Interflora Inc & Anor v Marks and Spencer Plc & Anor

On 24 February 2012 I made an order, which to this extent was by consent, in the following terms:

“4 Survey evidence

4.1

neither party has permission to adduce survey evidence without first obtaining the leave of the Court. Any application for such leave is to include details of any questions proposed to be used in any such survey and details of the method and procedures proposed to be adopted in relation to the conduct thereof

4.2

any application for leave under paragraph 4.1 shall be made no later than 27 April 2012.”

On 27 April 2012 Interflora applied by an application notice of that date for an order that they be given permission to adduce witness evidence at trial from witnesses gathered from two pilot surveys conducted by Interflora on 4 and 6 February 2012 and 24 and 26 March 2012. In terms, this is not an application for permission to adduce survey evidence. Nevertheless, counsel for Interflora accepts, in the light of the reasoning of Mann J in Specsavers International Healthcare Ltd v Asda Ltd [2010] EWHC 1497 (Ch), [2010] FSR 28 at [15]-[19], that, by virtue of paragraph 4 of the order of 24 February 2012, Interflora requires the permission that it is seeking from the court.

The principles applicable to an application of the present nature were reviewed, again by Mann J, in his later judgment in A&E Television Networks LLC v Discovery Communications Europe Ltd [2011] EWHC 1038 (Ch), [2011] FSR 31:

“5.

It is common for parties to actions such as this to seek to introduce survey evidence, and that evidence is invariably expensive and time-consuming, and quite often not particularly probative because of the manner in which it was conducted, or the questions asked, or both. Accordingly the courts have, as a matter of practice, required leave to be sought before such evidence is adduced. As Arden LJ said in E-Sure Insurance v Direct Line Insurance [2009] IP & T 706:

‘Secondly, consumer surveys are costly to produce. They can, moreover, sometimes be based on the wrong questions and thus produce irrelevant or unhelpful responses or for some other reason, as in this case, be of no evidential value. There is much to be said for the practice initiated by the late Pumfrey J (as he then was) in O2 Ltd v Hutchison 3G UK Ltd [2004] EWHC 2571 (Ch)… and subsequently followed by Rimer J in UK Channel Management Ltd v E! Entertainment Television In [2008] FST 120. Under this practice, case management directions are given at an interim stage requiring the parties to seek the directions of the court as to the scope or methodology of any proposed consumer survey that the parties may desire to put in evidence at trial. Those directions can then be given in advance of the trial …’ (para. 63).

6.

This was adopted by Rimer LJ at para. 78, where he went on to say:

‘I would only add that even where the practice has been followed, so that the questions and methodology are settled with the agreement of the court in advance, it may be that things will go wrong. Experience of polls shows that they can be conducted in a slipshod manner, so one would need to examine the detail. And even the unexpected can happen – a question which one thought in advance was clear, may turn out not to have been so when the polls has been conducted. And a question which one thought was non-leading, may turn out to be so or to have caused the pollees to speculate in a way which they would not have done but for the question itself.’

7.

The justification for the practice was elaborated in Rimer J's decision in UK Channel Management where he said:

‘The reason is … because such evidence does not simply put before the court the spontaneous reaction of members of the public who have been exposed to … [a party's] advertising material in actual use or the defendant's allegedly offending advertisement in actual use; it is evidence obtained as a result of the artificial application to members of the public of stimuli directed at provoking reactions and responses. The reason for Pumfrey J's earlier direction was to ensure that their reactions and responses were provoked by appropriate questions.’

8.

In the case before me, as will appear, it is necessary to bear in mind the juridical basis of what it is that the court is doing when exercising its control. In my view it is doing (at least) the following:

i)

So far as a party is going to seek to put expert evidence before the court, the court is exercising its power to control the amount and nature of expert evidence in order to make sure the expert evidence is proper evidence, admissible, and proportionate.

ii)

So far as a party seeks to put in the actual answers to questions, the court is ensuring the evidence is admissible and probative.

iii)

So far as the court is controlling the calling of live witnesses obtained as a result of some form of survey evidence (so-called witness collection exercises) it is again ensuring that the evidence is admissible and probative. In particular, it is acting to prevent a party seeking to call a witness whose evidence is going to be tainted to an unacceptable degree by the mechanism under which it is collected (an inappropriate question).

iv)

In so doing, the court is ensuring that costs are not wasted and are proportionate. It is wrong for costs to be wasted in conducting hopeless surveys, for the other party to have to waste costs dealing with that evidence, and for court time to be wasted in dealing with it at trial.

v)

When a court is acting in this capacity it must bear in mind that it is acting at some remove from the trial. If it disallows a survey it is concluding, short of a trial, that evidence which one party wishes to adduce should not be allowed in because it will be of no or insufficient value. In embarking on that exercise it must acknowledge that there will be cases in which it is not wholly clear that the evidence in question will be valueless. In those circumstances the right course may be not to bar the evidence or survey at the interim stage, but to allow it and to have more informed argument at the trial (or conceivably at another interim stage, provided that that is a cost-effective way of going about the matter).

9.

The points raised in paragraph 8 above can be demonstrated from what actually happened in the UK Channel Management case itself before Lewison J [2007] EWHC 2339 Civ. There were 3 surveys. The first was an ‘omnibus’ survey, which was not directly focussed on the marks in question but was commissioned by many clients. The actual questions and answers were not going to be available to be put in evidence (see para. 7 of the judgment), and it was going to require an expert to interpret the survey. In relation to this survey the court was invoking its jurisdiction to control expert evidence. The survey was intended to prove ‘acquired distinctiveness’, of which there was apparently rather better evidence (para. 7). Lewison J came to the conclusion:

‘In my judgment the omnibus survey is of such little value that it is not worth the time and effort that would be expended in considering it. I therefore rule that the omnibus survey is not to be adduced in evidence.’

It was therefore ruled out as being insufficiently probative in the circumstances.

10.

The second survey (‘Malivoire 1’) suffered from a ‘thoroughly misleading question’, which ‘got the survey off on the wrong foot’ (para. 11). Rimer J had previously ordered that the questions being put to the public be disclosed to the defendant and that the defendant inform the claimants whether it agreed with the appropriateness of the question or not, and it seems that the survey had been completed, without the benefit of a further court ruling, by the time of the hearing before Lewison J. He ruled that the survey was not admissible in evidence, because of the misleading questions. He also excluded the possibility of live (or perhaps written) witness evidence coming from it because:

‘the witnesses produced as a result of a misleading survey are likely to give misleading and tainted evidence through no fault of their own’ (para. 12)

11.

The court was therefore ruling in advance that the evidence would be inadmissible as being of no probative value, and that the court could determine the question at that stage even in advance of seeing the actual evidence that the witness was going to give.

12.

The third survey (‘Malivoire 2’) was apparently more balanced and there was less objection to it (see para. 14). Lewison J was sceptical about its value but did not feel confident enough that it ‘proves nothing’ and did not rule it out; he left it to the trial judge. That illustrates my proposition (v) above.

13.

The court must, however, be alive to the dangers of the pre-trial assessment exercise, in that the permission hearings themselves may give rise to another raft of costs in what is already likely to be a costly exercise. The case before me is a case in point. The application before me was one direct to the judge (it was not an appeal from the Master). There was a leader on one side, and a leader and junior on the other. The two leaders gave me estimates of the costs involved in the CMC so far as they related to the survey evidence point (the CMC raised other points which are not dealt with in this judgment). The claimants' costs were estimated at about £30,000; the defendant put its costs of the exercise (including the correspondence between the parties debating the various issues) at between £20,000 and £30,000. A cost-saving exercise which incurs so much cost must be handled with care. It must not be allowed to become a significant cost-generator in its own right. There will be a danger of this happening if the court is asked to micro-manage an intended survey.

…

26.

I remind myself of the nature of this exercise. It is not to settle the terms of some sort of joint survey. It is not to produce a survey which the defendant finds acceptable. And it is not to rule at this stage on the force and effect of apparently subtly crafted questions, divorced from answers and on an interim hearing. The first and main task is, in my view, to give the court an opportunity to prevent a survey going forward when it can be seen at this stage, with sufficient clarity, that it will produce results which will not help, will not be relevant and/or will not be probative or sufficiently probative to make considering it worthwhile. The results may be of various kinds. They may be results which are presented by an expert who analyses them and presents conclusions. They may be results presented by putting forward the results of the survey as a whole and inviting the admission of the answers under the Civil Evidence Act. The survey may enable the surveying party to call individual witnesses to give live evidence, or separate statements of evidence under the Civil Evidence Act. In each case the court will have to consider what the result may be, and if it is sufficiently clear that it will not be useful or probative, or perhaps that the effort will be disproportionate to the benefit achieved, then the evidence will not be allowed. If the evidence has the potential to be useful, and its real significance cannot be determined at the interim stage, then by and large the party applying should be allowed to have the survey it wants. Of course, the procedure provides an opportunity for the opposing party, and indeed the court, to contribute, and an appropriately humble applicant should be prepared to acknowledge that some of the objections are good ones and modify the surveys accordingly. But at the end of the day if the applicant wishes to stick with a particular form of survey, then the real question is whether it can be demonstrated sufficiently clearly to be flawed, misleading, insufficiently probative, or disproportionate (or some combination of those all four).”

In the present case Interflora has conducted two exercises, which it describes as pilot surveys. In the first survey on 4 and 6 February 2012, 95 questionnaires were administered to survey respondents brought in from the street into a hall set up with four laptop computers. The survey respondents were seated in front of a laptop computer and asked to type in the word “Interflora” as a search term for the purposes of generating a search engine results page on screen. The search engine results page displayed to the survey respondent was pre-ordained, in the sense that it consisted of one or other of two pages which those responsible for conducting the survey had built into the presentation.

The procedure followed in the second pilot survey on 24 and 26 March 2012 was the same, save that on that occasion 102 questionnaires were administered to survey respondents. In addition, there was a difference in that, whereas on the first occasion one of two forms of search result page was presented to the respondents, on the second occasion a third form of search reports page was used. As is customary in these circumstances, a structured sequence of questions was asked and the market researchers attempted as best they could to record verbatim the survey respondents answers. Those answers were subsequently transcribed and tabulated.

As I have already indicated, it is not the purpose of Interflora’s application to adduce by way of evidence at trial either of the two pilot surveys in themselves. Nor, on the basis of those pilot surveys, does Interflora seek permission to carry out a full survey, that is to say, a survey following the same methodology, but using statistically significant numbers of survey respondents. Instead, Interflora merely seeks permission to adduce witness statement evidence, to be subject if need be to cross-examination at trial, from witnesses identified by means of the two pilot surveys.

Counsel for Interflora submits that this is a course that has been adopted in a number of cases in the past and has been accepted by judges of this Division of being at least capable of resulting in admissible evidence which can assist the trial judge. Accordingly, he submits that permission should be given.

Counsel for Marks & Spencer advances a number of reasons as to why permission should be refused. Before addressing those reasons I wish to record my concern that, contrary to the guidance given by Mann J in the A&E case, Marks & Spencer has treated the present application as a major piece of satellite litigation in its own right. To that end, they have served what purports to be a witness statement of Adam Philips dated 14 June 2012. Mr Philips is a market research expert and his witness statement concludes with the following statement in paragraph 27:

“I understand that my role as an expert witness in these proceedings is to assist the court in making its determination and my paramount duty is to the court. I have complied with and will comply with that duty.”

As that statement correctly recognises, Mr Philips’ evidence is expert evidence. Not surprisingly, Interflora felt obliged to respond in kind. Furthermore, Marks and Spencer has also served a third witness statement of Theodoris Savvides of Osborne Clarke, Marks & Spencer’s solicitors, which sets out a number of what are to my mind at this stage of the proceedings relatively minor objections regarding the conduct of the survey. That evidence reveals that Marks & Spencer’s solicitors have engaged in an exercise of re-transcribing all of the answers given by the survey respondents from the questionnaires. That will inevitably have been a costly exercise. While it may be justified for the purposes of a trial, I do not consider it appropriate on an application of this nature. It is precisely the kind of exercise that Mann J in A&E indicated should not be undertaken at this stage of the proceedings.

I turn then to the substantive objections advanced by counsel for Marks & Spencer. The most far-reaching of his objections may be summarised as follows. He submitted that a survey of this kind is in the nature of an experiment. As with any experiment, the validity of the experiment and results obtained from it depends on the precise protocol for the conduct of the experiment. Accordingly, survey evidence could only be admissible and probative if it was the result of a properly conducted and statistically reliable survey. If it was not the result of a properly conducted and statistically reliable survey, then evidence obtained by means of that survey was of no value. Furthermore, the evidence of survey respondents could not be divorced from the survey to which they were subjected. Accordingly, the evidence of selected survey respondents could not be relied in the absence of a properly conducted and statistically reliable survey.

In support of that submission, he relied strongly upon the observations of the full court at the Federal Court of Australia in Arnotts Ltd v Trade Practices Commission [1990] 97 ALR 555 at page 606:

“One theoretical possibility, in a case like the present, would be for a party to call such of the 1200 respondents to the Roy Morgan survey as were contactable. This course would have the advantage of providing a fairly selected group of witnesses, subject to any distortion which might be caused by difficulties in locating respondents. But it would add enormously to the cost and duration of a trial. The second possibility would be for a party to call evidence from a lesser number of selected witnesses. This course was taken in Ritz. The plaintiff there called 152 members of the public. The majority of these witnesses were stopped in a public place by a representative of the plaintiff and questioned as to the significance to them of the word ‘Ritz’. It seems that those who gave answers favourable to the plaintiff’s case were asked to give evidence. Those who did not, were not. As a result, the evidence of these persons was of negligible value. All that it established was that, with the expenditure of sufficient effort and money, 152 people could be found somewhere in Australia who claimed to associate the word ‘Ritz’ with the plaintiff. The 152 witnesses were not a fair sample of the general public; so that, as McLelland J noted (NSWLR at 215) there was ‘no ground in the evidence for any extrapolation on a statistical basis, or on the basis of any mathematical or logical probability, of the views of the ‘public’ witnesses (or any selection from them) as representing the views of the relevant class of the Australian public or a significant section of that class’. The plaintiff was not even willing to reveal the total number of persons interviewed; for all the judge knew, the persons who associated the word ‘Ritz’ with the plaintiff may have been a tiny minority. The tender of such partisanly selected evidence was an absurdity.”

On that basis, he argued that, given that Interflora was not seeking to put before the court survey evidence which was contended to be statistically representative, it was illegitimate for Interflora to seek to put before the court the evidence of selected witnesses obtained by means of the pilot surveys with the implicit representation that those witnesses were representative of the relevant public.

In my judgment, that is an argument which has considerable force. The problem, as I see it, at this level of the judicial hierarchy is that there is considerable precedent for judges of this Division taking into account the evidence of witnesses obtained by means of a survey in circumstances where the survey itself is either not relied upon or is conceded not to have statistical validity. Examples of cases in which this has been done include Neutrogena Corporation v Golden Ltd [1996] RPC 473, United Biscuits UK Ltd v Asda Stores Ltd [1997] RPC 513, Julius Samaan Ltd v Tetrosyl Ltd [2006] EWHC 529 (Ch), [2006] FSR 42, D Jacobson & Sons Ltd v Globe GB Ltd [2008] EWHC 88 (Ch), [2008] FSR 21 and Numatic International Ltd v Qualtex UK Ltd [2010] EWHC 1237 (Ch), [2010] RPC 25. Furthermore, in a number of those cases the judges in question not merely admitted, but also relied upon the evidence of witnesses obtained by such surveys in circumstances where they acknowledged that the questions posed in the surveys were flawed.

In those circumstances it seems to me that the present state of the law is that evidence obtained from witnesses by virtue of the fact that they have been respondents to a survey is admissible and may be probative, even though the survey itself is not and cannot be relied upon and even though the questions posed in the survey were flawed. It follows that I am unable to accede to the first submission made by counsel for Marks & Spencer.

The second submission made by counsel for Marks & Spencer relates to the questions asked in the two surveys themselves. In survey one question six was in the following form:

“Thinking specifically about this search result (points to Marks & Spencer result), what, if anything, do you think this tells you about any relationship between Interflora and Marks & Spencer.”

Counsel for Marks & Spencer submits that that is a leading question. He argues that it invites the respondent to consider a possibility of a relationship between the two companies, whether or not the existence of such a relationship had ever occurred to the respondent previously or not.

Counsel for Interflora accepts that the question is leading to a degree, but he submits that it is not so objectionable that the evidence of witnesses who answered that question should be regarded as so tainted as to be completely unreliable. Counsel for Interflora drew a distinction between questions which were positively misleading, as for example, in the well-known case of Scott Ltd v Nice-Pak Products Ltd [1989] FSR 100, on the one hand, and questions which were merely leading on the other hand.

Counsel for Marks & Spencer took issue with that distinction, referring to the judgment of Jacob LJ in Reed Executive PLC v Reed Business Information Ltd [2004] EWCA Civ 159, [2004] RPC 40 at paragraph 140.

In my view, it is not possible to draw a hard and fast distinction between questions which are objectionable and questions which are not objectionable. There is a spectrum of questions between questions which are completely open-ended on the one hand through questions that prompt speculation, to questions that have a degree of leading quality to them, to questions that are completely misleading on the other end of the spectrum. Where on the spectrum a particular question lies is a matter of judgment in the individual case. In the present case it seems to me that question six in survey one is open to criticism, but I do not consider that it is so flawed as to taint any evidence that may be obtained from witnesses as a result.

I turn then to consider question six in survey two. This was in the following form:

“Thinking specifically about this search result (points to Marks & Spencer result), is there anything that tells you it is not related to Interflora?”

Counsel for Marks & Spencer submits that that form of question is even more grossly leading than question six in survey one. Furthermore, he submits that the negative emphasis in the phrasing of the question is such as to imply, contrary to the fact, that a connection between Marks & Spencer and Interflora does exist. Accordingly, he submits that that question is even more objectionable than question six in survey one. To that, counsel for Interflora makes essentially the same answer as before.

In my judgment, question six in survey two falls on the other side of the line to question six in survey one. In my view, the implication in the question that there is a connection between Marks & Spencer and Interflora, contrary to the fact, is misleading. In those circumstances it seems to me that the responses to that question are of no value whatsoever to the court in resolving the issues between the parties in this case. Furthermore, it seems to me that the evidence from witnesses obtained by answering that question will be so tainted to be of no value. In that respect, it seems to me that survey two falls into the same category as the survey referred to as “Malivoire 1”, which was disallowed by Lewison J (as he then was) in UK Channel Management Ltd v E! Entertainment Inc [2007] EWHC 2339 (Ch) at [11]-[13].

The third major point made by counsel for Marks & Spencer was that Interflora’s application raised serious practical questions which it did not address. The primary point he made was with regard to selection of the witnesses. He pointed out that the form of order sought effectively gave Interflora carte blanche as to which of the survey respondents who had agreed that they could be re-contacted should be contacted in order to obtain witness statements from them.

So far as that is concerned, counsel for Interflora made no bones about the fact that it was Interflora’s intention to seek to obtain evidence from those survey respondents who had agreed that they could be re-contacted and which appeared to be helpful to Interflora’s case. As he pointed out, there is precedence for that approach in a number of the earlier cases to which I have already referred.

It seems to me that this is a point which goes to the weight of the evidence upon which Interflora seeks to rely. It is not of itself a reason to refuse the permission that is sought.

Counsel for Marks & Spencer went on to observe that, if such evidence was adduced at trial, there were other questions which would need to be addressed. First, the question of privilege so far as communications between Interflora’s solicitors and the witnesses were concerned. Secondly, the question of whether Marks & Spencer should have permission to adduce expert evidence from a market research expert dealing with the validity of the pilot survey by means of which the witnesses had been identified. Thirdly, the extent to which it would be possible for Marks & Spencer to establish what, in fact, had been done.

So far as the last of those points is concerned, counsel for Interflora accepted that it was incumbent upon Interflora to put before the court full details of the pilot survey that had been carried out. He submitted that that had already been done. Indeed, counsel for Marks & Spencer did not identify any particular pieces of information which had not been supplied in the light of the reply evidence served by Interflora on this application.

The questions raised as to privilege and expert evidence seems to me to be valid questions, but I agree with Mann J in the A&E case that they are not questions that need to be addressed at this stage. As I see it, if the court gives permission as sought by Interflora, as I am intending to do in relation to survey one, then the court will have resolved the issue of principle as to the admissibility of the resulting evidence. That will leave open the other matters to which counsel for Marks & Spencer referred. In the first instance, it seems to me that they are matters which the parties should seek to reach agreement. In the absence of agreement then they can be resolved on a further application.

For all of those reasons I will give Interflora the permission that it seeks in relation to the first pilot survey, but not in relation to the second pilot survey.