Argos Ltd v Argos Systems Inc

The judge dealt with the issue of targeting at paragraphs 144-224 of his judgment. AUL relied heavily on an allegation that it was ASI’s subjective intention to target the AdSense ads on its website at users in the UK. The judge analysed the contemporaneous and other evidence concerning ASI’s intentions at an earlier section of his judgment from paragraphs 87 to 97. The judge held that the documentary evidence showed:

The judge nevertheless pointed out that the documents did not tell the whole story. Although they revealed that part of ASI’s motivation was to earn money from the ads, ASI was also driven by other motivations, including how best to address the unwanted traffic to the website (judgment [95]). Providing a means for “lost” visitors to get back to AUL was one of their concerns and objectives. The judge continued:

“96.

… The fact that ASI realised that this was likely to be achieved if AUL's ads were displayed on ASI's website assists AUL's case in that it supports the conclusion that ASI could foresee and did in fact intend that such displays would occur (although ASI did not have any control over AUL's ads being placed there, and in practice had to leave the presence of AUL's ads to be determined by Google's algorithms). At the same time, it assists ASI's case in that it shows that ASI's motives were far from purely mercenary. On the contrary, ASI had problems it wanted to solve, and it saw a way of doing so that was not inimical to AUL. In fact, ASI's display of ads not only took misguided visitors back to AUL but also generated revenue for AUL, and it seems likely that at least some of this was money that AUL would otherwise not have earned. In this regard, although some users who were trying to navigate to AUL's website and who reached ASI's website by mistake might have found their way back to AUL's website and made purchases from AUL regardless of whether they found AUL's ads displayed on ASI's website, I consider it likely that at least some such users would have given up looking for AUL's products if they had not seen AUL's ads on ASI's website.

97.

On a careful reading of the contemporary documents, it is apparent that the documents themselves reflect these different considerations. For example, one email records the view that the content of the ads is immaterial as long as visitors to ASI's website either click on them or "what's more likely in this case (Argos the Retailer) they simply see ads and figure out that they are in the wrong place". The writer of that email believed that, either way, ASI was likely to generate revenue from the ads because "their browsers have already downloaded the ads and showed on the screen". However, even with that element of focus on the revenue prospects, it was also the writer's perception that the ads would be an effective way of telling visitors who were looking for AUL's website that they were in the wrong place. As a matter of logic, although the emails do not descend to such details, this means of notifying visitors that they were lost would apply most clearly to ads which were not for AUL, as visitors would not expect AUL's website to display ads for others.”

The parties were divided over the relevance of the trader’s subjective intention to targeting and the judge resolved the argument in the following way:

Ultimately, the judge decided that the question was whether the average consumer “will regard ASI’s website (or part of it) as being “for them” … or “aimed and directed at them” … in sufficient numbers to justify banning ASI’s use of the sign ARGOS.”

The judge considered a number of qualitative factors in reaching his decision as to whether ASI’s website, when carrying AdSense ads, would be regarded by the average consumer as targeted at them. These were, in summary:

The judge’s factual conclusions can be summarised as follows:

Accordingly, the judge held that AUL’s case could only be sustained by considering the content of the home page alone, including the ads: judgment [206].

The judge went on to analyse the available evidence as to the appearance of the website, which he described as “scant”. The screenshots in evidence fell into two categories, those produced by ASI and those produced by AUL. The sample screenshots produced by ASI did not support AUL’s case. Of these, the first two advertised products or services were either quoted in US dollars, or used Americanisms, and would not be regarded as targeted at UK consumers. The third was more equivocal, but the judge concluded that the average user would conclude that the ads had been selected to complement ASI’s software, and therefore to be directed at the same customer base as the ASI’s core US business. Those users who were uncertain whether the ads were directed at them would have to enquire further. There was no reliable basis for saying what proportion of UK internet users would fall into that category or what their further enquiries would reveal.

The judge then turned to the screenshots produced by AUL. These had been produced by Mr Keane, who was part of AUL’s legal team. The judge considered that the way in which these screenshots had been obtained did not replicate the behaviour of the average consumer, because Mr Keane had cleared his browsing history before accessing the site, or alternatively had accessed the site after visiting only a few UK retailers’ sites (after having cleared his browsing history). Nevertheless the judge meticulously reviewed each of the screenshots, concluding that there was no page which was clearly directed at the UK by ASI.

The judge summarised his conclusions on the evidence at paragraph 214:

“Pulling all this together, and focussing on ads alone, the advertising content of the sample screenshots which are in evidence varies between (a) that which is entirely and unequivocally not directed at UK consumers, (b) that which is in part not directed at UK consumers and is in part directed at a territory which is unclear but, in the context of the part that is not directed at UK consumers, appears also likely not to be directed at UK consumers, (c) that which is in part not directed at UK consumers and in part directed at UK consumers, and (d) that which is entirely directed at UK consumers (of which the only example which has been produced is Mr Keane's twelfth screenshot, which was displayed to a UK user whose entire history consisted of browsing the websites of AUL, John Lewis, Tesco and Very).”

The judge then factored in the following considerations and conclusions:

A recurring theme in AUL’s case had been the suggestion that it was likely that ads for AUL’s competitors appeared on ASI’s site. Although, at a high level of generality this was a not unreasonable inference, there was no way of estimating to what extent this occurred: judgment [222] and [285]. The judge therefore concluded at [223]:

“For all these reasons, and having regard to the perceptions and expectations of the average consumer, I am unable to hold that the proportion of UK visitors to ASI's website who would have regarded the site or any part of it as aimed or directed at them was such as to warrant the conclusion that it was targeted at them. This result is reached more readily if am wrong in assuming certain matters in favour of AUL (for example, that it is not essential to consider the website as a whole; and that AUL's own ads should be taken into account when assessing the issue of targeting).”

Mr Mellor QC, who presented the case for AUL with Mr Jonathan Hill and Mr Maxwell Keay, opened his submissions with some contentions on the facts which, he said, had not been fully acknowledged by the judge. He submitted that ASI’s subjective intention of targeting the UK had been clear from ASI’s internal emails in 2008 when ASI was deciding whether to become a member of Google’s AdSense programme. Thus, Mr Outinen in Finland had corresponded with John Vigilante, the managing director of ASI, and a person called Jessi who was responsible for the website www.argos.com at ASI. Mr Outinen had made it clear that he, and therefore ASI, knew exactly what was happening with the misdirection of internet traffic intended for AUL, and that this represented an opportunity for ASI to make money. The emails showed that ASI was aware that in October 2008 10,000 different visitors a day were visiting www.argos.com. Of these only 105 a week were looking for ASI. ASI appreciated that the content of the ads would not necessarily be related to its business. All that was important was that visitors to the website either clicked on the ads (earning per click revenue from Google) or simply saw the ads, because at that point their browsers would already have downloaded the ads (earning per impression revenue). Mr Outinen pressed for ASI to subscribe to AdSense in the following terms:

“I still think you may not have understood the idea of putting these Google Ads. 99% of visitors of argos.com are people NOT interested into getting CAD software or potential customers in any way in the future. They are simply there because they typed something in their browsers. 10000 different visitors a day! And simply providing ads for them to VIEW is enough to make money (they don’t even need to click on the ads!) Is this money worth the hassle who knows, but without trying how can you know??”

In the light of these exchanges Jessi went ahead and registered the account with Google, and it was approved on December 12 2008. The account immediately started earning money for ASI: $US 300 in the remainder of December 2008. By August 2009 the account was earning approximately $45 per day and ASI anticipated $1300 dollars for the month. Over Christmas 2008/09 the website was receiving approximately 20,000 impressions per day. Mr Mellor submitted, therefore, that so far as subjective intention was concerned, ASI’s plan to make money out of the misdirected traffic arriving at their website was clear.

AUL accepts that there were different views within ASI about Google AdSense. Mr Outinen was keen to take steps to maximise revenues, whereas a Mr Fox, whose direct responsibility the website was, was less keen on the ads appearing on the website, which he thought looked bad. In November and December 2010 and again in those pre-Christmas months in 2011, ASI had made the page more ad-friendly, when they expected the traffic from the UK to increase.

Against this background, Mr Mellor submitted that the decision in January 2012 to split the home page so as to have one home page for the Americas (without ads) and one for the rest of the world (with ads) reinforced its decision to target the 10,000 UK internet users who accessed ASI’s website by mistake in the belief that they would find AUL.

In December 2013 Mr Outinen expressed himself as satisfied with the income from Google AdSense, which was running at $30,000 per year at that point. He described this as “money for nothing”. Accordingly Mr Mellor submitted that the subjective intention to target the 10,000 UK visitors to the website and earn money out of them through AdSense was realised.

Mr Mellor also relied on the fact that the Google algorithm is designed to provide ads which will be of interest to the individual internet user, based on his or her browsing history. Google’s charges for the AdSense service, as the judge had found, depended on the level of interest that was generated by the advertisements, so Google plainly had an incentive to provide relevant advertisements. The more relevant the advertisements to the surfer, the more likely that click-through income would be generated for ASI.

Mr Mellor went on to submit that the judge had placed undue reliance on the screenshots produced by ASI, as it was not clear that these had been downloaded from the UK. Moreover, the judge had been unduly critical of AUL’s evidence of what the consumer would see if accessing the site from the UK. What the consumer would see was dependent on a variety of factors which made it difficult to present representative evidence. The judge had wrongly declined to draw the inference that a UK internet user would be likely to be served with ads of relevance to that user, when that was plainly ASI’s intention, and when that was exactly how the Google AdSense algorithm would be likely to work, and when both Google and ASI would plainly benefit if the ads caught the interest of the consumer.

The judge had also failed to take account of further evidence that different ads were appearing on ASI’s website depending on the location of the user. This was based on an email from Mr Outinen in January 2012 in which he said that he had checked with a friend who lived in London that he was really given different ads, and he confirmed, so far as Mr Outinen could recollect at the time, that his friend had been fed with more than one AUL ad.

Based on these submissions on the facts, Mr Mellor submitted that there were two transactions or events on which it was necessary to focus, the “impression event” which brought the user to the ASI website and the “click-through event”, if it occurred, when the user clicked on an ad. Both events were extremely quick, and the role for the average consumer in deciding whether these events were targeted at the UK was limited if it had a role at all. In fact it was only necessary to ask whether it was intended to target the 10,000 users a day who visited ASI’s website by mistake. Put in that way, the question answered itself.

Thus, in Case C-173/11 Football Dataco Ltd v Sportradar GmbH [2013] FSR 4, in the context of re-utilisation of data from a database protected by the sui generis right under Directive 96/9/EC (“the Database Directive”), the Court of Justice had approached the issue of the localisation of the act on the basis of targeting, but had regard to matters other than those which are visible to the average consumer: see paragraphs 39-43 of the court’s judgment.

The approach of asking, through the eyes of the average consumer, whether the website or webpage was “for me” was not universally applicable, and the judge had erred in relying upon it exclusively in the present case. This error led the judge to ask the wrong question, because he was not considering the relevant events or transactions. Moreover in his application of the viewpoint of the average consumer the judge had indulged in too detailed an analysis, asking for example whether the consumer would think that the ads were targeted by Google or ASI. Finally, the judge had declined to take account of any factor which would not have been apparent to the average consumer.

Mr Mellor submitted at one point that, viewed in this way, it did not matter what ads were shown. ASI’s intention was to make money from the 10,000 visitors and they would do so whether or not the visitors clicked through on ads which interested them.

Mr Howe QC, who appeared for ASI with Mr Jaani Riordan, submitted that it was wrong to place so much reliance on Mr Outinen’s emails. Mr Outinen was not an employee of ASI at the relevant time. He was a director only of Vertex Systems in Finland, where his role was Senior Software Consultant. He had little direct role in the business. Since 2008 he had been Company Secretary of ASI. He could not therefore be taken to be representative of the corporate mind of ASI. Although he had recommended putting ads on the website, he had not taken the decision to do so. The judge did properly consider Mr Outinen’s emails, in the context of other evidence, and was entitled to make the carefully nuanced findings he did as to ASI’s motivations in using the AdSense ads.

The judge had made some important findings relevant to ASI’s intention and motivation which were not open to challenge on appeal:

Mr Howe submitted that it was not open to AUL to object to the evidence of the screenshots produced by ASI, given that it had not challenged Mr Moilanen about how they had been produced. The burden lay on AUL to show that the website would be regarded as targeted at the UK consumer, and if it was being suggested that these screenshots were not representative of what a UK consumer would see, that should have been investigated at the trial.

Mr Howe further submitted that it was not open to AUL to run a case based exclusively or predominantly on subjective intention when AUL’s case before the deputy judge had been that subjective intention has the subsidiary role which the judge ultimately found it to have, namely as an ingredient in the court’s overall evaluation of whether, viewed objectively through the eyes of the average consumer, the relevant activity was targeted by ASI at the United Kingdom. Had the case been run below as AUL now wished to run it, the course of the trial might well have been different.

Mr Howe then focused attention on the use of the sign ARGOS of which complaint is made. This was the use of the sign in relation to a service involving the provision of advertising space. The customers for such a service were either Google itself or its customers. In signing up to Google AdSense, ASI (in the US) would have used the sign ARGOS. That, however, was not a use which was said by AUL to be a use in the course of trade in the UK. If ASI was providing this service it was providing it to customers other than the users of the ASI website, and was not doing so through the website.

Mr Howe accepted that a sign can nonetheless be used in relation to people other than customers or prospective customers. So for example an employment agency’s customers are the employers. Nevertheless an employment agency’s mark is used in dealings with appointees, and the mark will be of use in attracting good potential appointees. However, it was difficult to see how the mark ARGOS was being used by ASI towards consumers in the UK in relation to the service of provision of advertising space. The consumers would see an obviously foreign website concerned with computer software which was carrying Google AdSense ads.

AUL’s case was dependent on the content of the ads being seen as relevant to UK consumers, and therefore targeted at them. The CJEU had distinguished between providing a channel for commercial communication by others and providing the commercial communication oneself: see Joined Cases C-236/08 to C-238/08 Google France [2011] Bus LR 1, at [55] to [57]; and Case C-324/09 L’Oreal v eBay [2011] RPC 27 at [99]-[102]. ASI were likewise merely providing a channel for the communication of the AdSense ads. If Google were selecting ads by reference to the browsing histories of users in particular countries, it was Google who were targeting those countries and users and not ASI. ASI was not sued for any form of accessory liability with Google. These considerations had not been factored into AUL’s case on targeting.

Targeting is not an independent doctrine of trade mark law. It is, in essence, a jurisdictional requirement. Because trade marks are territorial in effect, those who are doing business exclusively outside the United Kingdom should not have their dealings subjected to the trade mark law of the United Kingdom. Failure to recognise this principle is a failure to give effect to the territoriality of the underlying rights. Moreover the fact that a website is accessible from anywhere in the world, and therefore may attract occasional interest from consumers there when this is not intended, should not give rise to any form of liability. Thus, in order to make good its claim of trade mark infringement, it was necessary for AUL to establish that ASI was using the sign ARGOS in the course of trade in relation to goods or services in the United Kingdom.

Since the deputy judge gave his judgment, this court has had the opportunity to consider the law on targeting in Merck v Merck Sharpe & Dohme [2017] EWCA Civ 1834. Kitchin LJ, who gave the leading judgment (with which Patten LJ and I agreed) conducted a review of the EU and domestic authorities, including the deputy judge’s decision in the present case and the foundational Joined Cases C-585/08 and C-144/09 Pammer v Reederei Karl Schluter GmbH & Co. KG and Hotel Alpenhof GmbH v Heller. As to the relevance of subjective intention, Kitchin LJ said this at [165]:

“One of the issues which arose for consideration in Argos was the relevance of the subjective intention of an operator of a website in one territory in assessing whether its internet activity is targeted at the consumers in another territory, in particular the UK. The deputy judge held and I agree that if, viewed objectively from the perspective of the average consumer, a foreign trader’s internet activity is targeted at consumers in the UK, the fact that, viewed subjectively, the trader did not intend this result will not prevent the impugned use from occurring in the UK. But that is not to say that the actual intention of the website operator is irrelevant. If the foreign trader does intend to target its internet activity at consumers in the UK then it seems to me that this is a matter which the court may properly take into account. After all, a trader may be expected to have some understanding of the market it intends to penetrate and it may not be difficult to infer that this intention has been or is likely to be effective (see, by analogy, Slazenger v Feltham (1886) 6 R.P.C. 531 at p.536, per Lindley LJ).

Kitchin LJ then went on to summarise the effect of the authorities, when considered by reference to an advertisement for goods offered for sale on a website, in the following way:

“167.

First, in determining whether an advertisement of goods bearing a trade mark on the website of a foreign trader constitutes use of the trade mark in the UK, it is necessary to assess whether the advertisement is targeted at consumers in the UK and in that way constitutes use of the mark in relation to goods in the course of trade in the UK.

168.

Secondly, the mere fact that a website is accessible from the UK is not a sufficient basis for concluding that an advertisement displayed there is targeted at consumers in the UK.

169.

Thirdly, the issue of targeting is to be considered objectively from the perspective of average consumers in the UK. The question is whether those average consumers would consider that the advertisement is targeted at them. Conversely, however, evidence that a trader does in fact intend to target consumers in the UK may be relevant in assessing whether its advertisement has that effect.

170.

Fourthly, the court must carry out an evaluation of all the relevant circumstances. These may include any clear expressions of an intention to solicit custom in the UK by, for example, in the case of a website promoting trade-marked products, including the UK in a list or map of the geographic areas to which the trader is willing to dispatch its products. But a finding that an advertisement is directed at consumers in the UK does not depend upon there being any such clear evidence. The court may decide that an advertisement is directed at the UK in light of some of the non-exhaustive list of matters referred to by the Court of Justice in Pammer at [93]. Obviously the appearance and content of the website will be of particular significance, including whether it is possible to buy goods or services from it. However, the relevant circumstances may extend beyond the website itself and include, for example, the nature and size of the trader’s business, the characteristics of the goods or services in issue and the number of visits made to the website by consumers in the UK.”

These passages make it clear that evidence of subjective intention is a relevant, and possibly (where the objective position is unclear or finely balanced) a determinative consideration in deciding whether the trader’s activities, viewed objectively from the perspective of the average consumer, are targeted at the UK. Subjective intention cannot, however, make a website or page (or part of a page) which is plainly, when objectively considered, not intended for the UK, into a page which is so intended.

It is important to note that the summary of principles in Merck relates to the example of an advertisement for goods, where the role of the average consumer will be to determine whether the advertisement is targeted at him or her. In each case it will be necessary to look at the acts which are asserted to be use of the trade mark, and to focus on whether those acts are targeted at the United Kingdom. The scope of the enquiry will vary from case to case, as will the factors which are relevant to its determination. To that extent, I am prepared to accept that the role of the average consumer on the issue of targeting may differ from case to case.

It is not correct in my judgment to split the activity in question into the impression event and the click-through event. The acts in question concern the use of the sign in relation to the provision of advertising space. Before the internet user arrives at the site, he or she is unaware of any advertisement. There is only one event, if one must call it that, and that is when the consumer sees the landing page on the website. The answer to whether that page is targeted at the UK depends on the consumer’s reaction to that page.

I also do not think it is of assistance simply to ask whether it was intended to target the 10,000 consumers a day who visited ASI’s website. That question conflates subjective and objective intention, fails to focus on the relevant acts and excludes the average consumer altogether from the exercise.

In the present case the question was whether the average consumer would regard the service of the provision of advertising space on ASI’s website as targeted at UK consumers. The judge appears, at some points in his judgment, to have gone further and used the construct of the average consumer to determine the identity of the person who is targeting the activity at the UK. As I shall explain I do not think that is an appropriate use of the average consumer. If the relevant acts would be taken to be targeted at the UK consumer, the identity of the person who is seeking to target the UK is a question of fact to be determined from an evaluation of all the evidence.

Because AUL’s case is framed under Article 9(1)(c), it can be based on use of the sign in relation to any goods or services, whether or not similar to the goods or services for which the mark is registered. AUL contends that the domain name is used in relation to a service of the provision of advertising space. In argument this was likened to the electronic equivalent of the provision of a physical billboard or advertising hoarding. The relevant enquiry on the issue of targeting is therefore whether ASI’s use of the sign ARGOS in relation to its electronic billboard service is a use in the course of trade in the UK.

The question raised in this case is thus different in material respects from that raised in conventional online retailing cases, where the question for the UK average consumer is along the lines “is this website operator offering goods or services under the sign, in the course of trade, which are intended for me in the UK?” If one were to seek to encapsulate the question in the present case it would be “is this website operator providing an electronic billboard service under the sign, in the course of trade, which is intended for me in the UK?”

Absent advertisements with content relevant to UK consumers, the answer to the question thus asked is clearly “no”. The average consumer would see that a US based website owner, who does not target its core services at the UK has given up space on its web page to allow for the insertion of advertisements which are not aimed at the UK. The average consumer would conclude that the billboard service was not for him (or her). Although such consumers would conclude that ASI was operating a billboard service, it was one which was not being used in the course of trade in, or intended in any way to face the UK. The physical equivalent of this billboard is one alongside an Interstate Highway deep in the United States, displaying an advertisement for a local baseball game with ticket prices in dollars and cents.

The conclusion in the example in the previous paragraph is not altered because ASI knows that a large number of UK based consumers will mistakenly arrive at the website, immediately realise that they are in the wrong place and go away again, even if they do so by clicking on ads. It is also not altered by the fact that ASI will earn the per-impression and click-through income from these visits. No part of the web page I have taken as an example is objectively targeted at the UK, and no amount of knowledge or subjective intention on the part of ASI can change that objective effect.

By contrast, once the billboard is populated with ads of relevance to UK customers, there can be no doubt that the reaction of the UK consumer to the question I have posed will be different. It is true that the webpage as a whole presents conflicting messages by advertising ASI’s core business as well as presenting ads, but the billboard part of the webpage would be understood by the consumer as directed at them to the extent that it carried ads of obvious interest to UK consumers. The judge was inclined to accept this, at least in respect of some of the ads, and I agree. In those circumstances it is difficult to escape the conclusion that the average consumers who looked at the contents of the billboard when they include ads of interest to them would conclude that someone was targeting the billboard service at them.

I agree with Mr Mellor that it is wrong to ask whether the whole website, or even the individual page, is “for me” in the circumstances of this case. That may well be the right question when one is asking whether goods or services offered for sale from a website are offered to consumers in the UK. The issue here is whether there is use of a sign in relation to the service of the provision of advertising space, and so it is necessary to focus on the ads and ask whether a consumer would consider the ads as directed at him or her. Where the ads feature UK businesses the answer will usually be clear.

The question posed by Mr Howe’s argument is whether the targeter is Google (and its advertisers) or ASI. The judge did not make any finding that the average consumer would be able to come to a conclusion about who was selecting the ads for display on the website. It would all depend on context and expectation, and the consumer would think that some were selected or at least influenced by ASI. Mr Mellor submitted, and I agree, that there is an unreality in ascribing to the average consumer a conscious deliberation about who is targeting the ads, when the two events on which he relies occur very quickly, and necessarily without much thought. But quite apart from this, the purpose of the targeting enquiry is to determine whether the act complained of is a use of the sign in the course of trade in the United Kingdom in relation to goods or services. Once there is an affirmative answer to that question, the question of the responsibility for that act is a question of fact determined from the whole of the evidence, not just through the eyes of the average consumer. I think that this is a case, therefore, where it was necessary to conduct a fuller evaluation of the relevant circumstances than the judge performed in order to determine whether the billboard service was in fact directed by ASI at consumers in the UK.

There was abundant material on which to find that ASI was aware of the misdirected traffic arriving at its website seeking AUL not ASI. ASI knew also that the vast majority of this traffic emanated from the UK and Ireland. Equally ASI knew that the inherent probabilities were that that these visitors would be shown ads of interest to them, because this is how Google’s algorithms were likely to work so as to maximise income to itself and to its partners, and the visitors were shown such ads, in particular ads for AUL. The judge held at [97] that ASI could foresee and did in fact intend that such displays of AUL ads to UK users would occur.

It is true that Google ultimately has control over what ads appear on ASI’s billboard. The right conclusion from that is that Google had a role in causing the ads to be targeted at the UK. Nevertheless, the consequence of the selection of those ads was that they appeared on ASI’s billboard, which thus became targeted at UK customers. I think the right conclusion is that both Google and ASI were targeting the ads at the UK, using ASI’s billboard on their website argos.com for that purpose. I therefore reject Mr Howe’s argument that it was Google and not ASI who was targeting the UK.

The present case is not analogous to Google France, where the reference service provider permitted advertisers to use keywords which embodied the signs complained of, which were not subsequently displayed on the website advertisements. Google were not using the trade mark in its own commercial communications because no consumers saw the sign which they stored in their systems. L’Oreal v eBay is also not analogous to the present case. There, the offending sign was used in the advertisements placed by sellers on the eBay online marketplace. eBay may have facilitated this by providing the service which enabled this to take place, but they were not using the offending signs in their own commercial communications.

By contrast to those cases, ASI is using the sign in a commercial communication, not least because the domain name argos.com is used to direct users to its website where they can see its billboard.

How did the judge reach a different conclusion? Firstly, it seems to me that he was very much influenced by the fact that the overwhelming majority of consumers who arrived at the site arrived there by virtue of their own mistake. I do not think that the reason why internet users arrive at a site is of relevance to the issue of targeting in this case. Whatever the reason, substantial numbers of visitors were arriving at the site and were exposed to the ads. Once they are at the ASI website, there is no reason for them to suppose that ads of interest to UK consumers were not intended for UK consumers.

Secondly, the judge gave careful consideration to the question of whom the average consumer would think was responsible for the ads. He considered that if the ads complemented the trader’s business, then the user would be likely to conclude that they had been selected by a combination of the advertiser, the trader and any algorithm involved in selecting those ads for display, on the basis that they would be of interest to the US customer base (judgment at [210(5)]). If the ads were for products unrelated to the business of the website operator, then the consumer who got as far as considering them would assume that they were not chosen by the website operator, but by a combination of the advertiser and the Google algorithm (judgment at [212(4)]).

As I have explained, I think this analysis of who the user would think is responsible for the ads is wrong in principle as well as being somewhat unreal. Once it is clear that the ads, and in consequence the billboard service, are targeted at the UK, the only further question is who is using the sign in relation to the billboard. The identity of the person using the sign is a question of fact to be determined in the light of all the evidence.

In any event, the judge did not conclude that the average consumer would infer that the responsibility for the ads lay exclusively with Google. It all depended on context and expectation, and the consumer would think that at least some of the ads of interest to UK users were selected by ASI.

The judge’s overall conclusion (judgment [223]) was that the proportion of UK visitors to ASI's website who would have regarded the site or any part of it as aimed or directed at them was not such as to warrant the conclusion that it was targeted at them. I think the judge erred here as well. There were plainly some ads which visitors would regard as not directed at the UK, such as ads for US businesses. But ads for UK businesses were plainly aimed or directed at UK visitors, and it is difficult to see how the average consumer could sensibly decide otherwise.

I think the judge’s analysis of the issue of targeting did become, with respect, over-complex. Whatever may have been the circumstances surrounding the screenshots produced by AUL, it was plain that the appearance of ads of relevance to UK consumers was neither de minimis nor accidental. Google would not have been doing their job properly if ads with potential for click-through did not appear, and that they did appear was both established and entirely to be expected.

Standing back, it is clear that ASI was providing a billboard service which included ads of interest to UK consumers. To the extent that it contained such ads it was targeted at the UK, but not otherwise. In circumstances where those ads appeared, as they undoubtedly would, ASI was targeting UK users of its billboard although Google and its advertisers were targeting ads at the UK as well. ASI was using the sign ARGOS in relation to its billboard when it performed those targeted activities, not least because it was accessed via the website argos.com. I therefore respectfully disagree with the judge on the issue of targeting.

The judge dealt with the requirement that the use complained of must give rise to a link between the sign and the trade mark at paragraphs 265 to 270 of his judgment. He concluded at [270]:

“On the evidence, it seems to me that it is not any use by ASI of the sign ARGOS which creates a link in the mind of the average consumer between that sign and the 263 Mark, but, rather, that the connection undoubtedly made in the minds of many UK consumers between ASI's domain name and AUL arises from pure supposition. Logically, in light of the contents of ASI's website, and having regard to the fleeting if not vestigial duration of the overwhelming majority of UK visits to that website, any such supposition will be dispelled by a single visit to the website.”

Mr Mellor submitted that the requirement for a link between the use of the sign and the mark was made out in the present case. When the consumer mistakenly types the domain name argos.com into a web browser, he or she is automatically served the website containing the advertisements. Whilst in some cases of Article 9(1)(c) infringement it was necessary to show that the use of the sign brought the mark to mind, this was simply to establish causation, in the sense that it was the use of the sign which caused the advantage to be taken. What was needed to establish causation varied from case to case. Here causation was automatic and mechanical.

The requirement for a link arose from Case C-252/07 Intel Corporation Inc v CPM United Kingdom Ltd [2009] RPC 15 where the mark and sign were INTEL and INTELMARK and the goods and services for which Intel had a reputation (computers and computer-linked goods) were quite different to those marketed by CPM under the sign INTELMARK (marketing and telemarketing services). The requirement that the mark be “called to mind” was necessary on the facts of that case, to establish the necessary causation. In the Interflora case, Marks & Spencer bid for the keyword INTERFLORA so that its ads would appear on the same page as the search results. The average consumer never sees the keyword, and the mark cannot therefore have been called to mind by the relevant use, yet there was held to be a link. Article 9(1)(c) infringement only failed in that case because the use was with due cause.

Here the use of the domain name (which occurred because of AUL’s reputation) caused the web page to be served to the user and thus caused the use of the sign in the UK, which enabled ASI to make money. If ASI did not use that domain name it would not receive the landing page and would not be able to target UK consumers. That was sufficient to establish the link.

Mr Howe submitted that a mechanical link of the kind contended for by AUL was not sufficient to establish Article 9(1)(c) infringement. The paradigm or core example of the application of that provision was where there was a transfer of the image of the mark from the proprietor’s goods or services to the infringer’s goods or services. He relied on the statement of the CJEU in Case C-487/07 L’Oreal v Bellure [2010] Bus LR 303 (a case about comparisons made between perfume brands) at [41]:

“As regards the concept of “tak[ing] unfair advantage of . . . the distinctive character or the repute of the trade mark”, also referred to as “parasitism” or “free-riding”, that concept relates not to the detriment caused to the mark but to the advantage taken by the third party as a result of the use of the identical or similar sign. It covers, in particular, cases where, by reason of a transfer of the image of the mark or of the characteristics which it projects to the goods identified by the identical or similar sign, there is clear exploitation on the coat-tails of the mark with a reputation.”

Although Mr Howe accepted that it does not necessarily follow from this passage, or the court’s more general statement at [50], that Article 9(1)(c) is limited to cases involving a transfer of image, he submitted that the courts should be very cautious in extending the concept of a link to cases where no such transfer takes place.

The Interflora case showed that the relevant link must be created in the mind of the consumer. The distinction made in that case between advertisements served up by use of a keyword which offered imitations of the trade mark proprietor’s goods or services and advertisements which were merely offering alternatives to those good or services showed that it was a link in the mind of the average consumer which was relevant. To say that a mere mechanical link was enough would create a type of advantage which is different in kind from any considered in the case law.

In the present case the visitor to ASI’s website was no doubt thinking of AUL when it entered argos.com in his or her browser, but at that stage he or she had not been exposed to any ads. Any link with AUL was, on the judge’s findings, immediately dispelled on arrival at the site. It followed that at no stage was there a relevant link between the use of the sign (in relation to the ads) and the mark.

The requirement for a link between the sign and the mark with a reputation has been analysed in two well-known cases before the CJEU: Case C-408/01 Adidas-Salomon AG v Fitness World Trading Ltd [2004] Ch 120; and Case C-252/07 Intel Corp Inc v CPM United Kingdom Ltd [2009] RPC 15. It is sufficient for such a link that the sign will call the trade mark to the mind of the average consumer.

It does not follow from the CJEU’s jurisprudence that it is always necessary to establish that the use of the sign complained of will bring the trade mark to mind for the average consumer. That is one of the ways, but not the only way, in which it is possible to establish the necessary link. Where the taking of unfair advantage depends on a comparison of mark and sign in the mind of the average consumer, then it is obvious that the necessary link cannot be established without the use of the sign bringing the mark to mind: see Adidas at [49]. That is because it is by bringing the mark to mind that it is possible for the unfair advantage to be taken. It does not follow, however, that it will be necessary to show a “bringing to mind” in all cases. In Interflora there can be no doubt that the use of the keyword, which is not visible to the consumer, took advantage of the repute of the trade mark by directing ads for a competing product at an internet user searching for the trade mark owner’s product. As the court explained:

“86 In those circumstances, as the Advocate General observes at para. 96 of his Opinion, it cannot be denied that, where a competitor of the proprietor of a trade mark with a reputation selects that trade mark as a keyword in an internet referencing service, the purpose of that use is to take advantage of the distinctive character and repute of the trade mark. In fact, that selection is liable to create a situation in which the probably large number of consumers using that keyword to carry out an internet search for goods or services covered by the trade mark with a reputation will see that competitor’s advertisement displayed on their screens.

87 Nor can it be denied that, when internet users, having studied the competitor’s advertisement, purchase the product or service offered by the competitor instead of that of the proprietor of the trade mark to which their search originally related, that competitor derives a real advantage from the distinctive character and repute of the trade mark.

88 Furthermore, it is not disputed that, in the context of a referencing service, an advertiser which selects signs identical with or similar to the trade marks of other persons does not, as a general rule, pay the proprietors of the trade marks any compensation in respect of that use.

89 It is clear from those particular aspects of the selection as internet keywords of signs corresponding to trade marks with a reputation which belong to other persons that such a selection can, in the absence of any “due cause” as referred to in art.5(2) of Directive 89/104 and art.9(1)(c) of Regulation 40/94, be construed as a use whereby the advertiser rides on the coat-tails of a trade mark with a reputation in order to benefit from its power of attraction, its reputation and its prestige, and to exploit, without paying any financial compensation and without being required to make efforts of its own in that regard, the marketing effort expended by the proprietor of that mark in order to create and maintain the image of that mark. If that is the case, the advantage thus obtained by the third party must be considered to be unfair (L’Oréal [2009] E.T.M.R. 55 at [49]).

90 As the Court has already stated, that is particularly likely to be the conclusion in cases in which internet advertisers offer for sale, by means of the selection of keywords corresponding to trade marks with a reputation, goods which are imitations of the goods of the proprietor of those marks (Google France [2010] E.T.M.R. 30 at [102] and [103]).

91 By contrast, where the advertisement displayed on the internet on the basis of a keyword corresponding to a trade mark with a reputation puts forward—without offering a mere imitation of the goods or services of the proprietor of that trade mark, without causing dilution or tarnishment and without, moreover, adversely affecting the functions of the trade mark concerned—an alternative to the goods or services of the proprietor of the trade mark with a reputation, it must be concluded that such use falls, as a rule, within the ambit of fair competition in the sector for the goods or services concerned and is thus not without “due cause” for the purposes of art.5(2) of Directive 89/104 and art.9(1)(c) of Regulation 40/94.”

In the case of keyword advertising, the consumer already has the trade mark in mind at the moment the search is commenced. The choice by the advertiser of the keyword ensures that the advertiser’s advertisement appears on the screen as well as that of the trade mark proprietor. The court was clearly of the view that the use of the keyword took advantage of the distinctive character and repute of the trade mark. That was so notwithstanding the fact that it was not the keyword which brought the mark to the mind of the user. The activity as a whole escaped a finding of infringement only because the taking of advantage was with due cause where the consumer was offered a fair choice between normal competing alternatives. I reject Mr Howe’s argument that the conclusions of the CJEU in relation to due cause and the need for the advertisements to be for “imitation goods” before there will be unfair advantage have anything to do with the requirement for a link.

In this case it is the use of the sign ARGOS in relation to an electronic billboard service which must take unfair advantage of the distinctive character or repute of the mark. The internet traffic which arrives at ASI’s website on the strength of AUL’s reputation already has AUL’s reputation in the mark ARGOS in mind. Although the visitors immediately realise that they are in the wrong place, they are also immediately confronted by ASI’s billboard service. At this point they either leave the site altogether or click on an ad in order to leave. Nevertheless, by presenting the billboard service to the internet traffic which has arrived at the website on the strength of AUL’s reputation, ASI gain at least the impression fee earned by the downloading of the ads. Advantage is thereby taken of an opportunity which arises only because the internet traffic has arrived at the site on the strength of AUL’s reputation. For my part, I would regard that as sufficiently establishing the necessary link.

The judge’s conclusion that the relevant link was not formed was based on his conclusion that it was mere supposition which brought the traffic to ASI’s website. I do not think that is a reason why a relevant link is not present. The impression formed in the more traditional case where the mark is brought to mind might also be described as a supposition, but a link is nevertheless formed. It is true that the supposition is formed at a stage before the user is exposed to the billboard, but I do not see why that prevents a link being present at the point when the billboard service is so exposed.

Equally, I do not agree with the judge that the link is broken because visitors to the website immediately realise that they are in the wrong place. AUL’s case is that advantage is taken of the repute as soon as the ads are downloaded, because at that point the per-impression fee is earned. Advantage continues to be taken when the user clicks on the ads, even if, by that stage, they realise they are in the wrong place.

Contrary to Mr Howe’s submissions, I do not think that this conclusion represents a departure from the mainstream of the CJEU jurisprudence. The language of Article 9(1)(c) is general, and is not limited expressly or by implication to cases involving a transfer of image from the reputation in the mark to the infringer’s goods or services. It would be inconsistent with that general language to create a closed list of the ways in which advantage can be taken of the reputation of a mark, by excluding the type of mechanical link which is relied on in the present case.

It follows that, in my judgment, the absence of a link was not a correct basis on which to dismiss the claim.

The judge dealt with unfair advantage at [279] to [294] of his judgment. In summary his conclusions were:

The judge’s final conclusion was expressed at [294] as follows:

“In sum, I agree with [counsel for ASI] that the advantage that ASI gained from the ads was not significant in the context of the business of either ASI or AUL, was not without an element of benefit to AUL, ought to be seen in the context that the traffic which enabled it to be gained was not sought out by ASI and was not without some unwanted adverse effects for ASI, was not accompanied by any transfer of reputation or brand characteristics to ASI's goods or services (although I accept this is not a determinative consideration), and, overall, was not, in my view, unfair.”

Mr Mellor submitted, on the basis of statements of the CJEU in L’Oreal v Bellure, that it was enough for AUL to show that ASI had gained an advantage without paying compensation to AUL. ASI’s use was a plain case of free-riding on the coat tails of AUL’s reputation. This followed from the fact that ASI was making money out of AUL’s reputation.

Mr Mellor also relied on the fact that ASI could simply have used a landing page which re-directed the incoming traffic, so that users wanting AUL would be directed to AUL’s website and users wanting ASI would be directed into subsequent pages of ASI’s American website. Ads could have been shown on ASI’s American website, although of course these would not have generated nearly as much revenue. Instead ASI chose to include advertisements on the UK-facing site, when they had no need to do so.

Mr Mellor also suggested that there was some diversion of trade from AUL, and ASI had been reckless as to the types of ad which appeared. He went on to submit that it was not necessary to show a change in economic behaviour by the consumers in question. However if that was a condition of liability, it was satisfied here.

Next, Mr Mellor submitted that it was no answer to say that the ads provided a means for misdirected customers to find their way back to AUL. If this was ASI’s objective it could have been done by a link. In any event the bounce rates did not change significantly when the ads were removed, which showed that ASI did not need the ads to re-direct the users.

Mr Howe first drew attention to certain findings of the judge. The judge had found:

Mr Howe submitted that, to be capable of infringing under Article 9(1)(c), the use of the sign ARGOS by ASI must cause consumers to change their economic behaviour. His submissions depended on the fact that consumers immediately appreciate when they arrive at ASI’s website that the website is not for them. Accordingly it is not the use of the sign which causes them to do anything further at this point, and so does not change economic behaviour.

Next, Mr Howe drew attention to the judgment at [60] where the judge had made reference to the benefit or advantage which AUL had itself derived from the fact that their advertisements were placed on argos.com. A review of the performance for argos.com for the three months to July 2013 said that it had been “performing well, and has therefore not been excluded during our optimisation”. The review had found that in the period there had been 265,521 impressions, 19,763 clicks giving a click-through rate of 7.44% yielding 1,631 sales with a revenue of £106,214. It was calculated that for a spend of £300 a month a revenue of £34,000 a month was being generated. An email report of the review said:

“You would probably argue that it should be there – at least it stops people having to re-search and find an Argos click through. Conversely turning it off would not have a massive impact.”

Although there was a benefit to ASI in terms of revenue, there was also a discernible benefit to AUL. In the absence of the ads, most visitors would find their way back to AUL’s website, but not all would do so.

Mr Howe accepted that a simple link, without any AdSense ads, was one way of re-directing the lost traffic back to AUL’s website. Such a link would not have been a reasonable alternative, however. ASI was entitled to present to its US clients, wherever they were and including when they were travelling in the EU, an attractive website of ASI’s choice. A front page which said “we are not AUL” would not be acceptable.

Mr Howe submitted that the judge had conducted an entirely correct multi-factorial assessment. He had taken proper account of the problems which ASI had faced when the traffic overfilled its log files in 2004, and used up to 75% of the available bandwidth, which it was still experiencing as late as 2008. As a result, this court should not interfere with his evaluation that the use of the sign by ASI was not unfair.

Although categorised (in a manner which may seem odd to English lawyers) as a form of “injury”, the taking of unfair advantage of the distinctive character or the repute of a mark, within Article 9(1)(c), does not require that there be a likelihood of confusion or a likelihood of detriment to the distinctive character or the repute of the mark or, more generally, to its proprietor. The CJEU in L’Oreal v Bellure at [50] described the relevant advantage in the following way:

"… the taking of unfair advantage of the distinctive character or the repute of a mark, within the meaning of that provision, does not require that there be a likelihood of confusion or a likelihood of detriment to the distinctive character or the repute of the mark or, more generally, to its proprietor. The advantage arising from the use by a third party of a sign similar to a mark with a reputation is an advantage taken unfairly by that third party of the distinctive character or the repute of the mark where that party seeks by that use to ride on the coat-tails of the mark with a reputation in order to benefit from the power of attraction, the reputation and the prestige of that mark and to exploit, without paying any financial compensation, the marketing effort expended by the proprietor of the mark in order to create and maintain the mark’s image."

On a literal reading of that paragraph, any advantage taken by a third party of a mark with a reputation will be unfair if the third party seeks by its use of the similar sign (a) to ride on the coat-tails of the mark in order to benefit from its attraction, reputation and prestige, and (b) to exploit, without paying any financial compensation, the marketing effort expended by the proprietor of the mark in order to create and maintain the mark’s image. Such a general statement is comprehensible in the context of a case where there is, or is likely to be, a transfer of the image of the proprietor’s reputation to the defendant’s goods or services. However the present case does not involve any such transfer of image and it is in the highest degree improbable that the court had in mind a case on these highly unusual facts.

Subsequent to L’Oreal, this court decided in Whirlpool Corp v Kenwood Ltd [2009] EWCA Civ 753; [2010] RPC 2 that mere commercial advantage was not sufficient to render the taking of advantage unfair. Lloyd LJ (with whom Wilson and Rix LJJ agreed) stated at [136]

“There must be an added factor of some kind for that advantage to be categorised as unfair.”

In Specsavers International Healthcare v Asda Stores Limited [2012] EWCA Civ 24; [2012] F.S.R. 19, Kitchin LJ, after citing L’Oreal and Whirlpool reiterated that there were limits to the broad statement of principle in L’Oreal:

“128.

But plainly there are limits to this broad principle. For example, as stated, it would apply to comparative advertisements which comply with all the conditions set out in the Directive 2006/114/EC concerning misleading and comparative advertising (the "Comparative Advertising Directive"). Yet it is clear that such advertisements are permissible. So also the Court has recently given guidance on the application of Article 9 of the Regulation (and Article 5 of the Directive) in the context of the use of internet keywords in four judgments, namely: Joined Cases C-236/08 to C-238/08 Google France SARL v Louis Vuitton Malletier SA [2010] ECR I-2417; Case C- 278/08 BergSpechte Outdoor Reisen under Alpinschule Edi Kobmuller GmbH v Guni [2010] ECR I-2517; Case C-558/08 Portakabin Ltd v Primakabin BV [2010] ETMR 52; and Case C-323/09 Interflora Inc v Marks & Spencer plc [2012] ETMR 1.”

Then, at [141], Kitchin LJ summarised the position in the light of those decisions:

“In my judgment these cases do reveal a development by the Court of Justice of its jurisprudence on the scope of Article 9(1)(c) of the Regulation. They establish that a proprietor of a trade mark with a reputation is not necessarily entitled to prohibit the use by a competitor of his mark in relation to goods for which it is registered even though the mark has been adopted with the intention and for the purpose of taking advantage of its distinctive character and repute, the competitor will derive a real advantage from his use of the mark, and the competitor will not pay any compensation in respect of that use. Consideration must be given to whether the use is without due cause. Specifically, the use of a trade mark as a keyword in order to advertise goods which are an alternative to but not mere imitations of the goods of the proprietor and in a way which does not cause dilution or tarnishment and which does not adversely affect the functions of the trade mark must be regarded as fair competition and cannot be prohibited.”

So far as a requirement for a change in economic behaviour is concerned, the CJEU has held that proof that the use of the sign is or would be detrimental to the distinctive character of the trade mark requires evidence of a change in the economic behaviour of the average consumer of the goods or services for which the trade mark is registered or a serious likelihood that such change will occur in the future: see Case C-383-12 Environmental Manufacturing LLP v OHIM (judgment of 14 November 2013) (at [34]–[43]). It by no means follows that there is a requirement for evidence of a change in the economic behaviour of consumers of the trade mark proprietor’s goods or services in order to establish the taking of unfair advantage of the distinctive character or repute of the trade mark. In my judgment, it should be sufficient to show a change in economic behaviour of customers for the defendants’ goods or services in order to show that the use of the sign is taking unfair advantage. In Jack Wills Ltd v House of Fraser (Stores) Ltd [2014] EWHC 110 (Ch); [2014] F.S.R. 39, Arnold J proceeded on an assumption to that effect as a result of a concession by counsel (see paragraph 82) but I consider the concession to be correctly made. I do not think, however, that change of economic behaviour provides the answer to this case. The relevant economic behaviour must be that which occurs in response to ASI’s use of the sign ARGOS in relation to the service of the provision of advertising space. Had the sign not been used, internet users would neither have arrived at the site nor clicked on the ads. Even if the arrival at the site is not a relevant change of economic behaviour, clicking on the ads must amount to such a change.

That brings me to the central question of whether ASI’s use of the sign ARGOS in relation to the service of provision of advertising space took unfair advantage of the trade mark. I reject Mr Mellor’s contention that, in a case such as the present, unfairness is established by the fact of economic advantage and no more. So to hold would be to empty the word “unfair” of any meaning. Like the Court of Appeal in Whirlpool I do not consider the effect of the CJEU’s judgment in L’Oreal to go that far.

It follows that I do not accept that the judge made any error of principle in his approach to the issue of unfair advantage. He correctly identified that the case was not one involving any transfer of image to ASI’s goods or services. He went on to consider factors which have a bearing on unfairness. These were that (a) ASI had not done anything to seek out the unwanted internet traffic which arrived at its website, and which it had no power to prevent; (b) ASI’s display of AdSense ads was of some benefit to AUL by restoring misdirected customers to AUL who might otherwise have lost interest; (c) participation in AdSense was a normal and commercially unobjectionable activity; (d) the income stream derived from it by ASI was small in the context of both parties’ businesses; (e) on arriving at the website even moderately observant customers would see it had nothing to do with AUL.

The judge’s findings that the use of the sign by ASI did not cause AUL’s customers to click on ads for competitors, or cause any diversion of business away from AUL were plainly findings which were open to him. Although there was, of course, no need to show damage to AUL as opposed to advantage to ASI, these findings were plainly relevant to the issue of whether any advantage was unfair.

It is true that ASI could have ceased to participate in the AdSense programme. In those circumstances it could have left the misdirected traffic to work out for itself that it was in the wrong place or, perhaps more helpfully, have provided a simple link for internet users to click on if they were seeking AUL instead to take them back to AUL’s website. The fact that these alternatives exist does not, in my judgment, make it follow that the AdSense solution to the problem was unfair. I see no reason why, as a matter of fairness, ASI should adopt the least advantageous or most burdensome way of dealing with the unwanted traffic. Equally, having gone down that route, there is no reason to say that ASI should not take steps to optimise the advantage they receive from the AdSense programme, provided the other factors relied on by the judge are unaffected.

I therefore agree with the judge that AUL does not succeed in establishing that ASI’s use of the sign ARGOS took unfair advantage of the distinctive character or repute of the mark. There is no basis for interfering with the judge’s evaluation of the relevant factors on this issue.

It follows that the remaining issues of “due cause”, the own name defence and consent do not arise. It is fair to record that Mr Howe recognised the difficulties with “due cause” if the advantage taken was unfair, and both parties regarded the own name defence as turning on the issue of fairness as well. It is also not necessary to decide the appeal against the judge’s conclusion on the partial defence of consent.

For the reasons I have given, I would uphold the judge’s conclusion on the issue of unfair advantage, and therefore dismiss the appeal.

I agree.

I also agree.