Interflora Inc & Anor v Marks and Spencer Plc & Anor

This is an application for an interim injunction. The claimant is Interflora. They are a well-known flower delivery network in the United Kingdom. I think about a third of florists in the UK are members of the network. They are a substantial and successful business, and the name Interflora is well known.

The defendant is Marks and Spencer. They are a very well known retailer. They are particularly well known for clothing and food. They are a household name in the United Kingdom. They also sell flowers.

Both operate online. Interflora has a website, and so does Marks and Spencer. Both take orders online for flowers and flower deliveries. I believe they are the two top online flower delivery businesses in the UK, at least by volume.

This dispute relates to internet searching. It has a long history. It really began in 2008, when Google changed its policy and allowed third parties to bid for adwords which were trade marks but did not belong to the bidder. So, from that time, Marks and Spencer began bidding for the adword “interflora”. Accordingly, a person typing the word "interflora" into the Google search engine, whom, one might suppose, was looking for Interflora, perhaps to buy flowers, could be presented with a so-called sponsored link that is an advertisement for Marks and Spencer. They would click through the advertisement and get to the Marks and Spencer site and, one might imagine, buy flowers. An important point relied on by both sides is that the Marks and Spencer advertisement does not contain any reference to Interflora.

Interflora contends that this amounts to trade mark infringement of its trade marks for the word INTERFLORA registered for flowers.

After proceedings started, the case was stayed for some years pending references to the Court of Justice to the European Union. At the hearing to decide on the references and to consider a stay, Arnold J held that no interim injunction pending the stay should be granted. One reason at the time was that the claimants had delayed too long before seeking interim relief.

There were two relevant decisions from the Court of Justice: so-called Google France and the Interflora decision itself. The Google France decision held, essentially, that adwords are at least capable of being infringements of trade marks. The Interflora decision considered the issues arising on the facts of this case.

It is difficult to summarise the conclusions of law that these judgments have reached. Recognising that the following is an incomplete summary, I will refer to paragraphs 96 to 100 of the judgment of the Court of Appeal in this case, which includes citations from the CJEU:

‘96. It was then for the national court to make its assessment, and to do so from the perspective of the reasonably well-informed and reasonably observant internet user:

"53.

Having regard to that situation and to the other matters that it may consider relevant, the referring court will, in the absence of any general knowledge such as that referred to at [51] of this judgment, have to determine whether or not the use of words such as "Marks and Spencer Flowers" in an advertisement such as the one set out at [19] of this judgment is sufficient to enable a reasonably well-informed and reasonably observant internet user who has entered search terms including the word "interflora" to tell that the flower-delivery service offered does not originate from Interflora."

97.

Guidance on the effect on advertising function is set forth in the decision from [54] to [59]. The Court restated that the use of a sign identical to another person's trade mark in the Google referencing service did not adversely affect the advertising function of that mark, even though it might have repercussions on the use of the mark in advertising. In this connection the Court recognised that the use of such a sign as a keyword meant that the proprietor of the trade mark might have to pay a higher price per click than the competitor if it wished to ensure that its advertisement appeared before that of the competitor. But this did not necessarily mean that the trade mark's advertising function was adversely affected:

"57.

However, the mere fact that the use, by a third party, of a sign identical with a trade mark in relation to goods or services identical with those for which that mark is registered obliges the proprietor of that mark to intensify its advertising in order to maintain or enhance its profile with consumers is not a sufficient basis, in every case, for concluding that the trade mark's advertising function is adversely affected. In that regard, although the trade mark is an essential element in the system of undistorted competition which European law seeks to establish (see, in particular, Case C-59/08 Copad SA v Christian Dior Couture SA [2009] E.C.R. I-3421; [2009] F.S.R. 22 at [22]), its purpose is not, however, to protect its proprietor against practices inherent in competition.

58.

Internet advertising on the basis of keywords corresponding to trade marks constitutes such a practice in that its aim, as a general rule, is merely to offer internet users alternatives to the goods or services of the proprietors of those trade marks (see, to that effect, Google France [2010] R.P.C. 19 at [69]).

59.

The selection of a sign identical with another person's trade mark, in a referencing service with the characteristics of "AdWords", does not, moreover, have the effect of denying the proprietor of that trade mark the opportunity of using its mark effectively to inform and win over consumers (see, in that regard, Google France [2010] R.P.C. 19 at [96] and [97])."

98.

Again this is a significant passage for it shows an appreciation by the Court that internet advertising using keywords which are identical to trade marks is not an inherently objectionable practice. On the contrary, its aim is, in general, to offer to internet users alternatives to the goods or services of trade mark proprietors and it is not the purpose of trade marks to protect their proprietors from fair competition.

99.

Similar guidance was given by the Court in relation to the investment function. First, it explained the nature of this function:

"62.

When the use by a third party, such as a competitor of the trade mark proprietor, of a sign identical with the trade mark in relation to goods or services identical with those for which the mark is registered substantially interferes with the proprietor's use of its trade mark to acquire or preserve a reputation capable of attracting consumers and retaining their loyalty, the third party's use must be regarded as adversely affecting the trade mark's investment function. The proprietor is, as a consequence, entitled to prevent such use under Article 5(1)(a) of Directive 89/104 or, in the case of a Community trade mark, under Article 9(1)(a) of Regulation 40/94.

63.

In a situation in which the trade mark already enjoys such a reputation, the investment function is adversely affected where use by a third party of a sign identical with that mark in relation to identical goods or services affects that reputation and thereby jeopardises its maintenance. As the Court has already held, the proprietor of a trade mark must be able, by virtue of the exclusive right conferred upon it by the mark, to prevent such use (Case C-324/09 L'Oréal SA v eBay International AG [2011] R.P.C. 27 at [83])."

100.

But again, the investment function does not provide a means to protect trade mark proprietors against the effects of fair competition, even if such competition means that these proprietors have to adapt their efforts to acquire or preserve a reputation capable of attracting and retaining customers:

"64.

However, it cannot be accepted that the proprietor of a trade mark may—in conditions of fair competition that respect the trade mark's function as an indication of origin—prevent a competitor from using a sign identical with that trade mark in relation to goods or services identical with those for which the mark is registered, if the only consequence of that use is to oblige the proprietor of that trade mark to adapt its efforts to acquire or preserve a reputation capable of attracting consumers and retaining their loyalty. Likewise, the fact that that use may prompt some consumers to switch from goods or services bearing that trade mark cannot be successfully relied on by the proprietor of the mark."’

The Court of Justice in Interflora makes the point that a key aspect of this case was the fact that Interflora is a network. The point Interflora makes is this: because Interflora is and is known by the public to include a network of florists, there was nothing inherently improbable about the idea that Marks and Spencer might be part of the Interflora network. So, when a user searches for "interflora" and sees the Marks and Spencer advert, they or some of them will assume that Marks and Spencer is part of the network, and the fact that there is no reference to Interflora does not tip them off that Marks and Spencer is independent or is a rival of Interflora.

Also before trial, there were two outings to the Court of Appeal in this case relating to survey evidence.

In the end, the matter came to trial before Arnold J and he gave judgment for Interflora. After trial, he granted an injunction to restrain trade mark infringement. He heard argument on the form of the injunction. Interflora wanted it to restrain, in effect, any Marks and Spencer advertisement produced following typing the word "interflora" into your search engine. Marks and Spencer wanted the court there and then to ratify some advertisement wording which had not been considered at trial.

Arnold J rejected both approaches, and granted an injunction in the normal form to restrain infringement of the trade marks. A stay pending appeal of the injunction was refused.

The matter came before the Court of Appeal. The Court of Appeal accepted Marks and Spencer's submission that the trial judge had made important legal errors. At paragraphs 174, 177 and 183 the Court of Appeal decided as follows:

‘174. We have carefully considered the other complaints that Mr Hobbs makes about the judge's approach to the evidence but these were not developed in oral argument and we do not believe they have any independent force or add to the points we have addressed, although, as we reiterate below, we recognise that the judge's approach to the burden of proof has likely influenced his assessment of all of the evidence and consequently affected all of his findings.

…

177.

In assessing these rival submissions, we accept that appellate courts should not interfere with findings of fact made by trial judges unless compelled to do so. This is a matter we have had well in mind throughout. We also accept that, in a case such as this, a trial judge is called upon to make an evaluative or qualitative assessment from which an appeal court should be very cautious in differing. We acknowledge too the guidance given by the Court of Justice in its decision in this case, Interflora (CJEU). Nevertheless, we are far from confident that the judge would have come to the same conclusion had he not made the errors we have identified. We say that for the following reasons.

…

183.

Furthermore, despite all of the foregoing and having considered all of the matters to which the Court of Justice had indicated he should have regard, together with those matters relied upon by Interflora, the judge's overall assessment (at [310]) was that the majority of consumers who click on M & S's advertisements do so because they have been persuaded to take their custom to M & S and not because they believe that M & S is part of the Interflora network. Nevertheless, as he went on to explain, that did not exclude the possibility that a significant proportion did believe that there was a connection between M & S and Interflora. Then, in his overall conclusion (at [318]), he held that a significant proportion of the consumers who searched for "interflora" (and like signs) and then clicked on M & S's advertisements displayed in response to those searches, were led to believe, incorrectly, that M & S's flower delivery service was part of the Interflora network. This was therefore a finely judged decision and we are satisfied that it depended in significant part upon those particular findings in relation to which the judge had earlier fallen into error.’

The court decided to order a retrial and discharged all the orders made at or after the trial, including the injunction.

When the Court of Appeal indicated that it would discharge the injunction, Interflora indicated that it wished to seek an interim injunction pending the retrial. The Court of Appeal granted a short extension of time to allow Interflora to apply to the Chancery Division for an injunction if it wished. Interflora did so; and this is the hearing of Interflora's application for the interim injunction pending the retrial.

Normally in intellectual property cases, the question of whether or not to grant an interim injunction is governed by the famous American Cyanamid case. One considers a triable issue; if there is a triable issue in the claimant's favour and, for that matter, a triable defence, one considers uncompensatable harm and looks at the balance of convenience and, if appropriate, considers the status quo.

However, section 12(3) of the Human Rights Act provides as follows:

"No such relief [which might affect the exercise of the Convention right to freedom of expression] is to be granted so as to restrain publication before trial unless the court is satisfied that the applicant is likely to establish that publication should not be allowed."

In this case, the defendant submits that section 12(3) applies. That is on the ground that the injunction sought engages Article 10 ECHR because it involves the defendant's right to freedom of expression. So, the defendant says, there should be no interim injunction unless the claimant is likely to win at trial, and "likely" means more likely than not or probably.

The claimant reserves the argument that the Article 10 right is not engaged in this case, but contends that even if section 12(3) is engaged, the authority of Cream Holdings Ltd v. Banerjee [2004] UKHL 44 shows that it is not simply the case that one always has to ask if the claimant is probably going to win before granting an interim injunction. The test is more nuanced, involves a careful consideration and balancing of various Convention rights. In this case the balance is between the claimant's right of property, including intellectual property and the Article 10 right to freedom of speech. The claimant also submits that one needs to bear in mind that the "speech" in this case is commercial speech, and that is afforded lesser protection under the Convention than other kinds.

As I say, the claimant wishes to reserve whether an advertisement like this really engages free speech at all. It points out that this is not a case in which the media are seeking to publish facts or opinion unwelcome to the claimant which may be a breach of confidence or privacy. This is not even a case like comparative advertising -- and Boehringer Ingelheim Ltd v. Vetplus [2007] Bus LR 1456 is referred to -- in which a rival trader is seeking to make a public statement about its rival. Here, the speech in the advertisement is really just a statement "here I am".

The defendant contends that although of course the advertisement in question is commercial speech, nevertheless the injunction engages Article 10, because it is an advertisement. The defendant points out that the text of the advertisement includes statements to consumers about Marks and Spencer. The defendant also refers to Casado Coca v. Spain (1994) 18 EHRR 1 at paragraphs 35-36 and, again, to Vetplus. I have also been referred by the claimant to the judgment of Eady J in CDE v. MGN [2010] EWHC 3308 and to Markt Intern Verlag v Beerman ECHR Case No. 10472/83.

In my judgment in this case, the Article 10 right of the defendant is engaged. What the defendant wishes to do is publish an advertisement, and, to that extent, this injunction could engage its freedom of speech rights.

But it is important not to overstate of importance of the defendant's Article 10 rights in a case like this. First, it is, indeed, commercial speech. It is not about artistic or political freedom to express facts or opinions or to speak unwelcome truths to power. This case is really about competition, fair or unfair; and that is how the CJEU analysed it.

The important collision of rights here is between the claimant's intellectual property right, its trade mark, and the defendant's right to compete fairly with the claimants. Although the advertisement complained of is an advertisement and, in that sense, engages Article 10, a key element of it, which is an important part of the real dispute, is the fact that a user clicks on it and then is linked to the Marks and Spencer website. In that sense, it is not merely an advertisement; it is a link to a shop, and selling flowers can be the result.

However, since Article 10 is, in my judgment, engaged, it seems to me that section 12(3) does apply.

I turn to Cream Holdings Ltd v. Banerjee. In that judgment, Lord Nicholls considers the meaning of the word "likely" from paragraph 12 onwards. In paragraph 21 Lord Nicholls considers other contexts in which the word had been given a flexible interpretation and he concludes at paragraphs 22 to 23 as follows:

‘22. In my view section 12(3) calls for a similar approach. Section 12(3) makes the likelihood of success at the trial an essential element in the court's consideration of whether to make an interim order. But in order to achieve the necessary flexibility the degree of likelihood of success at the trial needed to satisfy section 12(3) must depend on the circumstances. There can be no single, rigid standard governing all applications for interim restraint orders. Rather, on its proper construction the effect of section 12(3) is that the court is not to make an interim restraint order unless satisfied the applicant's prospects of success at the trial are sufficiently favourable to justify such an order being made in the particular circumstances of the case. As to what degree of likelihood makes the prospects of success 'sufficiently favourable', the general approach should be that courts will be exceedingly slow to make interim restraint orders where the applicant has not satisfied the court he will probably ('more likely than not') succeed at the trial. In general, that should be the threshold an applicant must cross before the court embarks on exercising its discretion, duly taking into account the relevant jurisprudence on article 10 and any countervailing Convention rights. But there will be cases where it is necessary for a court to depart from this general approach and a lesser degree of likelihood will suffice as a prerequisite. Circumstances where this may be so include those mentioned above: where the potential adverse consequences of disclosure are particularly grave, or where a short-lived injunction is needed to enable the court to hear and give proper consideration to an application for interim relief pending the trial or any relevant appeal.

23.

This interpretation achieves the purpose underlying section 12(3). Despite its apparent circularity, this interpretation emphasises the importance of the applicant's prospects of success as a factor to be taken into account when the court is deciding whether to make an interim restraint order. It provides, as is only sensible, that the weight to be given to this factor will depend on the circumstances. By this means the general approach outlined above does not accord inappropriate weight to the Convention right of freedom of expression as compared with the right to respect for private life or other Convention rights. This approach gives effect to the parliamentary intention that courts should have particular regard to the importance of the right to freedom of expression and at the same time it is sufficiently flexible in its application to give effect to countervailing Convention rights. In other words, this interpretation of section 12(3) is Convention-compliant.’

It is clear from this that, in general, the threshold for the test of likelihood means “probably will win”, “more likely than not”, but there are clear exceptions and in the end, the court has to conduct a careful balancing of rights and freedoms.

I turn to the merits. Mr. Silverleaf, on behalf of his client, the claimant, says his client will probably win. He says that while he accepts that the Court of Appeal thought that the conclusion of the judge at trial was based on a view which was, in significant part, influenced by findings of fact reached as a result of legal errors and, hence, there was to be a retrial ordered, nevertheless, when one comes to look at facts left untouched by the Court of Appeal's findings and reasoning which were not focused on by the Court of Appeal, it can be seen that the claimant can truly say now that although it is not a case they will win, it is a case in which they will probably win.

The essential point is this. It is undeniable that the claimant is a network, and, submits Mr. Silverleaf, there is evidence that the public are used to seeing members of the network without Interflora branding, although they are Interflora members. So, for example, when a member of the public sees the Marks and Spencer advertisement without Interflora branding, having been shown it as a result of typing "interflora" into their search engine, the lack of Interflora branding would not indicate that Marks and Spencer is not part of the network.

The claimant also draws attention to Interflora's relationships with existing supermarkets, such as Tesco's, Asda and the Co-op (at least now, or recently), and so, it submits, there was nothing inherently unlikely about a link between Interflora and a supermarket.

At the same time, the claimant also points out that both Interflora and Marks and Spencer are in what has been called the top middle sector of the flower market, with a very similar or almost identical demographic of customers.

There are two other factors that I wish to mention. The claimant makes a point about an anomalous effect which is alleged to exist in relation to Marks and Spencer's use of the Interflora adword, as opposed to other adwords which Marks and Spencer used. This was something that was mentioned at trial. On that point, Miss Himsworth says that Marks and Spencer never had a chance to call evidence to deal with it at trial and it will be something to look at at the directions for the retrial.

The claimant also relies on the Fleurop decision of the Bundesgerichtshof (Case I, ZR 53/12). Fleurop is the German branch of Interflora. The case there involved the same argument: that a rival business not in the network was buying the trade mark as an adword and presenting an advertisement which does not mention the trade mark to internet customers. It was held to infringe.

On that, Miss Himsworth says the advertisement in question is not the same. The one in Germany, submits Miss Himsworth, uses entirely or essentially generic language, whereas in this case Marks and Spencer's advert emphasises their well-known and distinctive trade mark. Of course, Interflora say this does not meet their point, because they are a network. But I do note that the Court of Appeal in this case had the Fleurop decision in mind and, nevertheless, still ordered a retrial.

In my judgment, the correct analysis of the parties' position here on the merits is that each side has a properly arguable, sensible case. Either side could well win. In my judgment, the merits are finely balanced, and it is not realistic or possible to say that one side has a better case than the other.

Despite Mr. Silverleaf's submission, I am struck by the nature of the Court of Appeal's decision and its approach in this case. It seems to me that, given the Court of Appeal in this case, it is impossible to say that Interflora are more likely than not to win. So, if a proper approach under section 12(3) to this case requires the balance to tip in favour of the claimant on the merits, then I should not grant an injunction.

However, it seems to me that in a case like this, it is not enough to stop there. Cream Holdings shows that the balance of rights and of harm, and so on, may well mean that an interim injunction is appropriate even if the court cannot say that the claimant will probably win. In my judgment, this is especially so where, as here, this is in no sense a weak case by the claimant. As I said, the merits are finely balanced. It is clear to me that the merits come well past satisfying the American Cyanamid level.

Also, I note that, as was mentioned in argument, trade mark and passing off cases have long been recognised as ones in which the merits are often found to play a role when applying the American Cyanamid approach, because the strength of the case and the likelihood of unquantifiable harm to both sides are often likely to be interrelated.

As both sides accepted, a key problem in this interim situation is to face up to the consequences of being wrong. That is why there is such a problem with balancing privacy against free speech. Once something has been publicised, the privacy has gone, and so one might otherwise say that one should always grant an injunction pending trial; but then by taking that approach one may be stifling free speech, which is very important in a democratic society.

It is to the uncompensatable harm which attention has to be turned, because, necessarily, the court later cannot do much or anything about it. Considering the option that the court refuses an injunction but the claimant wins, then the focus has to be on what harm the claimant would suffer by the refusal of the injunction. If that harm can be compensated in damages, and assuming the defendant is good for the money, then, despite the claimant not liking it, the fact is that the lack of an injunction was not irredeemable. And vice versa if an injunction is granted but the defendant wins, then one needs to look at the harm the defendant suffers in the meantime.

I turn to consider the factors that apply in this case.

First, I will say that I was not impressed by arguments about negative matching. I will not define this in any more detail. One can find that explained in the judgments of the Court of Appeal and the trial judgment. It is an important element legally, and it did put a positive obligation on the defendant after trial, but, on the evidence, very few searches of the kind for which negative matching is important appear to be undertaken. For example, the frequency of search for the text string "interflora flowers" represents 0.6% of the searches made using the word interflora. The vast majority of searches using the word interflora just use the word interflora itself alone. No doubt, this is because Interflora has a connotation of flowers anyway; it is a mark known to be about flowers. I can see that a different search string -- for example, "Marks and Spencer's flowers" -- works in a different way. Here, the word "flowers" adds to Marks and Spencer from the searcher's point of view, because the trade mark in that string could relate to many different kinds of thing. In my judgment, since the relevant kind of searches using the word interflora as part of a larger composite are so rare, considerations based on them and on negative matching are not going to be decisive in this application.

Another minor matter is the impact of the Quality Score of the defendant. This is a score given to bidders by Google. I do not believe in this case that it is a source of any relevant irredeemable harm to the defendant.

Another matter I mention, to say that I have not forgotten it but I do not regard as important, is the argument about infringement under Article 9(1)(c) of the Regulation. There was no appeal from the judge's judgment that 9(1)(c) was not infringed. In my judgment, that has no significant consequences for the matters I have to decide this afternoon.

I will assume, without deciding or making any indication either way, that a retrial will take place in this case in approximately one year. Of course, it would be nice if it could be much earlier, and it may be a bit later. In any case, there will be a need for a CMC. The significance of that is that in the course of deciding whether to grant an injunction or not, I need to take into account roughly how long the market will need to exist in the relevant state.

Now I will consider the position of the claimant. In the period up to trial, Interflora adopted a strategy called Golden Box. The so-called Golden Box is, effectively, the top three advertisements on the search engine results page (or SERP). It seems to be the case that these are the ones that matter from an advertiser's point of view. This appears to be where the consumer looks.

Interflora adopted an approach of bidding in such a way that, despite Marks and Spencer bidding for the word interflora, Interflora Group itself outbid them, so that the top three advertisements (that is the ones in the Golden Box) were all Interflora advertisements and none were Marks and Spencer. But this was costly. There is clear evidence of what it cost Interflora for that period.

After trial, the cost of the Golden Box strategy dropped very substantially. On the figures, as I recall them, the cost was about a tenth of the cost before trial. In fact, it turned out there was no need for Interflora to adopt it at all.

I should mention third parties. There are others who might bid for the interflora keyword when they are not members of the Interflora network. On the evidence, I conclude that although many companies have not given undertakings to Interflora not to bid, they are, in effect, following what Marks and Spencer are permitted or not permitted to do in this case. When Marks and Spencer were enjoined after trial, third parties appear to have stopped bidding for interflora, too. It seems to me that the correct way to approach this case is on the basis that third parties will follow whatever Marks and Spencer are or are not permitted to do. I reject the submission that, in that sense, the injunction could be said to discriminate against Marks and Spencer. It might be true in theory but, in practice, as I have said, the evidence shows that third parties generally stay off the market if Marks and Spencer stay off as well. There was a blip recently in the evidence, but it was chased away by Interflora.

I also reject the argument from Interflora that harm caused by third parties will be significant if Marks and Spencer start bidding. There is no good evidence that harm caused by any third parties will be any more significant or a different character from the harm alleged to be caused by Marks and Spencer itself.

If no injunction is granted, then Marks and Spencer will start bidding for the interflora adword and Interflora will need to adopt its Golden Box strategy in order to protect itself (as it would see things). That strategy will work against third parties just as much as against Marks and Spencer.

Interflora submit that this extra cost is irrecoverable, or at least risks being irrecoverable, because of the judgment in the Google France case by the CJEU, in which they said that increased advertisement costs were not recoverable. In my judgment, what the CJEU were talking about was not damages flowing from a case of infringement, but increased advertising costs where no infringement took place. I have not had my attention directed to any part of the judgment of the CJEU which could be interpreted as indicating that they were focusing there on damages arising when the advertisement in question had been found to infringe.

In my judgment, the damages that I am talking about (that is to say, of the cost of running a Golden Box strategy) are recoverable, or at least are sufficiently likely to be recoverable, if the claimant wins the case, that any doubt about it is not great enough to mean that it should weigh substantially in the balance.

Taking the Golden Box approach by the claimant in response to alleged infringements is a natural, foreseeable and reasonable response. I do not see why damages would not be recoverable if the relevant acts were found to infringe. Moreover, this damage is quantifiable. The data already exists, or would exist, to calculate it, if required.

I turn to consider some other factors. There was an issue about the quality of the service offered by Marks and Spencer. There is nothing in this point. On the evidence, there is no reason to suppose that Interflora would be harmed by the quality of Marks and Spencer's service if customers mistakenly buy Marks and Spencer flowers thinking they were a part of Interflora. Both parties in this case are clearly well-respected and successful deliverers of a high quality flower service.

The other unquantifiable loss that would be suffered by the claimant is a loss of the distinctiveness of the brand and a harm caused by the effect of customers who will be looking for Interflora and find Marks and Spencer instead. As the claimants point out, they may be lost customers to the claimant and they may be very hard to find and to disabuse of their mistake.

I accept that there will be unquantifiable harm to Interflora if no injunction is granted. Nevertheless, I must say I am not convinced that it will be overwhelming in nature.

Turning to the defendant, it seems to me that the injunction, if granted, will cause unquantifiable harm to the defendant as well. Mr. Silverleaf urged that Marks and Spencer are a very big operation, and so the impact on their overall business by an injunction would be tiny.

I accept that, looked at that way, that is true, but I do not accept that that is the whole story. It seems to me to be legitimate to consider the impact on the defendant as an online flowers business, and I consider two aspects.

First, the claimant says that the defendant's evidence that was given in response to an application for Island Records v. Tring disclosure, and which was served after a trial, shows that the financial impact of the injunction will be insignificant; the harm to the defendant will be very slight.

There was a major dispute about this evidence. It is clear that Interflora have breached the order made by Arnold J, in that the contents of the disclosure were disclosed to individuals within Marks and Spencer whereas it was supposed to be disclosed in a way that was limited to independent lawyers and accountants only. This was a serious breach. But at the previous hearing in this case to give directions over to this hearing, I permitted Interflora to rely on that evidence. My reason was that although it was disclosed under a confidentiality order, the disclosure was not given in a manner which purported to restrict Interflora, via its lawyers, from using that information within the confines of the case at the point when it was disclosed to Interflora's lawyers. Clearly, they cannot use it outside the case and, clearly the information should not have been disclosed to the staff at Marks and Spencer itself, but Interflora's lawyers were entitled to know what the evidence of Ms. Devlin was and were entitled to deploy it in these proceedings, if it was relevant.

I should say that I have already directed Marks and Spencer to produce witness statements explaining in detail what had happened. That is for another day.

Marks and Spencer also argue that this episode indicated a lack of clean hands and that was a separate reason for refusing an injunction. I do not agree, and I do not propose to spend a lot of time on that point. As counsel for the claimant said, I permitted the claimant to use this material. The defendant submitted that the hands were still (as it was put by Miss Himsworth) "grubby". In my judgment -- and I am not belittling the significance of the breach of confidence -- as a factor to weigh in the balance of this case, it is very far from the central issues.

I will avoid stating the numbers in evidence, given the dispute. All I will say is this. The evidence shows that the direct financial impact on the defendant's flowers business is really quite low. However when it is looked at in the context of the overall profitability of the business as a whole, it is not clear to me that it is really far out of line, relative to the rest of Marks and Spencer's business.

The second aspect is this. Marks and Spencer say that the impressions created by the adverts produced in response to the keyword "interflora" represent about 10% of Marks and Spencer online flower impressions. So, it is submitted, a substantial part of its online advertising for its flowers business comes in that way.

I think that is right, on the basis of the data. It seems to me to show that this is a modest, but not insignificant, part of Marks and Spencer’s flowers online advertising. To lose it would impact on the business in an unquantifiable way.

The third point is that Marks and Spencer refer to the impact on other parts of its business. The situation under consideration is this: the customer types the word "interflora" into their search engine, they see the Marks and Spencer advertisement, they go to the Marks and Spencer website; they may buy flowers or may not, but they go on and perhaps buy or look at other things -- shoes, food or whatever. I accept that it exists and the inability to do it in the meantime will have an unquantifiable impact. However, there is no evidence before me that this is a very substantial element that needs to be taken into account in this case.

I will review where I have got to. Each side in this case will suffer unquantifiable harm which they ought not to suffer if they are right on the merits, whichever course I take. It is true that Interflora may very well also suffer compensatable harm from its Golden Box strategy but, as I have said, that is compensatable. I should say that I reject the argument that Interflora are not good for the cross-undertaking which they are offering.

I turn to consider the status quo. Mr. Silverleaf says I should weigh it in the balance rather than consider it at the end. In practice in this case, I am not sure it makes much difference on the findings I have reached. The claimant says that status quo is that Marks and Spencer is not bidding and that I should take this into account and continue that state of affairs up to trial.

The claimant refers to the decision in Garden Cottage Foods ([1984] 1 AC 130) and the speech of Lord Diplock at p140 B-D. It submits that if I apply what was said there, the status quo is either the start of the hostilities, which would be 2008, or it is today. Mr. Silverleaf says, taking the status quo as the position in 2008 is so long ago as not to be meaningful; so, I must choose today as representing what is the status quo.

The defendant says the status quo was the position just before trial, that is what the Court of Appeal intended to put the parties back to. The claimant says that although it is true that the Court of Appeal intended to put the parties back to that position in legal terms, as a matter of fact, irrespective of the parties' legal rights, the fact remains that Marks and Spencer is not bidding today and I cannot and should not ignore that.

First, it seems to me that Lord Diplock was not trying to lay down a test to identify what was or was not the status quo which would be applicable to resolve this particular argument. The facts of this case are a long way from the facts being considered by Lord Diplock. The factual position is, of course, that Marks and Spencer is not bidding because they were ordered not to by the court as the result of a judgment which has been set aside and a retrial ordered.

While I cannot ignore the facts as they are, I also cannot ignore the reasons why the facts are as they are. Also I should note that in the period up to trial, running for about five years, there was no injunction.

A further important factor is the following. I am not convinced that the state of affairs in issue in this case is vulnerable to perturbations. What I mean is this. I very much doubt that the Marks and Spencer's advertisement appearing on the SERP for “interflora” for five years and then not appearing for 17 months and then either appearing or appearing for another year, will have a major long-term effect on customer behaviour. In my judgment customers do not sit and look at how the SERP for “interflora” changes over time. I very much doubt anyone would notice. That is not to say that the issue of infringement or not is not important; it is about the long-term nature (or not) of this market.

This is not like a confidential information or privacy case in which the status quo is very possibly that an embarrassing fact is unknown. In that case the status quo is very vulnerable to being changed irrevocably. It is also not like the pharmaceutical injunctions cases with generics where there is tangible evidence that a change in the price will lead to a permanent alteration to the market.

The fact is that right now Marks and Spencer is not on the market. However, if that is the “status quo” then it is not a strong point in the claimant's favour. The position will change if I let Marks and Spencer on to the market and it will change back again if they are injuncted after trial. None of this will have irrevocable long-term consequences.

Another factor I need to take into account is the scope of injunction. The defendant points out that the injunction prevents all bidding for the word "interflora". The defendant says this goes wider than the judgment at trial and refers to the form of order hearing which rejected Interflora's attempt to obtain such an injunction. An order of the kind sought now assumes that no advert can be produced by Marks and Spencer which would not infringe. But, submits the defendant, the law and particularly Google France, makes it clear that Interflora cannot say that is so; Marks and Spencer are entitled today bid for adwords as long as the advertisement does not fail the relevant legal test. However, as the claimant fairly points out, no candidate advertisement has been advanced by Marks and Spencer; so, although the point is true as a matter of legal logic, in fact in this case it is not a major factor.

Although interim injunctions do always prevent the defendant from doing something which the court has not decided yet is unlawful and may therefore turn out to have prevented lawful acts, this point is different because this injunction will prevent acts which are, on any view, always lawful. That, in principle, is of more significance. It seems to me that the court can in an interim injunction case in proper circumstances grant an injunction which restrains lawful activity as part of the price of holding the ring, but the nature of the injunction will be important. I was attracted by Marks and Spencer's position initially but I think there is real force in Interflora's retort that although, in theory, Marks and Spencer is right, there is no evidence in practice that this will be a real fetter or Marks and Spencer. The kinds of advertisements which Marks and Spencer says they wish to run are the ones found to infringe at trial. There are no different advertisements advanced before me which the court could consider in order to rule whether or not they are infringing, even arguably. To hold this against Marks and Spencer that they have not produced such an advertisement is not, despite Miss Himsworth's submission, to put the legal burden on Marks and Spencer. This is an application for an interim injunction. If there was real substance to the point Marks and Spencer could and should have produced evidence before the court.

Mr. Silverleaf submits that the key task of the court is to hold the ring pending trial and that to grant an injunction will do less harm if I turn out to be wrong at trial and Marks and Spencer win; than the alternative, if I refuse the injunction and turn out to be wrong at trial. In that case more harm will be caused.

If I thought that was the right way of looking at it, I would be inclined towards granted in interim injunction pending trial. However, I am not satisfied that the claimant is right about the harm. It seems to me that whichever course I take will cause real harm to the other party if I turn out to be wrong, which harm will be uncompensatable. The uncompensatable harms, to the extent they can be compared, are comparable. It is not a case in which one harm is clearly more severe or larger (albeit uncompensatable) than the other. I also bear in mind that the Golden Box damages are real enough, but they are readily quantifiable and likely to be recoverable.

If I look at this on an American Cyanamid basis, I am not satisfied that I should grant an interim injunction. Considering the balance of justice and injustice, there are just as good reasons to do so as there are not to do so. Given the history of this litigation and the retrial, the state of affairs today is one entirely influenced by the litigation. To say that the fact that Marks and Spencer are off the market now should be a significant influence on the granting of an interim injunction seems to me to be unfair. To grant an injunction on the basis that Marks and Spencer is not now bidding would perpetuate the consequences of the trial, which since the Court of Appeal has decided it should be retried, would be unfair.

If I look at this on a Cream Holdings v Banerjee basis, and take into account the strength of the cases, I am not convinced this makes much difference compared to the analysis based on American Cyanamid. The free speech right of the defendant exists and so does the claimant have a right to protect its property, but free speech is not the primary issue. This case is concerned with commercial speech and the primary issues relate to alleged infringement of trade marks and fair competition. After all Marks and Spencer is not being prevented from speaking. They can and do publish their advertisement anyway. This case is only about whether they can put it forward in a particular circumstance.

It is not Marks and Spencer's right to free speech which justifies refusing the interim injunction in this case, it is the consideration of the consequences of the two options open to the court. I am not convinced that the irrevocable harm to the claimant if no injunction is granted is sufficient to justify an injunction, having regard to the irrevocable harm to the defendant if I grant one.

I will dismiss the claimant's application. There will need to be a CMC to sort out the next steps to trial.

The first question I have to decide is what to do about the costs. The defendant, who has been the successful party in relation to this application for interim injunction pending trial, submits I should make an order for costs in its favour. The claimant, who was the unsuccessful party, submits that since the decision was based not on the merits of the underlying proceedings but on the balance of convenience, the balance of justice, it would be fair and appropriate (and indeed is done in other cases) to reserve the costs over to trial.

While I see some attraction in Mr. Silverleaf's submission, I find that this is a case in which the normal rule should apply. It is an application which is freestanding and which the party who brought it failed and there is no reason why I should depart from the normal rule, which is that the unsuccessful party should pay the successful party's costs. That is what I will do.

The next matter is about a significant part of the costs which are included in the cost statement. Although they relate in some ways to the application, they are better characterised as being costs relative to the argument about breach of confidence which has been mentioned in my judgment given this afternoon.

I am told that it accounts for approximately 33%, one-third, of the defendant's costs. That seems like a realistic proportion. I accept Mr. Silverleaf's submission that the right way to deal with those is not to make an order for costs relating to them. I am going to do so on the footing that they are better characterised as costs relative to an alleged breach of confidence, than they are as costs of this application. The best time to deal with those costs will be at the further hearing, which is inevitable in this case and is essentially a case management conference to decide how to manage this case over to a retrial. As Mr. Silverleaf rightly says, the judge in that hearing will be faced with the issue of breach of confidence and will have to decide how to deal with it in whatever form it has taken by the time it gets to that judge. Although I have been the judge dealing with the case today and at the previous hearing on the application, nevertheless that judge will also be seized of the breach of confidence issue and they will be well able to deal with those costs.

I gather Mr. Silverleaf and I think Miss Himsworth suggested that it would be a good thing if I was the judge hearing the CMC, I am not going to say anything about that. It is a matter for Chancery Listing and liaison between that office and the parties.

That means the costs I have awarded today are the two-thirds of the actual costs of the application from the defendant's point of view. The defendant's bill comes to a total of £152,185.60. Two-thirds of that is just over £100,000.

Mr. Silverleaf submits that bearing in mind this was essentially listed as a one-day hearing, that is a very substantial sum and he submits that I should, on a summary assessment, reduce it substantially. Miss Himsworth reminds me that this is an important matter for her clients. It involved two significant hearings before the court, one on the 18th or 19th November giving directions over the hearing and this hearing which began yesterday, finished this morning and had judgment this afternoon. While Miss Himsworth is right about that, nevertheless £100,000 for what is an important but comparatively straightforward application for an interim injunction is quite a high sum.

Looking at the rates: those charged by the fee earners at the solicitors' firm do not seem to me to be unreasonable for this sort of intellectual property case and I infer that the explanation for the total is the time spent by various individuals, that is to say counsel and the solicitors. I appreciate I do not have an hourly rate for counsel, but I do not see that counsel are in a different position from the solicitors.

Doing the best I can with this information and bearing in mind that it is necessarily approximate, I will summarily assess the costs starting from a sum of about £100,000 for this application in the sum of £65,000. That is the sum I will award.