Interflora Inc and another v Marks and Spencer plc and another [2013] EWHC 273 (Ch), 21 February 2013 is another example of the application of the Court of Appeal of England and Wales’ two step test to assess whether survey evidence should be permitted in trade mark infringement proceedings. The test requires the party seeking to rely on the evidence to show that it i) will likely be of value at trial and ii) the costs involved are justifiable.


This judgment in the ongoing dispute between the parties over the use of keywords concerned the Claimants’ application for permission to rely on 13 witness statements at trial. It followed another application in the same dispute to rely on witnesses identified by way of surveys, which was recently determined by the Court of Appeal.

In the main action, the Claimants allege that Marks and Spencer infringed UK and Community trade marks through the purchase of keywords that generated advertisements that did not enable (or only with difficultly) reasonably well-informed and reasonably observant internet users to ascertain whether the goods advertised were from Marks and Spencer or the Claimants. The Claimants decided to look for evidence of confusion amongst its customers by emailing a survey to a subset of them. This survey resulted in the production of the 13 witness statements the Claimants sought to rely on.


Mr. Justice Arnold was tasked with applying to the present application the Court of Appeal’s guidance on when survey evidence was permissible. He began by considering the nature of the exercise undertaken by the Claimants, concluding that, whether or not it was a survey, the Court of Appeal’s approach should be applied. The two step test required Arnold J. to decide i) whether or not the evidence would be of value, and ii) whether or not the value of the evidence would justify the costs involved.

On the first limb of the test, Marks and Spencer argued that the evidence was not representative of the legal construct of the “average consumer” and that it had necessarily been distorted by the way in which it had been gathered by the Claimants. Both these arguments were rejected. In doing so, Arnold J quoted Lewison LJ in the Court of Appeal that there could be no objection to producing evidence of actual confusion. The judge noted that this method of gathering the evidence of confused consumers, rather than subjecting samples of the population to an artificial questionnaire, did not have to be based on a statistically reliable sample (unlike survey evidence generally). The consumers were asked about something they had experienced, rather than “an experiment which they had participated in”. Arnold J considered that the evidence was likely to be of some value in assisting the court at trial.

On the question of whether or not the likely value of the evidence justified its cost, Arnold J. noted that much of the work done by Marks and Spencer in preparing for this application would have been work that it would have undertaken in any event for trial. This exercise would now not need to be repeated. The exercise constituted approximately 5 per cent of the total cost estimate for the Claimants. Arnold J. also noted that there were large sums at stake for both sides in the dispute, and accordingly found that the costs were justifiable. Permission was therefore granted to rely on the witness statements at trial.


The Court of Appeal’s approach to survey evidence in Interflora has already been applied in a number of cases. Requiring the applicant to show the evidence will likely be of assistance to the court and, moreover, is justifiable in terms of cost, is likely to result in many more challenges to the use of such evidence by the other party to the dispute. It will also lead to, and already has in a number of cases, increased scrutiny of the way in which the evidence is gathered. It is notable that the Claimants in this application waived privilege over the entire exercise and that their solicitors gave a number of witness statements as to how the evidence was gathered. Parties seeking to rely on survey evidence should be aware that this is a trend that can be expected to continue.