey or other experiment and, in the case of the administration of a questionnaire disclose how many surveys have been carried out, exactly how those surveys were conducted and the totality of the number of persons involved and their answers to all questions posed;

iv)

disclose how the proposed witnesses were selected from among the respondents to the survey; and

v)

provide the court with the cost of carrying out the pilot survey and the estimated cost of carrying out any further work in relation to those witnesses.

153.

It may have a material effect on the court's decision whether (and if so to what extent) the applicant is prepared to waive privilege in so far as it attaches to the selection, interviewing and preparation of witness statements for the witnesses proposed to be called. In the absence of a waiver of privilege in this respect, a party who wishes to challenge the evidence is likely not to be able to do so effectively; and in particular would not know what questions were asked of the witness in order to prepare the witness statement. If the evidence proposed to be called cannot be effectively challenged, that may in itself reduce its potential probative value.”

Lewison LJ also upheld Marks & Spencer’s micro objection for reasons he expressed as follows:

“154.

… It will be recalled that one of the concerns expressed in the Whitford guidelines is that questions might lead a witness into a field of speculation upon which that person would never have embarked had the question not been put. That can be seen to have happened in this case. I give some examples, all taken from the answers to question 6 in the first questionnaire (‘Thinking specifically about this search result (POINT to MARKS & SPENCER RESULT). What if anything do you think this tells you about any relationship between Interflora and Marks and Spencer?’): …

155.

Mr Hobbs also pointed to a number of answers that either showed that the respondent drew nothing from the results, or appreciated that there was no link between Interflora and M & S. Again I give examples of responses to the first questionnaire …

156.

It would be tedious to multiply examples. I simply fail to see how against that background calling a selection of witnesses would advance the case that Interflora wants to make.”

Interflora contend that, as Lewison LJ put it at [136], “consumers have been confused in the real world” when searching for “interflora” on Google and being presented with advertisements for Marks & Spencer, and that the object of the present application is to put evidence of this before the Court. In order to understand both Interflora’s case for the admission of this evidence and Marks and Spencer’s objections to it, it is necessary first to describe what Interflora have done in some detail.

Interflora decided to look for possible witnesses among their own customers in the UK. They identified two subsets of customers from marketing databases held by eDigital Research, a digital research agency used by Interflora, and by Interflora. The first subset consisted of all those individuals who had taken part in an Interflora Net Promoter Score Survey (a short survey in which consumers rate the service provided) between November 2009 and August 2012 (the entire period for which such surveys had been carried out), had agreed to being re-contacted by Interflora and had not unsubscribed from Interflora emails or provided an invalid email address. This subset numbered 27,547 individuals. Summaries of the NPS surveys have been disclosed by Interflora.

The second subset consisted of a selection of customers who had ordered flowers from Interflora’s UK website in the periods 1-22 December 2011, 28 December 2011 – 27 January 2012 and 15-16 August 2012 (periods during which Interflora knew that Marks & Spencer had been bidding on “interflora” in Google AdWords) and had opted in to receiving marketing material. The total number of customers who had ordered flowers and opted in was 181,107. This group was filtered to remove duplicates, those who had provided invalid email addresses or had unsubscribed, users who appeared to be predominantly telesales customers, users who had ordered via Tesco’s website, Irish residents, new customers, Interflora staff, users who were either deceased or had ordered flowers for a funeral or other “sympathetic” occasion during the last three months and duplicates of the first subset. The resulting group comprised 78,459 individuals.

The total of the two subsets thus identified was 106,006 people. Interflora then arranged for eDigital to send each of these an email containing a link to a “prize draw” questionnaire on 13 September 2012. It was sent again on 20 September 2012 except to recipients who had opened and read the first email. The questionnaire is headed “Interflora Customer Survey” and contained the following multiple-choice questions:

“Have you used the internet to search for information about goods or services in the last year? Yes/No/Can’t remember

When you used the internet last year, did you use a search engine to search for ‘Interflora’? [A short explanation of the terms “search engine” was provided.] Yes/No/Can’t remember

Which search engines have you used to search for ‘Interflora’? A list of 10 of which Google was the fourth

Thinking about when you have searched for Interflora using Google, do you remember any search results for websites other than Interflora? Yes/No/Can’t remember

Which links to websites do you remember seeing on the search results page? A list of nine of which “M&S” was the first and “Don’t know/can’t remember” the last

We are interested in how people have used search engines to look for the Interflora website and the experience they had. If you are happy to be contacted by Interflora about the answers you have provided in this survey, please click this box and complete your contact details below.”

A total of 6,866 people completed the questionnaire. Of those, 434 were identified as having seen the M&S link in response to a search for Interflora. Interflora has disclosed a spreadsheet containing the contact details of all these individuals. This dataset was then edited to remove those who had provided obviously false telephone numbers (such as 999), leaving a total of 395 individuals.

These 395 individuals were then interviewed by Interflora’s solicitors by telephone following a set script in order to ascertain whether they would be prepared to be witnesses and would be available to do so. The solicitors made contemporaneous notes of the telephone calls which have all been disclosed. 43 could not be contacted because they had provided a false or incorrect telephone number, 122 could not be contacted despite being called three times, 172 declined to participate or were not available or could not remember completing the questionnaire and two scripts were mislaid.

This left 56 people who had indicated that they were prepared to be re-contacted with a view to giving evidence. Each of these individuals was then interviewed by Interflora’s solicitors by telephone following a second set script. Again, the solicitors made contemporaneous notes of the telephone calls which have all been disclosed.

Of the 56, witness statements were obtained from 13. These statements were prepared by the following method. The solicitors typed the statements into a laptop computer and printed them out in the presence of the witness. The witness was asked to review the statement and add or remove anything he or she wished. Any amendments were typed into the computer and the statement was then printed and signed by the witness.

Of the other 43, 11 could not be re-contacted despite three attempts, three changed their minds about participating, five said that they had imperfect recollection of what they had actually typed into Google and/or of the results presented and 24 stated that they did not think that there was any relationship or connection between Interflora and Marks & Spencer.

By way of example, I shall quote the operative part of one of the 13 statements without identifying the witness:

“2.

I use the internet quite a lot. I use it every day at work for emails. I also use it a lot personally for email, Facebook, researching and buying things online. The time I properly shopped online was probably last Christmas, but I did book and pay for MOT online. I’m always moving money around using online banking so it feels like I’m shopping!

3.

I remember that a friend of mine was in hospital earlier this year and I wanted to send her some flowers after she came out of hospital. I used the internet to order the flowers. I trust Interflora to be good quality and provide same day delivery service so I decided to use them. Off the top of my head they were the only ones I could get through to quickly who would do same day delivery.

4.

I wanted to order the flowers online because it was more convenient. To get to the Interflora site I went to the Google page to search for ‘interflora’. When I put ‘interflora’ into Google, I clearly remember M&S coming up on the search results page, together with other companies but I can’t recall what they were.

5.

When I saw the M&S search result, I thought that there must be some supermarkets which also do online flowers. I thought if I went through the M&S site I would still get Interflora flowers. I know that some companies probably act as agents for a company like Interflora. However, I wanted to make sure I got Interflora flowers direct and I trust the delivery service so I went through the official Interflora site.”

Interflora has expressly waived privilege over the entire exercise, and Interflora’s solicitors have given evidence that all documents created during the course of the exercise have been disclosed. In addition to the 13 witnesses whose evidence is sought to be adduced, the solicitor who was primarily responsible for the exercise has made two witness statements describing in detail how it was done. Interflora have identified four others who assisted him, and have stated that all five will be made available for cross-examination at trial. In addition, an employee of Interflora UK who was involved in the identification of the two initial subsets of customers has made a witness statement explaining how this was done. No statement has been provided by eDigital, but its role appears to have been entirely mechanical. In any event I have no doubt that a witness would be made available for cross-examination if required. In short, I am satisfied that Interflora has approached the present application in an entirely transparent manner.

There is a dispute between the parties as to whether this exercise is a “survey”. Marks & Spencer says that it is while Interflora say that it is not. It shares some of the attributes of consumer surveys, but on the other hand it differs from any survey that I have previously encountered in certain respects. Thus it involves taking a sample of consumers and asking them a series of questions. On the other hand, it differs from a typical survey aimed at establishing a likelihood of confusion in that the witnesses were not respondents to a market research experiment in which artificial stimuli by way of specially-created visual material and/or hypothetical questions were administered and the responses recorded. The 13 witnesses are all actual customers of Interflora who had actually searched the internet for Interflora and who had actuallybeen exposed to the M&S advertising complained of in this action. They have been asked about this and have given evidence as to what transpired, in the sense of what they saw at the time, and what they thought of this at the time.

In my view it does not matter whether one calls this a survey or not. Interflora having applied for permission to adduce the evidence in advance of trial, the court must decide whether or not to give such permission. For that purpose, it seems to me that the approach set out by Lewison LJ should be applied. Accordingly, what matters is not the terminology, but whether I am satisfied that (a) the evidence of the 13 witnesses would be of value and (b) the value of the evidence justifies the costs involved.

It is Interflora’s case that the evidence is of value, because it is evidence of actual “confusion” (using confusion as a shorthand for the test described in paragraph 3 above, which is akin to the test of likelihood of confusion). As counsel for Interflora pointed out, evidence of actual confusion has always been regarded as highly relevant in trade mark (and passing off) cases, and this has not ceased to be the case since the courts were required to consider the question of likelihood of confusion from the perspective of the average consumer: see, purely by way of example, Och-Ziff Management Europe Ltd v OCH Capital Ltd [2010] EWHC 2599 (Ch), [2011] ETMR 1 at [114] – [117].

Marks & Spencer contends that the evidence is of no value for two reasons. Although counsel for Marks & Spencer tended to run the two reasons together in his submissions, I think it is important to disentangle them.

The evidence is not representative. The first reason is that the evidence is not representative of the legal construct through whose eyes the question of likelihood of confusion falls to be assessed. In this regard counsel for Marks & Spencer relied upon (i) the statement of Morritt LJ in BACH and BACH FLOWER REMEDIES Trade Marks [2000] RPC 513 at [35], with which Chadwick LJ agreed at 535, that “I do not think the court is assisted by repetitious evidence from individuals put forward by the parties, whether expressly or not, as archetypal consumers or end-users for by definition, no one individual is such a consumer or end-user and the issues arising cannot be resolved by counting heads” and (ii) the endorsement of that approach by the Court of Appeal in the present case in the judgment of Lewison LJ at [38]-[43]. It can be seen from my summary of Lewison LJ’s judgment above that the reason why he agreed with this approach was that he considered that it is not possible to extrapolate from the respondents to a consumer survey (or “witness collection programme”) to the relevant population of consumers unless it can be shown that the respondents represent a fair sample of the population, and that can only be achieved by means of a properly conducted and statistically significant survey.

I do not accept this argument in the circumstances of the present application for the following reasons. First, Lewison LJ made it clear at [136] that “if there is evidence of consumers who have been confused in the real world, there can be no objection to calling it.” Claimants adducing evidence of actual confusion have never been required to show that the confused individuals are statistically representative of the relevant population, and Lewison LJ does not suggest that this should be a requirement henceforth. I doubt whether it would ever be possible to do so. It is, of course, always necessary to consider whether the evidence of actual confusion, assuming it stands up at trial, is actually probative of the claimant’s case. In a trade mark case, that will involve (among other things) consideration of the extent to which the confused individuals do or do not share the attributes of the relevant average consumer. That is an exercise in evaluating the weight of the evidence which judges who try trade mark cases are accustomed to.

Secondly, I think that one has to be careful about what is meant by “repetitious evidence from individual consumers”. Morritt and Chadwick LJJ’s statements were made in the context of a case where the respondent had commissioned NOP to carry out an omnibus survey of 9,596 people in which the respondents were asked whether they were familiar with flower remedies and, if so, which brands of flower remedies they were familiar with. 1,533 people answered yes to the first question and 215 answered the second question in a manner which referred to Bach Flower Remedies. 94 of those said that they were willing to be contacted by solicitors. Contact was made with 50. Of those, eight refused to discuss the matter, nine appeared not to know what they had meant by the word “Bach” and three were regarded as unhelpful to the respondent. Of the remaining 30, 18 were selected to give evidence on behalf of the respondent. It is not entirely clear from the judgment of Neuberger J (as he then was) at first instance whether all 18 were actually called, or whether the evidence of some was taken as read: see [1999] RPC at 21. Be that as it may, the nature of the exercise was the calling of selected witnesses identified by means of a consumer survey in an attempt to prove distinctiveness of the respondent’s trade marks. Furthermore, it appears that the respondent also relied upon the survey, but the judge held that it was unreliable given that it was an omnibus survey.

Courts are never assisted by repetitious evidence in any context. Furthermore, I entirely accept that one does not decide trade mark cases (or passing off cases) by counting heads. I also agree that, if one is calling pollees in order to exemplify the results of a statistically robust survey, as Lewison LJ held could be probative at [135] and [137(ii)], then a limit must be placed on the number of such witnesses called for reasons of cost and proportionality. Where there is evidence of actual confusion, however, there have been many cases in which the courts have found it of assistance to hear from people who were confused and who are prepared to come to court. Usually, it is quite hard to find people who have been confused, because confusion by its very nature often doesn’t come to light, and it is even harder to persuade confused people to come to court to give evidence. Thus the available sample is often a small one. But experience shows that it is rarely completely repetitious. On the contrary, it often shows a range of responses to the sign in issue. In my view, it is helpful for the court to receive first hand evidence of the range of responses from the people who had those responses.

Thirdly, I think it is important to bear in mind the nature of the legal question to be decided. In the context of a claim under Article 5(1)(b) of the Directive or Article 9(1)(b) of the Regulation, the question is whether “there exists a likelihood of confusion on the part of the public”. That question must be considered from the perspective of the average consumer of the relevant goods or services. It is not a binary question: is the average consumer confused or is the average consumer not confused? Rather, it is an assessment of whether it is likely that there is confusion. It is clear from the case law that this does not mean likely in the sense of more probable than not. Rather, it means sufficiently likely to warrant the court’s intervention. The fact that many consumers of whom the average consumer is representative would not be confused does not mean that the question is to be answered in the negative if a significant number would be confused. If there is evidence that some consumers have actually been confused, that may lead the court to conclude that there is a sufficient likelihood of confusion to warrant its intervention. As I have said, the test to be applied in the present case is akin to the test of likelihood of confusion even if it is not identical.

Fourthly, I consider that the analogy drawn by counsel for Marks & Spencer with other areas of the law tells against the argument, not in favour of it. Take patents, where the legal construct is the person skilled in the art. As many cases emphasise, he (or she) is not a real person and one cannot approximate real people to him (or her): see in particular Technip France SA’s Patent [2004] RPC 46 at [6]-[15] (Jacob LJ). Nevertheless, the primary evidence as to obviousness is that of properly qualified experts in the art in question: see Mölnlycke AB v Procter & Gamble Ltd [1994] RPC 49 at 113 (Sir Donald Nicholls V-C). As the late Laddie J was fond of pointing out, parties to patent litigation often interview a number of potential experts. The experts they call at trial are those whose opinions most closely accord with the cases of the parties calling them. It has never been suggested that a party calling an expert witness in a patent case must show that his expert’s opinions are statistically representative of the relevant community of scientists or engineers. Furthermore, there are cases in which weight is placed by the courts on secondary evidence as to obviousness. One class of secondary evidence consists of the contemporaneous reactions of workers in the field to the invention and another class consists of evidence that other workers faced with the same problem either did or did not make the invention: see Schlumberger Holdings Ltd v Electromagnetic Geoservices AS[2010] EWCA Civ 819, [2010] RPC 33 at [81]-[82] and [85] (Jacob LJ). Again, it has never been suggested that it must be shown that the persons in question were statistically representative of the relevant community. Rather, the task for the court is to evaluate the extent to which such evidence throws light on what the skilled person would or would not have done.

Finally, turning to the specific facts of the present application, counsel for Marks & Spencer pointed out that there had been a series of filtering processes leading to statements from just 13 consumers having been obtained from an initial pool of 208,654 consumers. Interflora have, however, explained the filtering that was applied, when it was applied and why it was applied. At this stage, I consider that the explanations that have been provided appear to be perfectly cogent and defensible. In particular, Interflora have been open that 24 of 37 witnesses from whom it would have been possible to obtain witness statements did not think that there was any relationship between Interflora and Marks & Spencer. Counsel for Marks & Spencer complained that no statements had been obtained from those 24, but I see no need for Interflora to have done so. If Marks & Spencer wish to obtain statements from those witnesses, their contact details have been provided. Furthermore, at least one of the 13 witnesses from whom Interflora did obtain a statement also appears to fall into this category.

The evidence has been distorted by the process by which it has been obtained. The second reason advanced by counsel for Marks & Spencer is that the evidence of the 13 witnesses has been distorted by the process by which it was obtained. Indeed, he went so far as to submit that the process amounted to creating evidence rather than uncovering it.

In my view this is the real objection to most survey evidence and “witness collection programmes”. As Mr Hobbs QC himself stated when sitting as a Deputy High Court judge of this Division at first instance in Whirlpool Corp v Kenwood Ltd [2008] EWHC 1930 (Ch), [2009] RPC 2 at [35] (in a passage quoted by Lewison LJ at [124]):

“The need for circumspection in the assessment of questionnaire survey evidence is well-understood. The concern is that to a greater or lesser degree: ‘Interviews and questionnaires intrude as a foreign element into the social setting they would describe, they create as well as measure attitudes, they elicit atypical roles and responses, they are limited to those who are accessible and will co-operate and the responses obtained are produced in part by dimensions of individual differences irrelevant to the topic at hand’ Webb, Campbell, Schwartz and Sechrest, Unobtrusive Measures (revised edition, Sage Publications Inc, 2000). There is, accordingly, a practical requirement for information relating to the structure, method and results of questionnaire surveys to be full enough to enable the strengths and weaknesses of the research work to be evaluated.”

As this indicates, there are two aspects to the enquiry. First, there is the question of transparency. How has the evidence been obtained and put before the court? Only once the court knows the full details of what has been done is it in a position to consider the second question: has the process by which the evidence has been obtained and put before the court resulted in a distorted picture being presented by that evidence?

The concern for transparency was at the heart of many of the guidelines laid down by Whitford J in Imperial. With the benefit of hindsight, it can be seen that he did not go far enough. In many of the subsequent cases involving survey evidence and witness gathering exercises, there have been problems over transparency: for a recent example, see A & E Television Networks LLC v Discovery Communications Europe Ltd [2013] EWHC 109 (Ch).

As I have already said, however, in the case of the present application I am satisfied that Interflora have provided full transparency. It is worth commenting that the level of transparency that has been provided is much greater than the court is generally given in respect of most other kinds of evidence. It is very rare for disclosure to be given of the entire process by a witness has been located and by which the contents of his or her witness statement were arrived at, including all the questions that were asked of the witness by the solicitor who prepared the statement.

I turn, therefore, to the question of distortion. Now it must be recognised that it is also very rare for witness evidence to be presented to the court in an entirely unmediated manner. Generally, someone (usually a lawyer) has found the witness, established by means of questioning that the witness has (at least arguably) relevant evidence to give and then has prepared a witness statement setting out the evidence which it is proposed that the witness shall give by way of evidence in chief. It is often been remarked that one of the problems with witness statements is that they often contain, in effect, the answers to a series of leading questions. In some cases judges have passed comments about witness statements being “heavily lawyered”, meaning that the wording of the statement was more the product of the lawyer than of the witness. That is, of course, to be deprecated. Nevertheless, witness statements are, and have for some years been, an established feature of English civil litigation procedure. That being so, it would not be justified to require higher standards for the admission of consumer evidence in trade mark cases. That is particularly so where the gathering of the evidence has been made as transparent as it is here.

The concern with surveys and “witness collection programmes” lies not so much in the drafting of witnesses’ statements, but with the process by which the witness became a witness in the first place. This was the other concern addressed by the Whitford guidelines. As Lewison LJ emphasised, a consumer survey of the kind that he was considering is an experiment. If a consumer is put in an artificial situation, presented with specially-created visual material and/or asked a series of hypothetical questions, how can one be sure that the consumer’s responses reflect the responses that the consumer would have in the real world situation of which the experiment is supposed to be a model? In my view, the answer to this question must depend on two factors. The first is how faithfully the experiment replicates the real situation. The second is the extent to which the questions influence the consumer’s responses as opposed to merely eliciting them.

On the application, Interflora is not seeking to put before the court the results of an experiment in the sense that I have just described. On the contrary, it is attempting to adduce evidence of what consumers’ responses were when presented with Marks & Spencer’s advertisements when they were in the real world situation. Counsel for Marks & Spencer nevertheless attacked the process by which the evidence had been obtained on both bases.

So far as the first is concerned, he pointed out that the witnesses had been the recipients of the prize draw questionnaire and then asked two sets of questions by solicitors. He submitted that this created an artificial situation. I am unimpressed with this argument. The questions were designed to elicit and record testimony from witnesses. Any process of asking potential witnesses questions with a view to eliciting and recording their testimony in a witness statement has an artificial aspect to it. But what the witnesses were being asked about was their historic responses in a real world situation.

Counsel for Marks & Spencer also complained that the questions did not tie the responses to any particular Marks & Spencer advertisements. To my mind, that is a more pertinent point, but I do not consider that it significantly weakens the value of the evidence. Marks & Spencer has used a large number of different advertisements, and I suspect it would be very difficult for consumers to remember which they saw. More importantly, the texts of the advertisements differ only in respects which at present appear to me to be unlikely to have any bearing on the issue to which this evidence is addressed. In the respects which presently appear to matter, they are all the same.

So far as the second aspect is concerned, counsel for Marks & Spencer made two main points. First, he pointed to the fact that the prize draw questionnaire was prominently headed “Interflora”. At the moment that does not strike me as significant. More importantly, he attacked the wording of some of the questions in the second script used by the interviewers (see paragraph 28 above), and in particular question 28 in that script. Question 28 was as follows:

“From your memory of these search results, what, if anything, do you think this tells you about any relationship between Interflora and M&S?”

Counsel for Marks & Spencer argued strongly that this was essentially the same question which had been disapproved by the Court of Appeal on the last occasion, and accordingly the answers obtained could not be relied on. (I should explain that this part of the work was carried out prior to the Court of Appeal handing down its judgment.) In my view this is the strongest argument in favour of the exclusion of this evidence. Having considered the matter, however, I am not persuaded by it for the following reasons.

First, as I have explained, the context in which the question was being asked was quite different. These consumers were not being asked about an experiment which they had participated in. On the contrary, they were being asked about something they had actually experienced.

Secondly, the wording of the first part of the question is different, and is correctly focussed on the consumer’s historic experience. The wording of the second part of the question is open to criticism, because it switches from the past to the present, but I do not regard that on its own as a significant objection.

Thirdly, the major objection to the wording is that it invites speculation. In this regard counsel for Marks & Spencer took to me some answers which suggested that the respondent had indeed been led to speculate. I acknowledge the force of this point, but I am not persuaded that it renders the evidence of little or no value. By this stage the respondents had already been asked questions that shed light on their responses to the Marks & Spencer advertisements, in particular question 26 (“What did you think when you saw M&S when you entered the search term ‘Interflora’?”) and question 27 (a probing question). Even if the evidence stemming from the answers given to question 28 by those who were asked it is discounted, the evidence stemming from the answers to questions 26 and 27 would still be of value. Furthermore, due to the structure of the script, five of the 13 witnesses were not asked question 28. Finally, in the case of those who were asked it, it is not apparent that everyone was prompted to speculate.

Conclusion on value. Overall, I consider that this evidence is likely to be of some value in assisting the court to answer the question identified in paragraph 3 above. I do not think it is likely to be decisive, but nor does it appear to be of little or no value. How much weight to give it will be a matter for evaluation after the witnesses have been cross-examined and once it has been put in the context of all the other evidence available at trial.

Interflora’s total cost of obtaining this evidence is £91,999.50. That comprises a (quite small) fee paid to eDigital and (much larger) fees paid to Interflora’s solicitors. These costs are sunk costs. If the evidence is admitted, it appears that the only additional costs that Interflora will incur will be the modest cost of calling the witnesses at trial.

As for Marks & Spencer, once again it has treated this application as a substantial piece of satellite litigation in its own right despite my comments in my earlier judgment: [2012] EWHC 1722 (Ch) at [8]. Once again, it has served substantial evidence in response to it. Once again, that evidence has included an expert’s report from Adam Phillips (although at least this time it is correctly described as an expert’s report), for which Marks & Spencer did not even seek permission under CPR r. 35.4. Once again, it has transcribed the responses to the questions (at least in the case of the second script). In addition, it has instructed not one, but two, leading counsel to oppose the application. Marks & Spencer’s cost schedule for the application reveals that its costs for this work are £80,891.45.

As a result, counsel for Marks & Spencer was in a position to make virtually all the submissions about the evidence that he is likely to make at trial, with the exception of submissions arising out of cross-examination of the witnesses. The only further costs it appears that Marks & Spencer are likely to incur are counsel’s fees for preparing that cross-examination and making those submissions. In short, it appears to me that Marks & Spencer has already spent the major part of the money needed to deal with this evidence.

The costs attributable to preparing and dealing with this evidence need to be considered in the context of the total costs of the litigation and what is at stake. Interflora has estimated its total costs of this litigation to the end of trial at approximately £2.1 million. I do not have the corresponding figure for Marks & Spencer, but it appears likely that its costs will be of the same magnitude. Thus the costs of this evidence appear unlikely to be more than about 5% of the total costs of the litigation.

As for what it is at stake, Interflora’s evidence is that it estimates that the effect of Marks & Spencer’s bidding on “interflora” (and variants) as keywords on Google AdWords has been to increase its costs of advertising on AdWords by at least £1 million per year since it started, and thus over £4 million to date. Interflora also points to evidence that Marks & Spencer has made well over £1 million in sales of flowers to customers who have made purchases from Marks & Spencer after searching for “interflora” on Google. Thus it is clear that large sums are at stake for both sides.

Having regard to all the factors considered above, I consider that value of the evidence does justify the costs involved.

I shall give Interflora permission to adduce the evidence of the 13 witnesses.