Interflora, Inc., Interflora British Unit v Marks and Spencer plc [2014] EWCA Civ 1403

This article was produced by Olswang LLP, which joined with CMS on 1 May 2017.

The latest judgment in the long running saga between Marks and Spencer ("M&S") and Interflora in the English courts was handed down by the Court of Appeal this month. On appeal by M&S, Lord Justices Kitchiner and Patten and Sir Colin Rimer found that Arnold J., on first instance, had erred in law on a number of points whilst reaching first instance judgment.

The Court of Appeal was far from confident that the judge would have reached the same conclusion had he not made the errors identified. However, it was not prepared to re-evaluate the first instance conclusion. Doing so would require an evaluation of the judge's finding of facts and his assessments of factual or qualitative criteria. The appellate courts should not interfere with first instance findings of facts unless compelled to do so, and should be cautious in differing from first instance qualitative or factual assessments. The Court of Appeal was also not willing to rule on infringement without the benefit of the documents or witness evidence presented to the High Court on first instance. Accordingly, the case was remitted to the High Court for retrial.

The High Court Judgment

In May 2013, Arnold J. at the High Court ruled in favour of Interflora, finding that the purchase of (and failure to negative match) the trade mark "Interflora" as an adword by M&S amounted to trade mark infringement pursuant to Article 5(1)(a) of the Trade Mark Directive and Article 9(1)(a) of the Trade Mark Regulation. The decision joins a line of case law including Google France (C-236, 237 and 238/08), Datacard [2011] EWHC 244 (Pat), and Cosmetic Warriors [2014] EWHC 181 (Ch) which consider trade mark infringement in the context of adword use. Broadly, it has been held that the origin function of a trade mark must be affected by adword use in order for an Article 5(1)(a)/9(1)(a) double identity claim to succeed.

Arnold J. held that the M&S Flowers advertisements returned by Google on the input by the user of the Interflora mark did not enable the reasonably well-informed and reasonably attentive internet users, or enabled them only with difficulty, to ascertain that the service referred to in the advertisements originated from M&S rather than Interflora or an undertaking economically related to Interflora. The origin function of Interflora's mark had therefore been affected and M&S had infringed the mark.

The Appeal

M&S appealed the decision on a number of key grounds.

The average consumer test

On first instance, Arnold J. held that a significant proportion of average internet users (being the reasonably well-informed and reasonably attentive internet users described above) would have believed there was a connection between M&S and the Interflora network. M&S appealed on the basis that the average consumer test poses only a binary question: is the notional average consumer liable to be misled, or not? M&S contended that the judge should not have applied a statistical test in the manner that he did.

The Court of Appeal upheld the judge on this point. The average consumer and average internet user are one and the same, and do not present a statistical test under which a certain percentage of people must be liable to be confused. Additionally, the Court of Appeal clarified that the "average consumer test" applied in trade mark infringement matters was not the same as the "substantial proportion of the public" test applied in passing off matters.

However, the Court of Appeal held that the judge was perfectly entitled to consider statistical evidence, or to have regard to a significant portion of the relevant population, in reaching a qualitative assessment as to the likely state of mind of the average consumer. The Court of Appeal went on, holding that: (i) a finding that a significant proportion would not be confused would not prevent a finding of infringement; and (ii) a finding that a significant proportion would be confused may allow a judge to properly find infringement.

The reversed burden of proof

A second point of contention was that Arnold J. had applied a reversed evidential burden of proof. In doing so, he had placed the onus on M&S to show that its use Interflora's mark in context was sufficiently clear so that there was no real risk of the average consumer being confused as to the origin of its services or goods. The judge had followed his own judgment in Datacard in this regard, but took the opportunity to qualify the test by reducing a requirement of "no possibility" to one of "no real risk". In Datacard, the judge sought to differentiate the "likelihood of confusion" requirement of Article 5(1)(a)/9(1)(b) infringement and the "effect on function" requirement imported into Article 5(1)(a)/9(1)(a) in respect of adwords. M&S appealed on the basis that Datacard was not the correct precedent and was not the correct application of the Court of Justice ruling in Google France.

The Court of Appeal was not convinced by any of the judge's reasoning, and held that the "effect on function" test set out in Google France was not a formulation of the "likelihood of confusion test" required for Article 5(1)(b)/9(1)(b). It held that Arnold's reversed "evidential" burden of proof was tantamount to a reversed legal burden and should not be applied. A party seeking to rely on the protection of the Trade Mark Directive (and subordinate legislation) should be required to make out all of the elements of infringement. The Court of Justice had not expressly stated in Google France that the burden of proof should be reversed, and the French court had not done so in deciding the case. The Court of Appeal held that there was no inconsistency between requiring advertisers to ensure that their adverts are transparent and placing the burden of proof of a failure to do so on the respective claimant. Accordingly, the judge was overturned on this point.

The doctrine of initial interest confusion

Arnold J. held that initial interest confusion (whereby consumers are initially confused as to the origin of goods, but are disabused of that confusion before ultimately making a purchase), would alone be enough to establish liability for infringement on the part of M&S. This doctrine originated in US trade mark law and was considered by Arnold J. at great length in Och-Ziff Management Europe Ltd & Anor v Och Capital LLP & Anor.

The Court of Appeal did not consider the importation of this doctrine into European law at all helpful in the circumstances. The Court of Appeal held that the doctrine has the potential to disrupt fair competition and did not incorporate any of the carefully considered checks and balances set out by the Court of Justice in Google France and subsequent rulings. The judge was, therefore, overturned on this point.

Failure to negative match

Google had not always offered sale of the mark Inteflora as an adword. Prior to Google's doing so, M&S had bid on a large number of generic flower related adwords and declined to "negatively match" the mark Interflora. Negative matching actively prevents advertisers' advertisements being generated as a result of the user inputting the negatively matched adword. Arnold J. had held that M&S' failure to negatively match amounted to infringement of Interflora's mark and obliged M&S to negatively match under an injunction.

The Court of Appeal held that failure to negatively match a trade mark as an adword may amount to "use" for the purposes of Article 5(1)(a)/Article 9(1)(a). A judge would be required to make an assessment on such a point, based on all the circumstances, including the amount of time the trade mark was "used" in this manner. Such use alone will not amount to infringement, unless the other elements of infringement are also satisfied.

Comments

Whilst the Court of Appeal has held that the "effect on function" test is not a formulation of the "likelihood of confusion" test, many commentators consider the distinction as unclear and regard the development of this requirement as a significant departure from the wording of the Trade Mark Directive. For the time being, the issue is restricted to the use of adwords, and has been formulated in order to strike a balance between the rights of trade mark proprietors and the need to promote fair competition. Should the "effect on function" requirement be imported into other contexts, legislative amendment may be required.

Additionally, this is possibly the first trade mark infringement case ever to be sent for re-trial in England, and will see both parties incur significant further costs. Trade mark infringement actions often require first instance judges to make a considerable number of factual findings, based on large amounts of evidence explored via cross-examination of witnesses. Additionally, as survey evidence is becoming more difficult to adduce and is generally being afforded little weight by the courts, judges are being called upon to make qualitative assessments of the state of mind of the average consumer, based on their impressions of the evidence presented at first instance. It is not unforeseeable that the appellate courts, therefore, will be faced with similar dilemmas in the future as to whether or not they should attempt to make a factual assessment or otherwise order a retrial.