Blooming surveys: consumer research evidence after Interflora

The ongoing battle between Interflora and Marks & Spencer regarding keyword advertising reached another climax when the Court of Appeal laid down new guidelines on the acceptability of survey evidence in trademark infringement cases. But will the strict approach adopted by the Court of Appeal really have that great an impact on how practitioners run such disputes in future, or reduce dramatically the role of market research companies?

Both M&S and Interflora offer flower delivery services. As part of its online advertising campaign, M&S purchased the word “Interflora” and other variations as a Google AdWord. As a result, when an internet user entered the term into the Google search engine, a sponsored link and advert for M&S’s flower delivery services would appear above the search results. There was no reference to “Interflora” within the M&S advert but, inevitably unhappy about the use of its trade mark by a competitor in this way, Interflora launched trade mark infringement proceedings against M&S.

The case was referred to the Court of Justice of the European Union, which ruled that in such situations, infringement may be found where a “reasonably well-informed and reasonably observant internet user” is unable to ascertain if the goods originated from the trade mark proprietor, someone economically linked to them, or from a third party (Case C-323/09).
In other words, would such users believe that the M&S flower delivery service belonged to or was part of Interflora’s network or was an independent competitor? Such a person is a variation of the ‘average user’ construct adopted in other trade mark infringement cases.

Interflora conducted two ‘pilot’ surveys to gather potential witnesses to provide evidence of confusion. At first instance, Arnold J held that the surveys were not statistically valid and included flawed questions, but he gave permission to adduce evidence from the witnesses generated by the first survey, as it was not so flawed that the witnesses should be excluded (Interflora v Marks & Spencer [2012] EWHC 1722 (Ch)).

Exercising discretion

On appeal from M&S, Lewison LJ held that the approach taken by the courts so far – that survey evidence should be allowed unless clearly valueless – was wrong (Marks & Spencer plc v Interflora [2012] EWCA Civ 1501). While such evidence is not inadmissible per se, the court should exercise its discretion to resist the inclusion of survey evidence. His view was that even if “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”. It is therefore for the party seeking to rely on the evidence to prove its worth and proportionality.

Lewison LJ went on to set down guidelines on the admissibility of survey evidence in future cases. He determined that the form of order adopted should make clear that:

? a pilot survey may be conducted without permission but at risk as to costs;

? no further survey may be conducted/adduced without the court’s permission; and

? no evidence from respondents to a survey may be adduced without permission.

He also set down guidelines as to what a party must show when seeking permission to adduce such evidence.

In this case, Interflora had selected the witnesses most favourable to itself from an unreliable dataset. The evidence was therefore valueless and inadmissible.

In his conclusions on the value of survey evidence, Lewison LJ determined that internet search results fall within the general description of ‘ordinary consumer services’. These are services with which a judge is likely to be familiar and as such he can determine the question of what the average consumer (or here the ‘reasonably well-informed and reasonably observant internet user’) would understand without the need for evidence from experts or real internet users. However, Lewison LJ observed that there are several situations when different considerations come into play and ?where survey evidence may have greater value. Among those are cases involving passing off.

Different tests

In most cases where trademark infringement can be alleged, so can passing off. Often both causes of action are run in parallel. So, will survey evidence still find its way into trademark infringement claims and influence the judge where passing off is also pleaded?

Not necessarily. The permission of the court would still be needed to admit survey evidence and an assessment as to its value and cost will be conducted. An application to admit survey evidence, including in passing off actions, should undergo a rigorous review to ensure compliance with the guidelines set down by Lewison LJ, including conformity with the Whitford Guidelines as set out in Imperial Group Plc v Philip Morris Ltd [1984] RPC 293, (which provides for full disclosure of the details of the survey or questionnaire and all answers given) and evidence of the survey’s value.

Nevertheless, due to the legal test that applies, survey evidence is generally of greater utility in passing off cases. This is illustrated by a case heard since the Interflora decision. In FAGE UK Ltd v Chobani UK Ltd [2012] EWHC 3755 (Ch) Hildyard J observed that all surveys are imperfect and “there is a risk if the threshold is set too high that a potentially beneficial, even if imperfect, filtering process will be thrown away”.

However, if survey evidence is admitted, the tests for trademark infringement and passing off are different. Infringement in trademark cases is assessed from the perspective of the average consumer, or equivalent. The test is singular. Arguably, only the judge need answer the question, “Would I be confused?”. For passing off, however, the test is whether on the balance of probabilities, a proportion of the public would be confused or deceived as a result of the alleged misrepresentation. This requires a quantitative test to be undertaken, and is something for which a judge may require evidentiary assistance.

If there is evidence that a substantial part of the public would be confused, is it likely then that a judge would determine that he nevertheless represents the ‘average consumer’ and would not be confused? In Reed Executive Plc v Reed Business Information Ltd [2004] EWCA Civ 159 Jacob LJ commented that “if a substantial proportion of the relevant consumers are likely to be confused, so will the notional average consumer and vice versa”.

But Lewison LJ disagreed that the substantial proportion of the public and the average consumer tests amount to the same thing. For passing off, it is not necessary to show that most people would be deceived, just more than a de minimis level. Thus 49 per cent of the public may be confused, meeting the test for passing off, but if 51 per cent are not confused, it seems unlikely to follow that the average consumer is.

Back door evidence

There are certainly previous cases where a judge has determined a trademark infringement dispute without reference to evidence of consumer confusion admitted in the same case for the purposes of passing off (see for example United Biscuits (UK) Ltd v Asda Stores Ltd [1997] RPC 513). But in reality, the likelihood is that where survey evidence is admitted for passing off actions, that evidence may still be useful in establishing trademark infringement (albeit through the back door). The chances of a judge finding that the average consumer would not be confused in the face of overwhelming evidence to the contrary ?from members of the public are likely to ?be slim.

That said, the Court of Appeal has clearly indicated in Interflora that if a party wishes to adduce any survey evidence, whether a statistical survey or a witness collection exercise, that evidence must be sound. ?Full disclosure of all facts relating to the creation of that survey is essential. The burden is on the applicant to demonstrate that the survey will be valuable and proportionate. As such, any survey will need to be well drafted and fit for purpose. This at least should reduce the risk of unreliable evidence colouring judges’ views.

But if the hurdles are overcome and the proceedings include a claim in passing off, then it seems reliable survey evidence may continue to play an important role in trademark infringement disputes.

 

---

 

Asking the right questions: how surveys can work in trademark cases

Trademark infringement or passing off can significantly devalue a brand and for lawyers looking to prove infringement, demonstrating that Joe Bloggs could mistake one brand for another is crucial. To do this, gathering the testimony of real consumers via a survey has often been on the agenda, since the testimony of real people who can be cross-examined has always been highly prized by the courts.

The Court of Appeal in Interflora found that trial courts can refuse the evidence presented through surveys, with concerted efforts to make them admissible.

Although a good idea in practice, in reality, these surveys are difficult to construct and expensive to carry out. Interflora spent 250,000 on surveys which werent even used in full  and the witnesses it gathered were ultimately rejected by the court. To avoid waste or inefficiency from costly and time consuming surveys in the future, the Court of Appeal has set out a number of guidelines to advise when, and how, these witness gathering exercises can be used:  the survey should resemble a real life situation as much as possible.

The witnesses are only called to support the evidence found in a reliable and representative survey.

If the trademark in question is related to a specialist or niche market which the judge cannot step into the shoes of a reasonable consumer to understand, a specialist expert witness may be called in to provide evidence.

The court recognises that consumer surveys can provide relevant evidence and has set out a framework for their use in court. In the commercial world, market research and surveys are used frequently in conjunction with expert interpretation. As such, there is always the potential for it to be relevant to a court.

This ruling could increase demand for expert witness evidence in trademark infringement claims  with the court placing more weight on this than on consumer evidence. A professional who can offer an expert view on the impact that infringement could have on a trademarks value could give more focused evidence that is of more use to the judge than the result of a survey  especially if the case revolves around a specialist or niche product.    

The ruling is no doubt inspired in part (subliminally if not explicitly) by the ongoing debate around cutting the costs of litigation and introducing costs budgeting measures under the Jackson reforms. With the spotlight shining on cost efficiency measures, practitioners should follow the courts lead in taking a fresh look at the various evidential sources they use to underpin trademark infringement or passing off actions. 

Thayne Forbes is joint managing director at Intangible Business and has acted as an expert witness (www.intangiblebusiness.com)