TY - JOUR
AU1 - Colangelo,, Giuseppe
AB - ABSTRACT The impressive growth of online shopping has had a significant impact on firms’ strategies and customer behavior, bringing to the fore new forms of trademark exploitation that may affect competition. A prominent role is played by keyword advertising services provided by internet search engines. Keyword advertising systems have been the subject of several litigations with regard to the legality of the use of keywords which correspond to trademarks, since trademark holders complain that the essential functions of trademarks might be detrimentally affected. However, given the importance of search engines for attracting customers to the websites of retailers and competitors, online advertising restrictions also raise anticompetitive concerns on both sides of the Atlantic. Indeed, the E-commerce Sector Inquiry carried out by the European Commission reported that some retailers are limited in their ability to use or bid on the trademarks of certain manufacturers to get a preferential listing on search engines’ paid referencing service or are only allowed to bid on certain positions. Furthermore, the UK Competition and Markets Authority encountered brand-bidding restrictions in the markets for broadband, credit cards, energy, flights, and home insurance, while the Netherlands Authority for Consumers & Markets analyzed the hotel sector. Moreover, in the US, the Federal Trade Commission has ruled that the largest online retailer of contact lenses unlawfully entered into a web of anticompetitive agreements with rivals, preventing them from bidding for search engine result advertisements that would inform consumers that identical products were available at lower prices. The aim of this paper is to shed light on the economic rationales and legal implications of keyword advertising to strike a proper balance between trademark protection and freedom of competition. I. INTRODUCTION The interface between intellectual property (IP) protection and antitrust law has, traditionally, been controversial.1 To define the boundary between competitive markets and IP rights, two opposing views have emerged. Some call for an intrinsic conflict, pointing out unavoidable tension between the main right granted by IP laws (and the basic reward pursued by authors and inventors), that is, the right to exclude others, and the natural goal pursued by antitrust law, that is, fostering free and open competition in the market. Others, despite these remarkably different basic principles, claim that they are complementary since, although in different ways, they share the same final and long-term goals because they both seek to promote economic welfare and innovation. Hence, in the light of this finalistic convergence, the conflict is merely apparent or at least oversimplified. Furthermore, IP rights are limited in terms of duration, scope, and requirements, and thus in capability and effectiveness, to insulate IP holders from the exploitation of their art. Finally, IP laws contain internal rules and statutory limitations expressly aimed at safeguarding competition, such as the fair use exceptions, the exhaustion of rights (or first sale doctrine), and the limitations on functionality protection. Nonetheless, the potential conflict between IP protection and competition law has fueled a long debate, fed by antitrust legislation and case law that has taken different approaches over past decades, swinging back and forth from the primacy of IP towards antitrust. Indeed, the history of the IP-antitrust interface has been characterized by cycles of over- and under-enforcement that have kept the two laws from settling into a healthy balance.2 Currently, to balance competition law and IP rights, both the US and the EU prescribe relatively limited prohibitions upon IP holders’ rights. However, the two jurisdictions adopt different policies with regard to specific practices, such as refusal to license (under the essential facility doctrine),3 post-expiration royalties,4 and (minimum) resale price maintenance.5 Recently, the discussion about the interaction between IP and antitrust has been enhanced by an additional layer. Along with the increasing role played by IP rights, the rationale underlying the very existence of IP came to the fore. Indeed, due to IP expansionism and the opportunistic use of IP protection, the link between IP, innovation, and firms’ productivity is being strongly questioned, and economic literature has so far been unable to provide clear-cut empirical evidence of the theoretical positive relationship between industry’s propensity to innovate and its productivity.6 As a consequence, a new and growing strand of literature has even begun to invoke the abolition of IP rights or at least the reduction of their length and scope.7 Against this background, IP rights are under fire. Critics are focused mostly on the patent system, which has been described as broken and unnecessary for securing innovation.8 In this respect, the emergence of patent assertion entities (PAEs, often referred to as patent trolls) is considered emblematic of the potential strategic use of IP rights. Indeed, the PAEs’ business model focuses primarily on purchasing, licensing, and enforcing patents, rather than actively developing or commercializing the underlying technologies. By acting as opportunistic litigation mills, PAEs engage in ex post licensing and abuse of patent remedies to extract unreasonable royalties from practicing firms. Other significant concerns have been raised about the holdup risks involving the strategic use of standard essential patents (SEPs). Notably, in the context of standards, the dispute is over the antitrust treatment for FRAND-encumbered SEPs. According to holdup proponents, the risks of opportunistic behaviors within the standard setting process are so severe as to require antitrust intervention, rather than private contracting remedies. Hence, a FRAND commitment must be considered as equivalent to a waiver of the right to seek an injunction. On the opposing side, an IP-oriented school of thought argues that there is no justification for preventing SEP holders from exercising injunction rights, since, with regard to the test for obtaining injunctive relief, patent law already strikes a careful balance that optimizes the incentive to innovate. Finally, it is worth mentioning the antitrust scrutiny of patent settlement agreements (also known as reverse-payment or pay-for-delay settlements) in the pharmaceutical industry between originators and generic companies. According to antitrust authorities, payments from the originator to the generic company might be assumed unlawful because they subvert the competitive process by giving generic manufacturers an incentive to accept a share of their rival’s monopoly profits (abandoning their patent challenges and refraining from launching their low-cost generic products) as a substitute for actual competition in the market. Indeed, generally, a payment from one business to another in exchange for the recipient’s agreement not to compete is a paradigmatic antitrust violation. However, as with the aforementioned case of SEPs, a different and IP-sensitive approach claims that these patent settlement agreements are lawful as long as they do not exceed the exclusionary scope of the patent in suit. This brief overview shows that the interaction between IP protection and antitrust enforcement mainly concerns patents. Besides, copyrights and trademarks seldom involve an allegation of market power. In comparison with other IP rights, trademark law expressly seeks to promote competition, at least in the core cases involving competing goods. Indeed, by preventing parties from using a sign that is likely to confuse consumers about the source of their goods, trademark law reduces consumer search costs, facilitating consumer decisions and contributes to economic efficiency, creating incentives for firms to produce goods of desirable qualities.9 Therefore, unlike other IP rights, wherein protection covers the object itself (inventions, works, and designs), trademarks are protected for what they do, rather than for what they are.10 However, it has been suggested that competitively effective marks are not limitless and that the monopolization of particular terms comes at the expense of both firms and consumers.11 Due to the alleged depletion and overcrowding of trademarks, new entrants face higher costs of entry or expansion, and consumers face higher search costs due to firms choosing less memorable marks and to the blurred distinctiveness of individual words.12 Moreover, concerns have been raised about the expansion of trademark rights to encompass noncompeting goods on the basis of market pre-emption and free-riding arguments, without an analysis of the risk of confusion.13 Hence, with the aim of restoring the confusion-based rationale, it has been proposed that well-known marks should receive a narrower scope of protection than marks of lesser strength since their distinctive strength reduces the likelihood of consumer confusion.14 Finally, it is disputed whether the protection of the reputation and image of brands may serve as an objective justification for selective distribution systems. Notably, the question is whether and to what extent restrictions on internet sales of luxury products aimed at prohibiting authorized retailers from using third-party platforms should be considered lawful. Indeed, the impressive growth of online shopping has had a significant impact on firms’ strategies and customer behavior, bringing to the fore new forms of trademark exploitation that may affect competition. A prominent role is played by internet search engines and their keyword advertising service.15 Internet search engines enable individuals to find websites and online content through the use of keywords. When an internet user performs a search on the basis of one or more words, alongside natural results search engines display advertisements, appearing under the heading ‘sponsored links’, either at the upper right-hand corner or on top of the search results. While natural results are displayed on the basis of objective criteria determined by the search engine, advertisements are triggered because advertisers pay for their sites to feature in response to certain keywords. In the most well-known version of the system (Google AdWords), advertisers are free to select any keyword against the payment of a fee for each click on the sponsored link (pay per click). Since advertisers are allowed to purchase the same keyword, they bid competitively against each other for page position on the search results page.16 Namely, the order in which sponsored links are displayed is determined according to the ‘maximum price per click’, that is the number of previous clicks on those links and the quality of the advertisement as assessed by Google. Keyword advertising systems have been the subject of several litigations with regard to the legality of the use of keywords which correspond to trademarks. Indeed, to increase the chance that a given advertisement is displayed, sponsors may be interested in selecting trademarks as keywords. Conversely, trademark holders complain (against both advertisers and intermediary search engines) that the use of a competitor’s mark to trigger the display of one’s own advertising as a sponsored link might have an adverse effect on the essential functions of the trademark. Hence, trademark holders try to prevent both the selection of such keywords by third parties and the display by search engine providers of sponsored links in response to those keywords, as this may result in sites for rivals or counterfeit products being displayed. However, given the importance of search engines for attracting customers to the websites of retailers and competitors online advertising restrictions could raise anticompetitive concerns. Indeed, the E-commerce Sector Inquiry carried out by the European Commission reported that some retailers are limited in their ability to use or bid on the trademarks of certain manufacturers to get a preferential listing on the search engines’ paid referencing service or are only allowed to bid on certain positions.17 Furthermore, the UK Competition and Markets Authority (CMA) encountered brand-bidding restrictions in the markets for broadband credit cards energy flights and home insurance,18 while the Netherlands Authority for Consumers & Markets (ACM) analyzed the hotel sector.19 Moreover, the US Federal Trade Commission has ruled that the largest online retailer of contact lenses in the United States (1–800 Contacts) unlawfully entered into a web of anticompetitive agreements with fourteen rivals preventing them from bidding for search engine result advertisements that would inform consumers that identical products were available at lower prices.20 Therefore, nonbrand bidding agreements appear to be common and widespread. From a policy perspective, the debate on whether competitive advertising on brand search should be allowed revolves around the question of whether competitors are mainly confusing or informing customers. Indeed, outside the cases of counterfeiting or of adverse effect on the origin function of goods and services, competitive advertising on brand search benefits consumers providing information about the existence and features of alternative products, thereby serving the ultimate goal of trademark protection, that is, to promote fair competition through increasing transparency in the marketplace and reducing consumer search costs.21 Therefore, in the absence of a genuine trademark dispute, nonbrand bidding agreements amount to an illegal market division and bid-rigging. The goal of this paper is to shed light on the economic rationales and legal implications of keyword advertising to strike a proper balance between trademark protection and freedom of competition. To this aim, and coherently with a comparative approach, Section 2 provides an analysis of keyword advertising under EU and US trademark law, while Section 3 discusses recent antitrust cases involving restrictions on the use of trademarks as keywords. Section 4 concludes by putting forward some policy recommendations. II. KEYWORD ADVERTISING UNDER TRADEMARK LAW From the trademark law perspective, keyword advertising litigations result from the ability of advertisers to bid on keywords that correspond to brand names to enhance the likelihood of reaching consumers interested in purchasing a certain type of product or service. Trademark owners on both sides of the Atlantic have argued that such use of a trademark as a keyword by a third party violates trademark law. Furthermore, they have also asserted claims of either direct or indirect infringements against intermediary search engines, arguing that they can be liable, respectively, for deceiving customers by selling keywords corresponding to trademarks (primary liability) and for facilitating the infringing conduct of advertisers (secondary liability). A. EU Scenario European courts have been involved in trademark disputes stemming from keyword advertising so many times that, on one occasion, Advocate General Jääskinen suggested that the situation be fixed through appropriate legislative measures, rather than a re-orientation of case law.22 The landmark cases are represented by Google France23 and Interflora judgements.24 In the former, Louis Vuitton and other parties became aware that the entry by internet users, of terms constituting their trademarks into Google’s search engine triggered the display, under the heading sponsored links, of links to competitors’ sites or to sites offering imitation versions of their products and services. In the course of the proceedings brought by these trademark holders against Google, the French Court of Cassation referred some questions to the European Court of Justice (CJEU) for a preliminary ruling, essentially asking whether the use in AdWords of a keyword which corresponds to a trademark can, in itself, be regarded as a use of that trademark which is subject to the consent of its owner. The references involved neither the use of trademarks on the advertisers’ sites, nor the use of trademarks in the text of the advertisements displayed. In his opinion, Advocate General Maduro raised serious doubts about trademark holders’ requests, concerned about a risk of significant expansion of the scope of trademark protection.25 Indeed, Maduro noted that trademark holders are not only requesting an extension of the trademark protection to cover actions by a party that may contribute to a trademark infringement by a third party, in accordance with what in the US is known as contributory infringement, but are also urging the CJEU to rule that the mere possibility that a system may be used by a third party to infringe a trademark means that such a system is, itself, an infringement. From the existence of a risk that AdWords may be used to promote counterfeit sites, trademark holders deduce a general right to prevent the use of their trademarks as keywords. Therefore, they “do not wish to limit their claims to cases where AdWords is actually used by sites offering counterfeit goods; they want to nip that possibility in the bud by preventing Google from being able to make keywords corresponding to their trademarks available for selection.”26 As a consequence, if a search engine were to be placed under such an unrestricted obligation, there would be serious obstacles to any system for the delivery of information and the very nature of internet would change.27 According to Article 5 of Directive 89/104 (currently Article 10 of Directive 2015/2436) as well as to Article 9 of Regulation 40/94 (currently Article 9 of Regulation 2017/1001), trademark proprietors are allowed to prohibit third parties from using their registered signs if the following three conditions are satisfied: (a) the use of the signs takes place in the course of trade; (b) it relates to goods or services which are identical or similar to those covered by the trademarks; and (c) it affects or is liable to affect the functions of the trade mark by reason of a likelihood of confusion on the part of the public. Unlike the situation in the US, the protection provided in case of identical signs and goods or services (the so-called “double identity” situation) is considered absolute in the sense that the trademark holder does not need to show likelihood of confusion. Moreover, in the case of reputable trademarks, an enhanced protection is granted since the exclusive right conferred to trademark holders also applies in relation to goods or services which are not similar and is not conditional upon there being a likelihood of confusion, but requires that the unauthorized use of those signs takes unfair advantage of, or is detrimental to, the distinctive character or the repute of the trademarks. Indeed, according to the dilution theory, trademark law should also protect the efforts and investments made by the trademark holder and the goodwill of the trademark. In the EU and the US the notion of dilution refers to the protection against blurring (uses involving the danger that the trademark loses its distinctive character) and tarnishment (uses that endanger the reputation of the trademark). Nonetheless, in EU law dilution also covers protection against free-riding, meaning the taking of unjustified advantage of the reputation or distinctiveness of another’s trademark. In this regard, the CJEU in L’Oréal/Bellure held that a mere attempt to ride on the coat-tails of a mark with a reputation could be sufficient to assume that unfair advantage had been taken.28 Namely, “where a third party attempts, through the use of a sign similar to a mark with a reputation, to ride on the coat-tails of that mark 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 his own in that regard, the marketing effort expended by the proprietor of that mark to create and maintain the image of that mark, the advantage resulting from such use must be considered to be an advantage that has been unfairly taken of the distinctive character or the repute of that mark.”29 Against this background, Google France essentially dug around the very first condition necessary for protection, notably whether the use, in an internet referencing service, of keywords corresponding to trademarks constitutes use in the course of trade. In this respect, after recalling that the use of a sign in the course of trade occurs in the context of commercial activity with a view to economic advantage and not as a private matter, the CJEU introduced a differentiation between advertisers and referencing service providers.30 Despite their uses of keywords being linked, according to the CJEU they constitute two different uses. Indeed, whereas advertisers are using signs selected as keywords in the context of commercial activity, because that selection has the object and effect of displaying an advertising link to the site on which they offer goods or services for sale, the same does not apply to the referencing service provider. Although it is carrying out a commercial activity with a view to economic advantage when it stores keywords corresponding to trademarks and arranges for the display of advertisements on the basis of those keywords, “[t]he act of creating the technical conditions necessary for the use of a sign and being paid for that service does not mean that the party offering the service itself uses the sign.”31 Consequently, the assessment of the other conditions, mainly the one related to the effect on the functions of the trademark, involves only the advertisers. Quoting L’Oréal/Bellure, the CJEU argued that the functions of the trademark include not only the essential function of indicating origin, but also that of guaranteeing the quality of the goods or services and those of communication, investment, or advertising.32 In the case in question, the relevant functions are the function of indicating origin and the function of advertising. The latter pertains to the trademark holder’s use of the mark for advertising purposes designed to inform and persuade consumers, and it is infringed if the use by a third party adversely affects the proprietor’s use of its mark as a factor in sales promotion or as an instrument of commercial strategy.33 With regard to the origin function, the CJEU pointed out the relevance of the manner in which the advertisement is presented. Namely, the function is adversely affected if the third party’s advertisement triggered by the keyword does not enable normally informed and reasonably attentive internet users (or enables them only with difficulty) to ascertain whether the goods or services referred to by the advertisement originate from the owner of the trademark or an undertaking economically connected to it or, on the contrary, originate from a third party.34 There is an adverse effect on the origin function when the advertisement suggests the existence of an economic link between the third party and the holder of the trademark, or even when it is so vague about the origin of the goods or services that normally informed and reasonably attentive internet users are unable to determine whether the advertiser is a third party vis-à-vis the proprietor of the trade mark or, on the contrary, economically linked to that proprietor.35 Internet users must be able to recognize the origin of the goods or services on the basis of the advertising link and the commercial message attached thereto. Thus, the origin must be determinable by virtue of the advertisement itself, without having to visit the advertiser’s website. Conversely, although the use by internet advertisers of a sign identical to another person’s trademark as a keyword has certain repercussions on the advertising use of that mark, these repercussions do not of themselves constitute an adverse effect on the advertising function of the trademark.36 Indeed, according to the CJEU, the visibility to internet users of the goods or services of the trademark holder is guaranteed, irrespective of whether that proprietor is successful in also securing the display, in one of the highest positions, of an advertisement under the heading ‘sponsored links’, because when internet users enter the name of a trademark as a search term, the home and advertising page of the proprietor of that mark will appear in the list of the natural results, usually in one of the highest positions on that list.37 As the Court will further argue in Interflora, the mere fact that the unauthorized use by a third party of a sign obliges the proprietor of that mark to intensify its advertising to maintain or enhance its profile with consumers is not a sufficient basis, in every case, for concluding that the trademark’s advertising function is adversely affected.38 Finally, insofar as the role played by the service provider is neutral, in the sense that its conduct is merely technical, automatic and passive, pursuant to Article 14 of the e-commerce Directive 2000/31 it cannot be held liable before its being informed of the unlawful conduct of that advertiser. Agreement between the keyword selected and the search term entered by an internet user is not sufficient in itself to deprive the service provider of the exemptions from liability provided for in Directive 2000/31, and to justify the view that the service provider has knowledge of, or control over, the data entered into its system by advertisers and stored in memory on its server.39 By contrast, the role played by the service provider in the drafting of the commercial message which accompanies the advertising link or in the establishment or selection of keywords is relevant.40 Therefore, if the service provider has not played an active role, it cannot be held liable for the data which it has stored at the request of an advertiser, unless having obtained knowledge of the unlawful nature of those data or of that advertiser’s activities, it failed to act expeditiously to remove or to disable access to the data concerned.41 The Google France approach has been confirmed by the Court on several occasions and also applied to keywords reproducing a trademark with minor spelling mistakes (Portakabin) and to customers/sellers of an online-marketplace operator (eBay).42 In the latter case, the CJEU distinguished the situation in which an online-marketplace operator chooses keywords corresponding to trademarks in a search engine’s referencing service to trigger a sponsored link to its website from the situation in which the same online-marketplace operator provides a service consisting in enabling its customers to display on its website signs corresponding to trademarks. While in the former situation the operator of the online marketplace is acting as an advertiser, hence it is using the signs identical with or similar to trademarks in the course of trade, in the latter situation the use of those signs is made by the sellers who are customers of the operator of that marketplace and not by that operator itself.43 Against this background, the Interflora decision deserves special attention since the CJEU dealt with the selection of a keyword corresponding to a competitor’s trademark with a reputation; hence, it addressed the issue of the potential adverse effect of keyword advertising on the investment function of trademarks as well. Notably, the CJEU stated that a trademark may also be used by its proprietor to acquire or preserve a reputation capable of attracting consumers and retaining their loyalty.44 Although the investment function may overlap with the advertising function, it is nonetheless distinct from the latter because, when the trademark is used to acquire or preserve a reputation, not only is advertising employed, but also various commercial techniques.45 The investment function is adversely affected whether the use by a third party of a sign identical to a reputed trademark in relation to identical goods or services affects that reputation and thereby jeopardizes its maintenance, and whether it substantially interferes with the proprietor’s use of its trademark to acquire or preserve a reputation capable of attracting consumers and retaining their loyalty.46 Nonetheless, to safeguard the freedom of competition, as in the case of the advertising function in Google France, the CJEU introduced a significant limit to leverage the investment function in keyword advertising cases. “[I]n conditions of fair competition that respect the trademark’s function as an indication of origin”, it cannot be accepted that the holder of a trademark may prevent a competitor from using an identical sign, if the only consequence of that use is to oblige the proprietor of that trademark to adapt its efforts to acquire or preserve a reputation capable of attracting consumers and retaining their loyalty.47 Furthermore, the fact that that use may prompt some consumers to switch from goods or services bearing that trademark cannot be successfully relied on by the proprietor of the mark. By the same token, Advocate General Jääskinen noted that one of the blessings of the internet is that it greatly enhances consumers’ opportunities to make enlightened choices between goods and services.48 Finally, the CJEU addressed the issue of the conditions under which keyword advertising must be regarded as causing detriment to the distinctive character of the mark with a reputation (dilution) or as taking unfair advantage of the distinctive character or repute of that trademark (free-riding). With regard to the former, a case-by-case assessment of the evidence is needed. Indeed, the selection as a keyword of a sign which is identical or similar to a reputed trademark does not necessarily contribute to turning the trademark into a generic term, since it may merely serve to draw the internet user’s attention to the existence of an alternative product or service to that of the trademark holder.49 Therefore, if it has been ascertained that the trademark’s function of indicating origin has been adversely affected because the advertising does not enable a reasonably well-informed and observant internet user to tell that the product or the service promoted by the advertiser is independent from that of the trademark holder, then it must be demonstrated that the advertising has also caused detriment to the distinctive character of the trademark by contributing to turning it into a generic term. On the free-riding allegation, the CJEU referred to the coat-tail formula developed in L’Oréal/Bellure arguing that the selection as internet keywords of signs corresponding to reputed trademarks can be construed as a use whereby the advertiser rides on the coat-tails of a trademark with a reputation 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 holder of that mark to create and maintain the image of that mark.50 Indeed, it cannot be denied that a competitor selects a well-known trademark as a keyword because it expects that a large number of consumers will use that keyword to carry out an internet search, thus with the intention to take advantage of the distinctive character and the goodwill of the trademark.51 Nonetheless, the Court delimited and circumscribed the free-riding hypothesis to the Google France situation, that is, to the case in which advertisers offer for sale goods which are imitations of the goods of the proprietor of reputed trademarks.52 By contrast, the holder of a reputed trademark is not entitled to prevent advertisements which put forward alternative goods or services.53 In summary, as shown by this brief overview of the European case law, keyword advertising represents the new scene of the traditional dispute over the balance between interests in protecting IP and the freedom of commerce and competition. The trademark is an essential element in the system of undistorted competition which European law seeks to establish, insofar as it fulfils its function. However, since it has been used the words “functions” of trademark in Arsenal Football Club,54 the CJEU has progressively endorsed an expansion of the scope of trademark rights indicating that the essential function of indicating origin is not the only function of the trademark that is worthy of protection.55 In L’Oréal/Bellure the Court held explicitly that the functions of the trademark also include those of guaranteeing the quality of the goods or services and those of communication, investment, or advertising.56 In this scenario, the keyword advertising jurisprudence allows the CJEU to reconsider the balance between trademark protection and freedom of competition, since the expansionist approach becomes even more questionable in the digital landscape.57 As clearly argued by Advocate General Maduro in Google France keyword advertising cases call for such a balance because the promotion of innovation and investment also requires competition and open access to ideas, words, and signs.58 Trademark holders do not have an absolute right of control over the use of their marks, and their interests are not sufficient to prevent consumers from benefiting from a competitive market.59 By claiming the right to exert control over keywords which correspond to trademarks in internet referencing services, trademark holders could prevent internet users from viewing rivals’ advertisements for perfectly legitimate activities.60 In the same vein, Advocate General Jääskinen in Interflora found the coat-tail formula elaborated in L’Oréal/Bellure “very problematic” from the point of view of competition resulting in a move away from a Pareto optimal situation.61 Indeed, “[t]he situation of the trademark proprietor would not improve as he by definition would not suffer any detriment because of the use, but the competitor’s situation would worsen because he would lose a part of his business. Also, the situation of the consumers who had not been misled by the ad but consciously preferred to buy the competitor’s products would be impaired.”62 By sharing the concerns about the anticompetitive impact of trademark protection in keyword advertising, the CJEU’s rulings are in striking contrast with the expansionist trend of the Court’s functions jurisprudence.63 Indeed, these rulings notably limit the control of trademark holders over their signs against both search engines and rival advertisers, restoring the primacy of the confusion-based rationale of trademark protection. Whereas for the former, the use of keywords corresponding to trademarks in an internet referencing service is not considered use in the course of trade, the latter are liable insofar as the origin function is affected. Admittedly, the origin function is interpreted broadly, since an adverse effect is found even when the advertisement, while not suggesting the existence of an economic link, is so vague about the origin of the goods or services that normally informed and reasonably attentive internet users are unable to determine whether the advertiser is a third party vis-à-vis the proprietor of the trade mark or, on the contrary, economically linked to that proprietor.64 Nonetheless, neither the advertisement function nor the investment function plays a significant role. Indeed, pursuant to EU case law, the repercussions on the advertising use of a mark identical to a keyword that triggers a competitor’s advertisement as a sponsored link do not of themselves constitute an adverse effect on the advertising function of the trademark, since the visibility to internet users of the goods or services of the trademark holder is guaranteed because the home and advertising page of the proprietor of that mark will appear in one of the highest positions in the list of the natural search results.65 The mere fact that the unauthorized use by a third party of a sign obliges the proprietor of that mark to intensify its advertising is not a sufficient basis for concluding that the trademark’s advertising function is adversely affected.66 Furthermore, regarding the investment function, whether or not the competition respects a trademark’s origin function, it cannot be accepted that the holder of a trademark may prevent a competitor from using an identical sign, if the only consequence of that use is to oblige the proprietor of that trademark to adapt its efforts to acquire or preserve a reputation capable of attracting consumers and retaining their loyalty.67 Likewise, the fact that that use may prompt some consumers to switch to rivals’ goods or services cannot be successfully relied on by the proprietor of the mark. B. US Scenario On the other side of the Atlantic, the expansive approach to trademark protection embraces an additional layer, broadening the class of entities which can be held liable for using keywords corresponding to trademarks. Indeed, in contrast to the European approach, internet intermediaries may face responsibility in terms of both secondary and direct infringement. Regarding the former, in Rosetta Stone the Fourth Circuit stated that Google could be held liable for secondary contributory infringement if it has allowed known infringers and counterfeiters to bid on the Rosetta Stone marks as keywords.68 Namely, pursuant to the well-known Inwood Laboratories test, if a manufacturer or distributor intentionally induces another to infringe a trademark, or if it continues to supply its product to one whom it knows or has reason to know is engaging in trademark infringement, the manufacturer or distributor is contributorily responsible for any harm done as a result of the deceit.69 This two-part test crafted by the Supreme Court for analyzing contributory trademark infringement has also been routinely applied in the digital environment. In particular, in Tiffany v. eBay the Second Circuit rejected a contributory infringement claim reiterating that, to satisfy the Inwood test, it is not enough to have general knowledge as to counterfeiting on its website to impose upon eBay an affirmative duty to remedy the problem.70 Rather, the defendant must supply its product or service to identified individuals that it knows or has reason to know are engaging in trademark infringement. Thereby, in the case at stake Tiffany would have to show that eBay knew or had reason to know of specific instances of actual infringement. However, a service provider is not permitted willful blindness. Hence, contributory liability may arise if eBay had reason to suspect that counterfeit Tiffany goods were being sold through its website, and intentionally shielded itself from discovering the offending listings or the identity of the sellers behind them. Willful blindness is considered equivalent to actual knowledge. In this regard, efforts undertaken by eBay were relevant for inducing the Court to dismiss the charges. Indeed, eBay implemented, among other things, a fraud engine dedicated to ferreting out illegal listings and employed manual searches for keywords in listings in an effort to identify blatant instances of potentially infringing activity. The transatlantic divergence emerges essentially in relation to internet intermediaries’ liability for direct infringement. Whereas according to the Google France jurisprudence, by storing and making available keywords corresponding to trademarks to trigger advertisements, an internet referencing service provider does not use those signs in the course of trade, in the US starting with Playboy Enterprises, courts found that search engines’ use of trademark as advertising keywords amounts to use in commerce.71 Courts thereby opened the door to direct infringement claims against internet intermediaries that create likelihood of confusion. The meaning of the Lanham Act §45 (15 U.S.C. §1127) definition of use in commerce in a keyword advertising scenario has been thoroughly addressed by the Second Circuit in Rescuecom.72 Stating that Google’s recommendation and sale of Rescuecom’s mark to its advertising customers were not internal uses, the Court clarified the reasoning of its previous ruling in 1–800 Contacts, where it dismissed the claim of a trademark holder arguing that the use of its mark to generate pop-up advertisements did not qualify as trademark use, but rather a company’s internal utilization.73 On this ground, some district courts dismissed similar suits holding that use of a trademark in metadata did not constitute trademark use within the meaning of the Lanham Act because the use is strictly internal and not communicated to the public,74 likewise the internal use of a keyword to trigger advertisements.75 However, according to the Second Circuit, these decisions over-read the 1–800 Contacts ruling. Indeed, keyword advertising services contrasts with some important aspects of the 1–800 Contacts context, since in the pop-up case, to the extent that an advertisement for a competitor was displayed when a user opened the plaintiff’s website, the trigger to display the advertisement was not based on the defendant’s sale or recommendation of a particular trademark. Conversely, in Rescuecom Google was recommending, displaying, and selling to its advertisers Rescuecom’s trademark.76 Therefore, the Second Circuit concluded that the use of keywords by Google in its AdWords program fitted within the terms specified by the Lanham Act.77 Once the use requirement is satisfied, the gate is open to direct infringement claims against internet intermediaries and courts are allowed to assess whether the unauthorized use of a trademark as keyword causes likelihood of confusion. In this regard, an expansive approach to trademark law in another dimension has also been noted, whereby courts have progressively endorsed an interpretation of the initial interest confusion doctrine that shifted the focus of infringement analysis away from consumer confusion and toward a generalized inquiry into whether a challenged use diverts attention away from the trademark holder.78 Initial interest confusion is customer confusion that creates initial interest in a competitor’s product.79 It occurs when a customer is lured to a product by the similarity of the mark, even if the customer realizes the true source of the goods before the sale is completed.80 Therefore, this doctrine allows for a finding of liability where a plaintiff can demonstrate that a consumer was confused by a defendant’s conduct at the time of interest in a product or service, even if that initial confusion is corrected by the time of purchase. In the cases at issue, trademark holders assert that, by keying advertisements to their marks, internet intermediaries actively create initial interest confusion because would-be purchasers of their products or services who search for their website are misleadingly directed to the advertisements and websites of competitors in a manner which leads them to believe that these ads or websites are sponsored by or affiliated with trademark holders. Furthermore, even if would-be purchasers realize upon accessing the competitor’s site that they have reached a site unrelated to the trademark holder, through initial consumer confusion the competitor will still have gained a customer by appropriating the goodwill that the trademark holder has developed in its mark. To analyze likelihood of confusion, US courts have elaborated tests including factors that are generally considered relevant to the inquiry.81 Nonetheless, these factors are not a rote checklist, but are intended as an adaptable proxy for consumer confusion, and thus, they are not exhaustive and are applied in a flexible way. In the internet scenario some factors may not be applied, and some are considered more important than others.82 Indeed, as argued in Network Automation, in the keyword advertising context the likelihood of confusion will ultimately turn on what the consumer saw on the screen and reasonably believed.83 Notably, given the nature of the alleged infringement at issue, the Ninth Circuit considered the labeling and appearance of the advertisements and the surrounding context on the screen displaying the results page the most relevant factor to the assessment of the likelihood of confusion. Furthermore, the court argued that the proximity of the goods would become less important if advertisements are clearly labeled or consumers exercise a high degree of care, because rather than being misled, the consumer would merely be confronted with choices among similar products. Moreover, explaining that the nature of the goods and the type of consumer is highly relevant to determining the likelihood of confusion in the keyword advertising context, the court specified that the default degree of consumer care is becoming more heightened as the novelty of the internet evaporates and online commerce becomes commonplace.84 However, the Fourth Circuit in Rosetta Stone was less prone to rely on consumer sophistication noting that, according to an internal Google study, even well-educated, seasoned internet consumers are confused by the nature of sponsored links and are sometimes even unaware that sponsored links are, in actuality, advertisements.85 By reversing the district court’s conclusion that no reasonable trier of fact could find that Google intended to create confusion by permitting the use of Rosetta Stone’s trademark as keywords, the Circuit opened the possibility of finding trademark infringement on keyword use. Moreover, the court found that Rosetta Stone proffered sufficient evidence of actual consumer confusion to survive summary judgment. In 1–800 Contacts v. Lens.com, the Tenth Circuit affirmed the district court’s grant of summary judgment to defendant on the ground that no likelihood of confusion existed, stating that traditional analysis and actual marketplace data revealed that the keyword use by Lens.com and its affiliates was highly unlikely to divert consumers.86 Notably, the court noted that, even if consumers in general may not much care what retailer supplies their contact lenses, the consumers relevant to the suit were looking for a particular retailer: “Presumably they have narrowed their search because they have already selected 1–800 as the preferred retailer and are searching for its website or perhaps commentary on its performance.”87 In the abstract, one who searches for a particular business with a strong mark and sees an entry on the results page will naturally infer that the entry is for that business. “But that inference is an unnatural one when the entry is clearly labeled as an advertisement and clearly identifies the source, which has a name quite different from the business being searched for.”88 Furthermore, the ratio of clicks to impressions associated with Lens.com’s own keyword use did not support an inference that Lens.com’s keyword activity was likely to lure consumers away from 1–800. Indeed, the clickthrough rate was very low, revealing that initial-interest confusion occurred at most 1.5% of the time that a Lens.com ad was generated by the challenged keywords.89 Finally, with regard to the relevance of the design of the web page that is displaying the competing mark and offering the competing products for sale, it is worth mentioning the Multi Time Machine v. Amazon ruling.90 Referring to Playboy Enterprises and Network Automation, the Ninth Circuit confirmed that the labeling and appearance of the products for sale on an internet intermediary’s web page is the most important factor, since clear labeling can eliminate the likelihood of initial interest confusion in cases involving internet search terms. In the case at issue, MTM watches were not listed on the page because neither Amazon nor MTM sold MTM watches on Amazon. However, MTM contended that initial interest confusion might occur because Amazon listed the search term used (the trademarked phrase “mtm special ops”) at the top of the search page, thereby generating the risk that a consumer might consider the products displayed as types of MTM watches. According to the trademark holder, to eliminate the likelihood of confusion, Amazon should change its search results page so that it explained to customers that it did not offer MTM watches for sale before suggesting alternative watches to the customer. The court disagreed and found consumer confusion highly unlikely in light of Amazon’s clear labeling of the products it carried, by brand name and model, accompanied by a photograph of the item. In summary, unlike EU case law, US courts apparently adopted an expansive approach without any significant self-restraint. Indeed, the abovementioned rulings do not contain any analysis of the potential anticompetitive impact of trademark protection in keyword advertising. Rather, by stating that search engines’ use of trademarks as advertising keywords amounts to use in commerce, US courts opened the door to internet intermediaries’ liability for direct infringement. Nonetheless, trademark holders may find it hard to prevail against both internet intermediaries and keyword advertising competitors. Indeed, keyword advertising, by itself, does not constitute an infringement. As recently declared by a district court, virtually no court has held that the purchase of a competitor’s marks as keywords alone, without additional behavior that confuses consumers, is sufficient for liability.91 Hence, the crucial question is whether competitor-targeted advertising is more akin to a bait-and-switch likely to confuse consumers or to the simple act of offering consumers a choice. In this regard, it is worth noting that in several cases courts did not find evidence supporting the finding that the use of a trademark as keyword was likely to cause confusion among consumers.92 C. Empirical Evidence Since the legal assessment of keyword advertising is centered on the question of whether competitors are mainly confusing or informing customers, empirical investigations aimed at determining the effect of keyword advertising policies on consumer behavior are of utmost relevance. Indeed, it has been suggested that, due to objective opaqueness, the searcher’s goals are not clear from the keyword, and thus, one cannot make any legally supportable inferences about searcher objectives based on the keywords used.93 Empirical studies may support or contradict this argument, shedding light on the actual existence of customer confusion as a result of competitive advertising. First insights are provided by Franklyn and Hyman.94 Their study reported little evidence of actionable consumer confusion regarding the source of goods, noting however that only a small minority of consumers correctly and consistently distinguished paid advertisements from unpaid search results. Exploiting the change in the European Google’s Adwords policy after the Google France ruling as a natural experiment, Bechtold and Tucker found no large measurable average effect on browsing behavior.95 The study confirmed the multifaceted ways in which search engine users are using trademarks. Indeed, only 20 percent of searches appeared to be purely navigational, meaning that few consumers were searching for the keyword because they were directly interested in using the search engine as a shortcut to find the trademark owners’ website. According to the study, the lack of net effect after the AdWords policy change stemmed from two opposing effects in searchers’ behavior: while navigational searches were less likely to lead to the trademark owner’s website (decreasing by 9 percent), nonnavigational searches were more likely to lead users to the trademark owner’s website (increasing by 5 percent). Further evidence that consumer motivations behind search queries are complex and heterogeneous is provided by Dotson et al., who investigated the relationship between brand attitudes and search engine queries using a micro-level data set collected from a panel of Google users who agreed to track their individual brand search behavior over 8 weeks and link this search history to their responses to a brand attitude survey.96 Notably, focusing on the smartphone and automotive markets, their analysis found that users who are actively shopping in a category are more likely to search for any brand. More recently, using a large-scale randomized advertisement allocation on Bing, Simonov and Hill attempted to measure the effectiveness of competitive advertising on brand search. In particular, relying on data on links on the results page and consumers’ click and post-click decisions, their study aimed at assessing whether competitors’ paid links attract a focal brand’s searchers, how effective and expensive defensive advertising is for the focal brand, and the success rates for clicks that are siphoned off by competitors’ paid links.97 The reported evidence shows that competitors’ paid links allow them to divert large shares of traffic from the focal brand, crucially depending on whether the focal brand’s link occupies the top paid position on the search results page.98 These results are in line with the field test run on Edmunds.com by Coviello et al., who found that only about half of the traffic normally flowing to Edmunds.com through branded search advertisements still flowed to the website when it relied only on organic search links, thus suggesting that, for a company in the top one percent of the most-visited websites, paid brand search advertising increases traffic to the website.99 In contrast, Blake et al.100 and Golden and Horton101, respectively, conducting an experiment with eBay’s sponsored search advertisements and with a company that temporarily suspended its sponsored search advertising campaign on Google, found that brand search advertising has a negligible effect on the focal brand’s traffic, thus supporting the navigational nature of brand search. However, according to Simonov and Hill, even if competitors’ paid links allow them to divert large shares of traffic from the focal brand, these diverted clicks are unlikely to lead to immediate purchases. Indeed, 46 percent of consumers return to Bing in less than 30 seconds after the click, compared with 3.5–6 percent for consumers clicking on the focal brand’s link. The high probability of quick returns after a click on a competitors’ link is due both to negative selection by customers and an incremental increase in the overall number of unsuccessful clicks, with the latter being consistent with both customer confusion and deliberate search. To sum up, despite different results about the navigational nature of brand search and thereby on the capability of advertising on competitors’ keywords to divert significant shares of traffic, the empirical studies seem to share a common view about the limited risk of consumer confusion generated by competitive advertising on brand search, suggesting that the nature of competitors’ click is more consistent with informational deliberate search. III. RESTRICTIONS TO KEYWORD ADVERTISING UNDER ANTITRUST LAW In addition to trademark litigation, keyword advertising has recently also been under the scrutiny of antitrust authorities. Indeed, given the impressive growth of internet shopping and role of search engines in attracting customers to the websites of retailers and competitors, brand-bidding restrictions could raise anticompetitive concerns. Notably, several strategies emerged. On the European side, within selective distribution networks, retailers may be limited to use or bid on the trademarks of certain manufacturers to get a preferential listing on the search engines’ paid referencing service or are only allowed to bid on certain positions. In this regard, in the aftermath of the E-commerce Sector Inquiry, the European Commission challenged the online search advertising restrictions imposed by Guess to reduce competition from independent online retailers.102 Furthermore, in the UK, the CMA assessed the effects of different types of advertising restrictions (narrow nonbrand bidding, wide nonbrand bidding, and negative matching agreements) in the markets for broadband, credit cards, energy, flights, and home insurance.103 Moreover, in the Netherlands, the ACM reported that hotels, by negotiating nonbrand bidding agreements with online travel agents (OTAs), can to some extent shield their own online sales channel against competition from other hotels present on the OTAs.104 Finally, on the US side, the Federal Trade Commission examined trademark settlement agreements aimed at preventing rivals from bidding for search engine result advertisements.105 A. EU Scenario: Online Advertising Restrictions Within Selective Distribution Networks In the EU, antitrust concerns about restrictions to keyword advertising have been raised in the context of selective distribution systems, notably on online advertising restrictions by which the head of a selective distribution network relies on its trademark rights to prevent admitted resellers from bidding on its marks as keywords for online advertising services. Such restrictions aim at preventing retailer websites from appearing prominently, thereby effectively competing with their own e-shops. Indeed, the E-commerce Sector Inquiry recently carried out by the European Commission reported that some retailers are limited in their ability to use or bid on the trademarks of certain manufacturers to get a preferential listing on the search engines paid referencing service or are only allowed to bid on certain positions.106 The Commission acknowledged that this may be in the interest of the manufacturer to allow its own retail activities to benefit from a top listing and keep bidding prices down. However, such restrictions could raise concerns as anticompetitive agreements under Article 101 TFEU, should they restrict the effective use of the internet as a sales channel by limiting the ability of retailers to direct customers to their websites. Conversely, restrictions on the ability of retailers to use the trademark name of the manufacturer in the retailer’s own domain name rather help avoid confusion with the manufacturer’s website. The Commission’s approach reflects the principles affirmed by the CJEU’s case law which has recently addressed some crucial issues regarding the scope and the effects of selective distribution systems (increasingly used in response to the growth of online sales) and the legitimacy of certain vertical restrictions limiting e-commerce (e.g., pricing restrictions, marketplace bans, restrictions on price comparison tools, and exclusions of pure online players). Indeed, a retailer’s choice to sell the relevant products through its own online shop, a physical store or a broader platform may eventually raise objections and disagreements among suppliers. This is because the cited means have a different significance to the category of suppliers, depending inter alia on the relevance attributed to the goals of preserving the product brand, maximizing the sales volume or providing a high-quality shopping experience. In this regard, while in Pierre Fabre the CJEU stated that the goal of maintaining a prestigious image was not a legitimate aim for restricting competition,107 the Coty Prestige ruling endorsed the view that luxury goods may require the implementation of a selective distribution system to preserve their quality.108 Indeed, the quality of luxury goods is not just the result of their material characteristics, but also of the allure and prestigious image which bestow on them an aura of luxury. In this scenario, the Court considered the principle established in Pierre Fabre to be confined to the context of that judgment. Namely, that assertion is related solely to the goods at issue (the goods covered by the selective distribution system at issue in that case were not considered luxury goods) and to the contractual clause in question in Pierre Fabre (a general and absolute ban on internet sales). Against this background, in Coty the CJEU considered lawful a clause prohibiting authorized retailers from using, in a discernible manner, third-party platforms for internet sales of luxury products.109 Hence, an absolute online marketplace ban should not be considered as anticompetitive insofar as it does not amount to prohibition on selling online and does not restrict the effective use of the internet as a sales channel. Furthermore, it is worth noting that the CJEU relied on a previous case (Copad), in which the concepts of prestigious image, allure, and ‘aura of luxury’ of a product in the eyes of consumers had been given prominent attention, though on the basis of trademark rights rather than competition law.110 In Copad the CJEU held that, in the context of a selective distribution system, a brand owner can rely on its trademark to prevent its licensees and authorized distributors from supplying discount stores where such resale would degrade the ‘aura of luxury’ and prestigious image of the goods, and the subsequent resale of the products by unauthorized dealers, where such resale would weaken the reputation of the branded goods. The judgment extended the right for manufacturers or brand owners to seek relief directly against unauthorized resellers, hence going beyond what is currently permitted under EU competition law. In fact, the supplier’s right to prevent, under certain conditions, its authorized dealers from supplying discount stores conflicts with the EU competition law principle that authorized distributors should remain free to sell to all end users. In addition, whereas EU competition law prevents manufacturers from imposing restrictions on their distributors’ customers, Copad allows the supplier or brand owner to prevent the unauthorized seller from selling (or advertising) the branded goods when the subsequent commercialization outside the selective distribution system weakens the image of the products. Against this background, as a follow-up to the E-commerce Sector Inquiry the European Commission launched an investigation into the Guess selective distribution network challenging, inter alia, the online search advertising restrictions imposed by Guess to reduce competition from independent online retailers.111 Guess established a hybrid distribution system within which it acted as wholesaler and retailer in competition with the resellers admitted to its selective distribution network. In addition to being sold in bricks-and-mortar stores, the branded products were also sold online both by Guess itself, through its own e-shop or e-commerce marketplaces, and by pure online retailers. To prevent its own e-commerce site from being cannibalized by rivals, Guess systematically banned the admitted resellers (both mono-brand and multibrand retailers) from using its brand names and trademarks as keywords in Google AdWords auctions. In general, a contractual clause within a selective distribution agreement is lawful provided that it has a legitimate objective, is laid down uniformly for all potential resellers, is applied in a nondiscriminatory fashion, and does not go beyond what is necessary. Guess claimed that these restrictions were conceived in line with the strategy to safeguard the prestigious brand image of its products and to enhance the value and reputation of its trademarks. However, according to the Commission, Guess pursued different objectives when it came to its AdWords policy.112 In particular, Guess sought to maximize traffic to its own website at the expense of the independent Guess distributors and to minimize its own advertisement costs. In other words, to reduce competitive pressure from authorized retailers on Guess’ own online retail activities by curtailing the ability of authorized retailers to use this advertising tool effectively, and to keep down its AdWords costs. Banning the use of the Guess brand names and trademarks in AdWords restricted the “findability” and ultimately the viability of authorized online retailers within Guess’ selective distribution system. Thus, an exclusive right reserved for Guess to use its brand names and trademarks in online search advertising provided Guess with a considerable competitive advantage over its retailers with whom it competed online, and restricted intra-brand competition.113 To strike a balance between the scope of trademark protection and the goals of competition law, the Commission assessed the restrictions at issue mainly referring to trademark law cases, that is, Google France and Interflora rulings.114 Firstly, the Commission noted that Google France cannot be relied on to justify a restriction of the ability of authorized retailers in selective distribution systems, who sell genuine Guess products, to use or bid on Guess’ brand names and trademarks, as in this case there is no risk of confusion as to the origin of the products. Furthermore, the reduction in advertising costs could not be regarded as a legitimate objective of a selective distribution network, since according to Interflora internet advertising using a referencing service on the basis of keywords corresponding to another person’s trademark constitutes a practice inherent to competition as it offers internet users alternatives to the trademark proprietor’s goods or services even if it leads to the trademark proprietor having to intensify its advertising to maintain or enhance its profile with consumers. Given that the restriction on the use of the Guess brand names and trademarks in online search advertising did not pursue any legitimate objectives in the context of the operation of the selective distribution system, the Commission considered the online search advertising restriction incompatible with antitrust provisions. B. The UK and the Netherlands Antitrust Authorities’ Studies Keyword advertising restrictions appear common and widespread. Indeed, in sectors where online sales play an important role it has been reported that agreements not to bid on other firms’ trademarked terms in search engine keyword auctions are increasingly being adopted. Notably, the UK CMA encountered these restrictions in the markets for broadband, credit cards, energy, flights, and home insurance, while the Netherlands ACM dealt with the hotel sector. Within its market study on digital comparison tools, the British antitrust authority analyzed the impact of different types of advertising restrictions.115 Namely narrow nonbrand bidding agreements (where the restricted advertiser agrees not to bid on another advertiser’s brand name when the search term only includes that brand name) wide nonbrand bidding agreements (where the restricted advertiser agrees not to bid on another advertiser’s brand name when the search term includes that brand name alone or with other nonbrand related words) and negative matching agreements (where the restricted advertiser agrees to add another advertiser’s brand name to its negative keywords which prevents its advertisement appearing when the search term includes that brand name alone or with other nonbrand related words). According to the CMA, wide nonbrand bidding and negative matching agreements could affect the paid search results that consumers see in response to their search terms, since they are likely to have a bigger impact on click-through and conversion rates (i.e. the extent to which consumers that click on the advertising go on to make a purchase) both for the restricted advertiser and for the brand owner, than narrow nonbrand bidding.116 Indeed, these agreements are more likely to cover search terms that consumers use to shop around, and hence, the consumer would be more likely to click on the link of the brand owner’s rival if that appeared in response to the search term. In contrast, narrow nonbrand bidding relates to brand name-only searches. Furthermore, negative matching agreements are likely to affect a greater number of searches than wide nonbrand bidding and narrow nonbrand bidding, because negative matching agreements affect all searches that include the brand owner’s name irrespective of the context whereas with wide nonbrand bidding a restricted advertiser may still appear due to relevance.117 However, despite the fact that wide nonbrand bidding and negative matching agreements have the potential to lead to consumer harm, the CMA concluded that the evidence gathered as part of the market study does not suggest that these agreements are currently having a significant impact on consumers.118 The Dutch antitrust authority instead addressed a specific strategy put in place in the hotel sector, which is also under scrutiny in the US.119 Indeed, in Tichy v. Hyatt Hotels a district court recently denied the motion to dismiss an antitrust claim against major hotel chains, which agreed to stop bidding on each other’s trademarks and to impose restrictions in their OTAs agreements that prohibited OTAs from bidding on their branded keywords.120 The aim of the Netherlands ACM was to test whether the difference in price on hotels’ websites and the price on OTAs are positively or negatively affected by the presence of a nonbrand bidding agreement. Indeed, the underlying theory of harm was that hotels, by negotiating nonbrand bidding agreements with OTAs, can to some extent shield their own online sales channel against competition from other hotels present on the OTAs. In particular, since a customer using a branded keyword on a search engine is considering a specific hotel, a nonbrand bidding agreement reduces the probability that the consumer will visit the OTA, where they are confronted with many other brands.121 Therefore, by reducing the extent to which consumers compare different brands, the nonbrand bidding agreement allows the hotel to charge a higher price on its own website, inducing market segmentation as OTAs cannot target customers that use a branded keyword with search ads. The results of the empirical analysis conducted by the ACM are consistent with the abovementioned concerns, showing that nonbrand bidding agreements lead to a higher price on the hotel’s website relative to the OTA price (around 2 percent).122 C. US Scenario: Trademark Settlement Agreements The US landscape has brought to the fore another keyword-related strategy that appears challenging for antitrust enforcers. Indeed, while nonbrand bidding agreements among competitors appear at first glance an illegal form of market division and bid-rigging, they may be justified by the aim of resolving infringement claims on trademarked keywords, thereby acting as search advertising settlement agreements. In general, both the trademark protection and the avoidance of litigation costs through settlement are valid procompetitive justifications. However, at the same time keyword bid-suppression agreements interfere with search engine auctions cutting off a mode of competition for customers. Affecting the ability to comparison-shop, these trademark settlement agreements may therefore harm both consumers, by causing higher prices and less choice, and the competitive process, by prohibiting competitors from having the opportunity to challenge a dominant player. The FTC addressed this conundrum in the 1–800 Contacts case, finding restrictions on search term bidding unlawful.123 Notably, on November 2018, the FTC issued an opinion condemning as an antitrust violation the trademark settlement agreements between 1–800 Contacts and fourteen online sellers of contact lenses that limit internet search advertising and restrict bidding in internet search auctions to the detriment of consumers. The settlement agreements arise from trademark infringement claims brought by 1–800 Contacts against online rivals which bought advertisements using 1–800 Contacts’ trademarks as keywords. In nearly all cases, the litigation settled before trial. Indeed, between 2004 and 2013, 1–800 Contacts entered into several settlement agreements to resolve these disputes. In the suit against Lens.com the case went to a judge, but the court did not resolve the question of whether use of challenged trademark keywords, divorced from the text of the resulting ads, could result in a likelihood of confusion, because it found that 1–800 Contacts’ infringement claim failed due to lack of adequate evidence of confusion.124 The settlement agreements required the parties to refrain from bidding on each other’s trademark terms in search-based keyword advertising and to employ negative keywords to prevent its ads from displaying when a consumer searches online for the other party’s trademarks. The agreements did not limit a party’s ability to advertise through other media and did not prohibit parties from bidding on generic keywords such as “contact lens,” so long as they employ negative keywords. The FTC majority found these agreements “unusual.”125 While trademark litigation typically seeks to bar the use on the infringer’s labels, ads, or other promotional materials of the plaintiff’s trademark or a similar mark in a way likely to confuse consumers, the settlement agreements at issue shut off an entire channel of advertising which is a relevant method for marketing contact lenses online to obtain new customers. Indeed, many online retailers devote most of their advertising expenditures to search advertising.126 Furthermore, 1–800 Contacts is a pure-play online retailer, which sells only online and does not have brick-and-mortar locations. Pure-play online sellers accounted for 17 percent of contact lens sales in 2015 and 1–800 Contacts accounted for more than 60 percent of the online contact lens market and more than four times the sales of the second-largest online retailer. Evaluating the challenged agreements as horizontal restraints aimed at restricting the information provided by advertising to consumers, the majority concluded that they were “inherently suspect”, thereby applying a truncated rule of reason analysis to evaluate them.127 An inherently suspect analysis and a per se analysis are close neighbors. If the conduct at issue is inherently suspect due to its likely tendency to suppress competition, a quick-look scrutiny is sufficient to condemn the restraint, unless the defendant can articulate legitimate justifications.128 The FTC rejected both the justifications articulated by 1–800 Contacts. On the one hand, the avoidance of litigation costs through settlement would have procompetitive effects only insofar as cost savings are passed through to consumers by lowering prices or improving service quality.129 On the other hand, although trademark protection can be a legitimate justification and an advertising restraint may conceivably avoid consumer confusion, “[c]onfusion must be probable, not merely possible.”130 The majority acknowledged that trademark litigations underlying the settlement agreements were not sham and that 1–800 Contacts had a brand identity that it wished to preserve.131 However, 1–800 Contacts never had judicial support for its claims: “apart from a single district court summary judgment decision from over ten years ago, no court has found bidding on trademark keywords to constitute trademark infringement, absent some additional factor, such as a misleading use of the trademark in the ad text that confuses consumers as to the advertisement’s source, sponsorship, or affiliation.”132 The justification for including negative keywords in the agreements is even weaker, since “no court has ever found that bidding on a generic keyword (like “contacts”), which may be broad or phrase-matched by the search engine to a trademark search, is even a “use.”133 In summary, 1–800 Contacts’ trademark claims appeared pretextual and the restrictions included in the agreements unreasonably necessary to protect its trademarks.134 Ultimately, the advertising restrictions had the anticompetitive effects of forbidding truthful advertising and increasing online contact lens prices.135 The challenged agreements reduced the number of competitor ads and increased sales for 1–800 Contacts while reducing the sales for its rivals. Moreover, as a consequence of 1–800 Contacts shielding itself from competitive pressure by preventing consumers from obtaining information that would enable comparison shopping, prices charged by 1–800 Contacts were on average higher than those of its online competitors.136 Finally, the agreements also harmed search engines by unreasonably restraining price competition in certain search advertising auctions and impairing the quality of service provided to consumers.137 Commissioner Phillips dissented, vigorously.138 Holding that the challenges settlements did not look suspiciously like per se illegal conducts and that the majority failed to give appropriate credit to 1–800 Contacts’ procompetitive justifications, Phillips’ opinion focused on the relevance of the IP interest at stake. At the outset Phillips underlined the massive endeavor undertaken by 1–800 Contacts to generate brand awareness and to pioneer the online contact lens business: “That is the type of conduct that antitrust and trademark law should, and does, encourage,” and it “created precisely the value that other retailers sought to derive by bidding on 1–800 Contacts’ trademarked terms.”139 Furthermore, the context surrounding the challenged settlements should not be disregarded. In the wake of the Rescuecom decision, uncertainty and legal risks arose, increasing the incentive for alleged infringers to settle, given the fact-specific nature of the inquiry into trademark confusion.140 Indeed, “[n]o one, even today, contends that the trademark claims asserted by 1–800 Contracts were shams.”141 Against this background, according to Phillips, the settlements at issue were appropriately tailored to achieve their goal of preventing trademark infringement while balancing the need to permit noninfringing advertising.142 Indeed, the agreements restricted only a subset of advertising, impacting only those consumers who search specifically for 1–800 Contacts’ trademarks, that is the searches through which users are most likely attempting to reach the 1–800 Contacts website.143 Therefore, the facts of the case do not fit into the jurisprudence on advertising restraints mentioned by the majority, since the latter regards cases involving complete advertising bans or limitations on the content that advertisements could contain.144 In addition, none of these cases implicates IPRs. In contrast, in the assessment of the challenged settlements, the majority did not recognize how trademark protections and the enforcement of trademarks encourage brand investment and promote competition. “In fact, the majority dismiss[ed] the benefits of trademark policy entirely.”145 Notably, Commissioner Phillips warned that the main argument brought against 1–800 Contacts’ justifications was that its trademark infringement claims were weak. However, at most, the case has shown that the legal status of using trademarked terms as keywords in paid search advertising was uncertain: “Predicating antitrust liability on an ex post judgement about the strength of intellectual property infringement claims—or ignoring the context of their protection entirely—not only will reduce clarity in the law, but also threatens to chill the procompetitive investment that is one of the hallmarks of trademark law.”146 Finally, the economic analysis endorsed by the majority was also insufficient to show direct price effects of the advertising restriction.147 Indeed, the majority failed to prove that the challenged settlements caused the price differential, that is, that consumers paid higher prices due to the challenged settlement agreements. Rather, its findings rested solely on the fact that 1–800 Contacts’ prices were higher than rivals’ prices.148 D. Are Trademark Settlements Different? The debate within the FTC over trademark settlements shows how challenging it is to strike a proper balance between the antitrust goal of avoiding any restriction to truthful advertising that makes consumer comparisons more difficult and safeguarding the incentives related to trademark protection, aimed at rewarding firms that pioneer markets and generate brand awareness. Antitrust enforcers have already faced IP settlement dilemmas investigating agreements whose suspect aim was to keep rivals off the market and reduce the risks of competition, rather than solving disputes. Indeed, EU and US courts and authorities have adopted a strict approach to reverse payment patent settlements (also known as pay-for-restriction or pay-for-delay agreements) between originators and generic companies in the pharmaceutical industry, containing an obligation on the latter not to use the invention covered by the patent during the period of its protection and/or an obligation not to challenge the patent concerned in court.149 Notably, the scope of the patent approach endorsed by several courts—according to which the settlement agreements should be considered lawful as long as they do not exceed the exclusionary scope of the patent in suit, the patent litigation is neither a sham nor otherwise baseless, and the patent is not obtained by fraud—has been rejected by the US Supreme Court.150 Nonetheless, stating that patent settlements must be judged under rule of reason analysis, the Court also rejected both a per se illegality151 and a quick look approach based on a presumption of illegality.152 Since these agreements provide for the patentee to pay the alleged infringer, rather than the opposite, with the aim of delaying its market entry, their anticompetitive effect is inferred from a large and unjustified value transfer to the generic company, without litigating on the validity of the patent at issue. Namely, “the size of the unexplained reverse payment can provide a workable surrogate for a patent’s weakness, all without forcing a court to conduct a detailed exploration of the validity of the patent itself.” 153 Hence, according to the so-called pay-for-delay inference, the US Supreme Court revealed its preference for determining patent strength by examining the payment rather than the patent.154 The European Technology Transfer Guidelines endorsed a similar approach, stating that it will be particularly attentive to the risk of market sharing if there is a significant value transfer from the licensor to the licensee.155 In the recent Paroxetine ruling, the Court of Justice, although considered that the fact that patent settlements involve transfers of value is not sufficient ground to classify them as a restriction by object, affirmed the view that such agreements should be qualified as restrictions by object when they involve a value transfer representing a significant inducement for the alleged infringer to abandon its competitive efforts and stay off the market.156 To meet the inducement criterion, it is not necessary to demonstrate that the transfer of value exceeds the profits expected by the generic manufacturer: all that matters is that the transfer is sufficiently beneficial to encourage the manufacturer of generic medicines to refrain from entering the market and not to compete on the merits with the manufacturer of originator medicines. In sum, by considering the probabilistic nature of IPRs, one suggestion would be to extend to keyword advertising settlements the approach adopted for reverse payment patent settlements, thus evaluating the significance of the restraint at stake as a surrogate for weakness of the trademark claim in question. However, several reasons suggest that the criteria elaborated for reverse payment patent settlements are not a good fit for analyzing keyword advertising settlements.157 First, the significant differences in terms of economic rationale and functions between patent and trademark protection cannot be overlooked. Besides, although all property rights share a certain degree of uncertainty, the uncertainty associated with patents is especially striking and fundamental to an understanding of their effects on innovation and competition.158 Indeed, in keyword advertising settlements the validity of the trademark is not disputed, whereas reverse payment settlements are problematic because it is apparent that the generic company’s agreement is not based on its recognition of the validity of the patent. Furthermore, the existence of a reverse payment is doubly suspect because a valid IPR, in principle, allows a transfer of value to its holder, rather than vice versa. Hence, the reverse value transfer cast doubts as to whether the settlement is based on the recognition of the validity of the patent in question. Instead, in trademark settlements, the value transfer goes in the usual direction, that is, from the defendant to the plaintiff. Moreover, in reverse payment settlements, the size of the transfer plays a pivotal role since it is considered an indicator of the strength of the patent and of the fact that the originator company is not convinced of its chances of succeeding in the event of litigation. Although certain trademark settlements contain monetary consideration, the value transfer is not comparable with the amount at issue in pharmaceutical patent settlements. Finally, reverse payment settlements are natural by-products of pharma regulation since they reflect the features peculiar to the pharmaceutical industry, which presents a perfect storm for regulatory gaming.159 Indeed, reverse payment patent settlements are investigated due to their effects of delaying the launch of new products, while trademark settlements involve an advertising restraint. More precisely, the latter may include different types of restraints (wide nonbrand bidding, negative nonbrand bidding, negative matching), whose potential impact on competition varies considerably. For these reasons, a case-by-case approach appears better suited to tackle the peculiar features of keyword advertising settlements. However, the comparison between reverse payment patent settlements and keyword advertising trademark settlements is useful to highlight similar critical aspects and concerns regarding the boundaries of antitrust enforcement against the scope of IP protection. In Actavis Justices Roberts, Scalia and Thomas complained that the majority conducted an antitrust analysis ignoring the presence of a patent: “But a patent holder acting within the scope of its patent does not engage in any unlawful anticompetitive behavior; it is simply exercising the monopoly rights granted to it by the Government. Its behavior would be unlawful only if its patent were invalid or not infringed. And the scope of the patent—that is, what rights are conferred by the patent—should be determined by reference to patent law.”160 Furthermore, “the majority would impose antitrust liability based on the parties’ subjective uncertainty about [the] legal conclusion [on the patent validity or infringement].”161 In the same vein, in 1–800 Contacts Commissioner Phillips argued that the challenged settlements resolved legitimate intellectual property infringement claims, accusing the majority of dismissing the benefits of trademark policy and disregarding the legal uncertainty surrounding the settlements.162 Following this line of reasoning, the only way to safeguard the balance between antitrust enforcement and IP protection is to assess the scope of the trademark. A trademark holder must act within the scope of the trademark and its actions are subject to antitrust scrutiny if they go beyond the power conferred by the trademark law. Accordingly, the allegation of misusing trademark law to advance an anticompetitive agenda is grounded only if the trademark holder acted outside the scope of the IP protection. However, the legitimate concerns about the progressive expansion of the scope of trademark rights, which is no longer limited to the traditional and essential function of indicating origin should not be disregarded. Therefore, to avoid the risk of defining the interface between trademark law and antitrust law according to a mere formalistic approach, in keyword advertising cases a balanced approach requires a proper consideration of the expansionist trend to trademark protection. In this regard, valuable insights are provided by EU trademark case law. Indeed, pursuant to Google France and Interflora, in the keyword advertising scenario, only the origin function has relevance. To safeguard the freedom of competition, trademark holders are not allowed to leverage the advertising and the investment functions in keyword advertising cases. “[I]n conditions of fair competition that respect the trademark’s function as an indication of origin”, it cannot be accepted that the holder of a trademark may prevent a competitor from using an identical sign, if the only consequence of that use is to oblige the proprietor of that trademark to intensify its advertising and to adapt its efforts to acquire or preserve a reputation capable of attracting consumers and retaining their loyalty.163 The fact that that use may prompt some consumers to switch from goods or services bearing that trademark cannot be successfully relied on by the proprietor of the mark. Moreover, free-riding claims are circumscribed to the case in which advertisers offer for sale goods which are imitations of the goods of the proprietor of reputed trademarks.164 By contrast, the holder of a reputed trademark is not entitled to prevent advertisements which put forward alternative goods or services. Finally, with regard to the origin function, as in the US, the focus is on the labeling and the appearance of the advertisements.165 There is an adverse effect on the origin function if the advertisement does not enable users to ascertain whether the goods or services originate from the owner of the trademark or from a third party, and if the advertisement suggests the existence of an economic link between the third party and the holder of the trademark. By restoring the primacy of the confusion-based rationale of trademark protection, EU trademark case law (rather than antitrust case law) delimits the scope of the trademark protection in the keyword advertising scenario reducing the control of trademark holders over their signs to ensure fair competition. In the light of the CJEU’s principles, it may be predicted that trademark settlement agreements limiting search-based keyword advertising will be considered within the scope of trademark protection and compatible with EU antitrust rules only if the underlying litigation relates to an infringement by confusion or dilution (by blurring). Indeed, in this case the settlement agreement does not give the trademark holder more rights than it would have been able to derive by virtue of the trademark regime. This approach is supported by the CJEU’s antitrust decision in BAT.166 Although the settlement at stake did not involve keyword advertising but a trademark delimitation agreement, the CJEU clearly held that addressing a trademark dispute is not in itself restrictive of competition if there is a genuine risk of confusion between the parties.167 By the same token, in the European Technology Transfer Guidelines, the Commission has acknowledged that IPRs settlements are “in principle a legitimate way to find mutually acceptable compromise to a bona fine legal disagreement” and they are capable of generating welfare enhancing benefits.168 Namely, parties may prefer to discontinue a dispute which it proves to be too costly, time-consuming or uncertain as regards its outcome; courts and administrative authorities may save effort and resources in deciding on the matter. Therefore, a trademark settlement agreement limiting search-based keyword advertising in a dispute over likelihood of confusion should not be regarded in principle as restrictive of competition by object since it falls within the scope of the trademark and it is capable of generating procompetitive gains. This does not nevertheless rule out the possibility that in certain circumstances a settlement agreement remaining within the scope of the trademark exclusive right might have the effect of restricting competition.169 It is for courts and antitrust authorities to establish its restrictive impact through a case-by-case assessment of the context and the provisions of the agreement. In this regard, as suggested by the British and the Dutch antitrust authority inquiries, it would be relevant to analyze the impact on consumers of the different types of advertising restrictions. In principle, wide nonbrand bidding and negative matching agreements are likely to have a bigger impact on click-through and conversion rates than narrow nonbrand bidding. Furthermore, by affecting all searches that include the brand owner’s name irrespective of the context, negative matching agreements are likely to have a bigger impact than (narrow and wide) nonbrand bidding. If the dispute relates to a risk of confusion, to consider the trademark settlement anticompetitive by its very nature it would be necessary to demonstrate that the underlying litigation is sham or vexatious.170 Indeed, abusive litigation with the aim of excluding rivals represents a typical form of nonprice predation. Further, if the trademark holder claims an infringement by free-riding, an eventual settlement should be considered anticompetitive by object, since in keyword advertising cases only the origin function is worthy of protection. However, the scope of the trademark approach leads to a different outcome in the US. Indeed, unlike EU case law, US courts did not introduce any limits to trademark protection in keyword advertising cases. In particular, despite it having been argued that US trademark law lacks a clear anti-free-riding principle,171 no court has suggested limiting its analysis of infringement claims on trademarked keywords to cases where only the origin function is adversely affected.172 As a matter of policy, the European approach is better suited to ensuring a balance between the interests at stake by circumscribing the scope of trademark protection in keyword advertising cases to confine its potential anticompetitive effects. It is worth noting that the principles of the limited trademark protection in the keyword advertising scenario have been set by the CJEU in trademark litigations, rather than in antitrust investigations. Therefore, the scope of the trademark has been determined by reference to trademark law. Moreover, this approach has been endorsed and applied by the European Commission in the Guess antitrust decision. Indeed, the Commission assessed the restrictions at issue referring to the Google France and Interflora rulings, thereby evaluating the online search advertising restriction as incompatible with antitrust provisions because there was no risk of confusion as to the origin of the products. An alternative to the scope of trademark approach is represented by the legal standard set by EU and US courts and authorities for reverse patent settlements. However, as already mentioned, crucial differences advise against adopting the same criteria for the antitrust assessment of keyword advertising settlements. Namely, inter alia, the two main conditions which make the former suspicious (i.e. the disproportionate amount of the value transfer and its unusual direction) are missing for the latter. Moreover, the study conducted by the UK antitrust authority suggests evaluating the different types of advertising restrictions at issue (narrow nonbrand bidding agreements, wide nonbrand bidding agreements, and negative matching agreements) and gathering more evidence on the impact of trademark settlements on consumers. Hence, it is questionable to consider trademark settlements as inherently suspect, thereby applying to them the same or an even stricter standard than reverse patent settlements. To support a case-by-case analysis it is also worth noting that the agreements under scrutiny in the Netherlands and in the US in the hotel sector show a context which is apparently different from the one surrounding the settlements evaluated by the FTC in 1–800 Contacts and by the UK CMA. Indeed, in these cases the trademark dispute seems to be absent: hotel chains negotiated nonbrand bidding agreements with OTAs, without any underlying litigation concerning a trademark infringement. Therefore, the risk of conspiracy is more concrete since these agreements appear to be aimed only at shielding online sales channels against competition from other hotels present on the OTAs. IV. CONCLUDING REMARKS: ANTITRUST, TRADEMARKS, AND THE VALUE OF INFORMATION In a world driven by rapidly evolving digital technology, the role of the internet as a game changer in global trade is all the more evident. Platforms, auction sites, price comparison search engines and online shops are all means which are widely known and used by consumers for purchasing goods and services on a daily basis. Hence, the Internet constitutes a powerful tool to provide information to consumers and promote competition among firms. In particular, search engines have reduced the cost for companies of reaching consumers. Therefore, policy makers and antitrust authorities are inclined to protect advertising’s information function and are legitimately concerned about online advertising restrictions.173 Against this background, it is suspected that trademark law could possibly jeopardize the internet’s potential as an information resource and a catalyst for competition.174 Indeed, trademark holders try to prevent the selection of their brand names as keywords by third parties and the display by search engine providers of sponsored links in response to those keywords. Trademark holders complain that the use of a competitor’s mark to trigger the display of one’s own advertising as a sponsored link might have an adverse effect on the essential functions of the trademark. However, by introducing unreasonably overbroad restrictions and/or raising pretextual claims, trademark holders may aim at insulating consumers from becoming aware of their rivals, rather than just protecting their IPRs. Indeed, outside the cases of counterfeiting, using brand names as part of keyword advertising increases the intensity of competition for customers, making it easier for firms to find consumers who would be interested in learning about their product.175 Restrictions on the use of trademarks as keywords may frustrate consumer ability to compare prices and competitor offerings, thus consumers may eventually be harmed as a result of higher prices and poorer products. For these reasons, keyword advertising represents the new frontier of the IP-antitrust interface. On the one side, the progressive expansion of the scope of trademark protection has raised significant concerns, which are further heightened in the digital scenario. Whereas, according to the traditional economic rationale, trademarks essentially serve the purpose of indicating the commercial origin of goods and services offered in the marketplace, the recognition of other functions worthy of protection, such as those of communication, investment and advertising, has allowed trademark holders to exert control over the use of their marks across all markets, regardless of the risk of confusion.176 As a result, the equilibrium between trademark protection and freedom of competition may be impaired. On the other side, an overbroad enforcement of antitrust rules risks undermining incentives and benefits related to IP protection. In this regard, tools and doctrines crafted by antitrust authorities to manage the intersection with IP, such as the essential facility doctrine, the willing licensee test for FRAND-encumbered SEPs, and the special treatment for pay-for-delay settlements, have been hotly debated. Furthermore, all these antitrust interventions have been characterized as exceptional in comparison with the proclaimed finalistic convergence and complementarity of the two disciplines. Keywords advertising restrictions set the scene for a new IP-antitrust clash and for the reshaping of exceptional circumstances which justify antitrust intervention. To avoid this risk and ensure an actual convergence between IP protection and antitrust goals, courts need to strike a proper balance that recognizes both the contribution that informative advertising makes to competition and the need to forbid misleading and deceptive strategies that adversely affect the origin function of trademarks.177 By restoring the primacy of the confusion-based rationale of trademark protection in the keyword advertising scenario, the CJEU’s Google France and Interflora rulings define the most appropriate path which should also be followed in assessing brand-bidding restrictions. Indeed no one questions the essential trademark’s function as an indication of origin. Concerns have been raised about the expansion of trademark rights to encompass other functions notably those of communication investment or advertising. Furthermore by limiting the scope of trademark protection in keyword advertising cases the CJEU explicitly refers to the need to safeguard the freedom of competition namely to ensure conditions of fair competition that respect the trademark’s function as an indication of origin. This approach has been endorsed by the European Commission in the Guess decision related to brand-bidding restrictions imposed within selective distribution networks to reduce competition from independent online retailers. The same approach should be applied to trademark settlement agreements limiting search-based keyword advertising in a dispute over likelihood of confusion. These agreements, as well as other agreements with brand-bidding restrictions, should not be regarded as per se (or by object) restrictive of competition as long as they address a risk of confusion as to the origin of products and services, since they fall within the scope of the trademark (i.e. they do not provide trademark holders more rights than they would have been able to derive by virtue of the trademark regime) and are capable of generating procompetitive gains. It is for courts and antitrust authorities to establish their restrictive effects through a case-by-case assessment of the market conditions and the provisions of the agreements, notably looking at the impact of the different types of advertising restrictions (i.e. narrow and wide nonbrand bidding, and negative matching agreements). This appears the only way to pursue the much invoked balanced approach between IP and antitrust. Footnotes 1 " For a recent overview, see H. Hovenkamp, Intellectual Property and Competition, in Research Handbook on the Economics of Intellectual Property (P. Menell & B. Depoorter, eds., Edward Elgar 2019), Vol. 1, p. 231. 2 " H. Hovenkamp, M.D. Janis, M.A. Lemley, and C.R. Leslie, IP and Antitrust, Wolters Kluwer, 2013, vol. 1, pp. 1–16. 3 " Whilst in Verizon Communications v. Law Offices of Curtis V. Trinko, 540 U.S. 398 (2004) the US Supreme Court has repudiated the essential facility doctrine, the doctrine has gained huge success in the European scenario, its application having been extended to IP rights (see CJEU, 6 April 1995, joined Cases C-241/91 P and 242/91 P, Radio Telefis Eireann (RTE) and Independent Television Publications Ltd (ITP) v. Commission; CJEU, 29 April 2004, Case C-418/01, IMS Health GmbH & Co. OHG v. NDC Health GmbH & Co. KG.; General Court, 17 September 2007, Case T-201/04, Microsoft Corp. v. Commission). 4 " In Brulotte v. Thys Co., 379 U.S. 29 (1964) and in Kimble v. Marvel Entertainment, LLC, 576 U.S. __ (2015) the US Supreme Court has stated that an agreement allowing a patent owner to collect royalty payments after a patent’s expiration is unlawful per se. Instead, in the EU the parties can normally agree to extend royalty obligations beyond the period of validity of the licensed intellectual property rights without falling foul of Article 101(1) TFEU, according to paragraph 187 of the Guidelines on the application of Article 101 TFEU to technology transfer agreements (OJ C 89/03 (2014)). 5 " Since Leegin Creative Leather Products v. PSKS, 551 U.S. 877 (2007) minimum resale price maintenance is judged in the US by the rule of reason, whereas in the EU it is considered a hardcore restriction according to Article 4 of the Regulation No. 330/2010 (Vertical Block Exemption Regulation, VBER; OJ L 102/1 (2010). 6 " See, e.g., J. Lerner, The Empirical Impact of Intellectual Property Rights on Innovation: Puzzles and Clues, 99 American Economic Review 343 (2009). 7 " See, e.g., M. Boldrin & D.K. Levine, The Case Against Patents, 27 Journal of Economic Perspectives 3 (2013). 8 " J. Bessen & M. Meurer, Patent Failure: How Judges, Bureaucrats, and Lawyers Put Innovation at Risk, Princeton University Press, 2008; A.B. Jaffe & J. Lerner, Innovation and Its Discontents: How Our Broken Patent System is Endangering Innovation and Progress, And What to Do About It, Princeton University Press, 2007. See also M.A. Lemley & C. Shapiro, Probabilistic Patents, 19 Journal of Economic Perspectives 75 (2005). 9 " W.M. Landes & R.A. Posner, Trademark Law: An Economic Perspective, 30 Journal of Law & Economics 265 (1987). 10 " A. Kur, Trademark Functions in European Union Law, Max Planck Institute for Innovation and Competition Research Paper No. 19–06, 2, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3425839 (accessed February 5, 2020). 11 " B. Beebe & J.C. Fromer, Are We Running Out of Trademarks? An Empirical Study of Trademark Depletion and Congestion, 131 Harvard Law Review 945 (2018). 12 " However, see L.L. Ouellette, Does Running Out of (Some) Trademarks Matter? 131 Harvard Law Review 116 (2018), contending there is not yet any concrete evidence that trademark depletion and congestion impose any significant harms for either firms or consumers. 13 " See e.g. M.A. Lemley & M.P. McKenna, Owning Mark(et)s, 109 Michigan Law Review 37 (2010). On the European side, see Kur, supra note 10, analyzing the steps undertaken towards the emancipation of trademark rights from their original raison d’être. 14 " B. Beebe & C. Scott Hemphill, The Scope of Strong Marks: Should Trademark Law Protect the Strong More than the Weak? 92 New York University Law Review 1339 (2017). 15 " It is worth distinguishing between search advertising and display advertising as two main forms of online advertising. While the former refers to advertisements provided in response to users’ search queries, the latter refers to visual-based advertisements, including banner and video advertisements, that appear on social media, video, as well as news media platforms alongside the content a user is interested in. Search advertising can be further divided into general and specialized, where the former is the advertising that appears on the search results of general search engines (e.g. Google and Bing), the latter is the advertising that appears alongside the search results of search engines that perform more specialised functions (e.g. the search engines on the platforms of Amazon or Expedia). Given the evidence of a user’s purchasing intention provided by her choice of keywords, search advertising is often favored over display advertising for direct response campaigns focused on conversion. Indeed, the user entering the search query is more likely to perform the action desired by the advertiser purchasing the advertised product or contacting the supplier of the product. Conversely, display advertising is often used for promoting general brand awareness. For an analysis of the competitive landscape of digital advertising markets, see Australian Competition and Consumer Commission, Digital platforms inquiry, 2019, https://www.accc.gov.au/focus-areas/inquiries-ongoing/digital-platforms-inquiry (accessed February 5, 2020); UK Competition and Markets Authority, Online Platforms and Digital Advertising, 2019, https://www.gov.uk/cma-cases/online-platforms-and-digital-advertising-market-study (accessed February 5, 2020). 16 " Outcomes are determined by second-price auction, hence an advertiser’s bid does not directly determine how much it pays, but affects its ranking in the auction and consequently in the search results. For the game-theoretic analysis of keyword auctions, see B. Edelman, M. Ostrovsky, and M. Schwarz, Internet Advertising and the Generalized Second-Price Auction: Selling Billions of Dollars Worth of Keywords, 97 American Economic Review 242 (2007); H.R. Varian, Position Auctions, 25 Industrial Journal of International Organization 1163 (2007). Advertisers may delegate some of this decision-making to the search engine by using automated bidding tools which automatically adjust an advertiser’s bid to optimize performance within their budget. See UK Competition and Markets Authority, supra note 15, 163–164, arguing that the complexity of the auction mechanism implies that buying search advertising is a data-driven process that requires advertisers to make many granular decisions (which keywords to bid for, how flexible to be with matching to those keywords, which consumers to target, and how much to bid for each keyword) attempting to optimize their expenditure on search advertising continuously over time, by setting their bids to allocate their expenditure to the keywords, search engines and target audiences where their return on investment is greatest. In addition, since search advertising prices are set by auction rather than being set directly by the search engine, this raises a relevant antitrust question as to whether a search platform with market power has the ability to influence the prices that advertisers pay: see UK Competition and Markets Authority, supra note 15, 172–175, noting that search engines have several levers through which they can influence advertising prices and conversion rates. Indeed, search engines determine the maximum number of ads that can be shown per search query, how these ads are presented, the way in which relevance is assessed, the level of reserve prices, and the way in which matching algorithms work. 17 " European Commission, Staff Working Document—Accompanying the Document Report from the Commission to the Council and the European Parliament—Final Report on the E-commerce Sector Inquiry, SWD (2017) 154 final, para. 632. 18 " UK Competition and Markets Authority, Digital comparison tools market study, 2017, 62–64, https://www.gov.uk/cma-cases/digital-comparison-tools-market-study (accessed February 5, 2020). 19 " Netherlands Authority for Consumers & Markets, Price effects of non-brand bidding agreements in the Dutch hotel sector, 2019, https://www.acm.nl/sites/default/files/documents/2019-06/working-paper-acm-price-effects-of-search-advertisement-restrictions.pdf (accessed February 5, 2020). 20 " In the Matter of 1–800 Contacts, Inc., 2018 WL 6078349 (FTC Docket No. 9372, November 7, 2018). 21 " E. Goldman, Brand Spillovers, 22 Harvard Journal of Law & Technology 381 (2009); S.L. Dogan & M.A. Lemley, Trademarks and Consumer Search Costs on the Internet, 41 Houston Law Review 777 (2004). 22 " Opinion of Advocate General Jääskinen, March 24, 2011, Case C-323/09, Interflora Inc. and Interflora British Unit v. Marks & Spencer plc and Flowers Direct Online Limited, para. 9. 23 " CJEU, 23 March 2010, joined Cases C-236/08, C-237/08 and C-238/08, Google France SARL v. Louis Vuitton Mallettier SA et al. 24 " CJEU, 22 March 2011, Case C-323/09, Interflora Inc. v. Marks & Spencer plc and Flowers Direct Online Ltd. 25 " Opinion of Advocate General Maduro, 22 September 2009, joined Cases C-236/08, C-237/08 and C-238/08, paras. 48–50. 26 " Id., para. 49. 27 " Id., paras. 121–122. 28 " CJEU, June 18, 2009, Case C-487/07, L’Oréal SA and others v. Bellure NV and others. 29 " Id., para. 49. 30 " Google France, supra note 23, paras. 50–59. The Court rejected the proposal of the Advocate General, according to which the selection of keywords should also be classified as private use by the advertiser (see Opinion of Advocate General Maduro, supra note 25, para. 150). 31 " Google France, supra note 23, para. 57. 32 " Id., para. 77. 33 " Id., paras. 91–92. 34 " Id., paras. 83–84. On the relevance of the design of the advertisement, for a recent national application of this principle, see German Federal Supreme Court (Bundesgerichtshof), 25 July 2019, Case No. I ZR 29/18, Ortlieb II, stating that Amazon’s advertising on Google using the branded term “Ortlieb” was a trademark infringement because the Google ad led to a list of results on Amazon that included competitors’ products. The Court found that the use of the sign Ortlieb could adversely affect the origin function of the trademark in suit because the relevant public expected to be showed offers from Ortlieb when clicking on the ads. Notably, the design of the ads gave the public no reason to expect to be presented with a list of offers that contained, besides Ortlieb products, corresponding products from other manufacturers without specially identifying them. This expectation is reinforced by the URLs shown in the ads, which also contained the term Ortlieb. Hence, by clicking the ad the public expected to be led to a list of Amazon’s offers that fulfilled the criteria given, that is, the respective products of the brand Ortlieb. Conversely, the fact that a branded search on Amazon resulted in a variety of manufacturer’s products does not necessarily imply a finding of trademark infringement, because an average Internet user conducting a search with a brand name on a digital marketplace does not expect to find only offers of the trademark holder, and more in general she is also aware that third parties regularly place paid ads with the operator of an Internet search engine (Federal Supreme Court, February 15, 2018, Case No. I ZR 138/16, Ortlieb I). 35 " Id., paras. 89–90. 36 " Id., paras. 93 and 95. 37 " Id., para. 97. 38 " Interflora, supra note 24, para. 57. 39 " Google France, supra note 23, para. 117. 40 " Id., para. 118. 41 " Id., para. 120. 42 " CJEU, March 25, 2010, Case C-278/08, Die BergSpechte Outdoor Reisen und Alpinschule Edi Koblmüller GmbH v. Günter Guni and trekking.at Reisen GmbH; 26 March 2010, Case C-91/09, Eis.de GmbH v. BBY Vertriebsgesellschaft mbH; 8 July 2010, Case C-558/08, Portakabin Ltd v. Primakabin BV; July 12, 2011, Case C-324/09, L’Oréal SA and others v. eBay International AG and others; Interflora, supra note 24. 43 " L’Oréal, supra note 42, paras. 84–87 and 98–103. 44 " Interflora, supra note 24, para. 60. 45 " Id., para. 61. 46 " Id., paras. 62–63. 47 " Id., para. 64. 48 " Opinion of Advocate General Jääskinen, supra note 22, para. 45. 49 " Interflora, supra note 24, paras. 80–81. 50 " Id., para. 89. 51 " Id., para. 86. 52 " Id., para. 90. 53 " Id., para. 91. 54 " CJEU, 12 November 2002, Case C-206/01, Arsenal Football Club plc v. Matthew Reed, paras. 42 and 51. 55 " According to Kur, supra note 10, 5–6, the trigger for what became the CJEU’s functions jurisprudence lay hidden in the text of the Directive 89/104 and the Regulation 40/94, precisely in the double identity clause. In a similar way, see CJEU, Interflora, supra note 24, para. 39, highlighting the use of the words ‘in particular’ in the tenth recital to Directive 89/104 and in the seventh recital to Regulation No 40/94. Differently, M.R.F. Senftleben, Adapting EU trademark law to new technologies: back to basics?, in Constructing European Intellectual Property (C. Geiger, ed.), Edward Elgar, 2013, 137. 56 " L’Oréal/Bellure, supra note 28, para. 58. 57 " Senftleben, supra note 55, 137. 58 " Opinion of Advocate General Maduro, supra note 25, paras. 101–103. 59 " Id., paras. 112. 60 " Id., paras. 111. 61 " Opinion of Advocate General Jääskinen, supra note 22, para. 94. 62 " Id., para. 94. 63 " A “welcome surprise” for Senftleben, supra note 55, 160. See also Kur, supra note 10, 12–13. 64 " Google France, supra note 23, para. 90. 65 " Id., paras. 93, 95 and 97. 66 " Interflora, supra note 24, para. 57. 67 " Id., para. 64. 68 " Rosetta Stone Ltd. v. Google, Inc., 676 F.3d 144, 163–165 (4th Cir. 2012). 69 " Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844, 854 (1982). 70 " Tiffany (NJ) Inc. v. eBay Inc., 600 F.3d 93 (2nd Cir. 2010). 71 " Playboy Enterprises, Inc. v. Netscape Comm. Corp., 354 F.3d 1020 (9th Cir. 2004). For a critical analysis see, e.g., Dogan & Lemley, supra note 20, 807, arguing that selling advertising space based on a keyword that is also a trademark does not use that trademark as a brand because internet intermediaries are not selling any product or service using those terms as an identifier. 72 " Rescuecom Corp. v. Google Inc., 562 F.3d 123 (2nd Cir. 2009). 73 " 1–800 Contacts, Inc. v. WhenU.Com, Inc., 414 F.3d 400 (2nd Cir. 2005). 74 " S & L Vitamins, Inc. v. Australian Gold, Inc., 521 F. Supp. 2d 188, 199–202 (E.D.N.Y. 2007). 75 " Merck & Co., Inc. v. Mediplan Health Consulting, Inc., 425 F. Supp. 2d 402, 415 (S.D.N.Y. 2006). 76 " Rescuecom, supra note 72, 129. 77 " Id., 127. 78 " Dogan & Lemley, supra note 21, 782. See also E. Goldman, Deregulating Relevancy in Internet Trademark Law, 54 Emory Law Journal 507, 509 (2005), arguing that initial interest confusion doctrine is predicated on multiple mistaken and empirically unsupported assumptions about searcher behavior; and M.A. Lemley & M. McKenna, Irrelevant Confusion, 62 Stanford Law Review 413, 450 (2010), proposing that “the law should require that trademark owners claiming infringement based on confusion regarding anything other than source or responsibility for quality must demonstrate the materiality of that confusion to consumer purchasing decisions.” 79 " Playboy Enters., supra note 71, 1025. See Groeneveld Transport Efficiency, Inc. v. Lubecore International, Inc., 730 F.3d 494, 519 (6th Cir. 2013) rejecting the initial interest confusion doctrine and stating that “what appears to concern Groeneveld is not so much initial-interest confusion, but initial interest, period. Groeneveld, in other words, simply does not want its customers to become interested in Lubecore as a potential competitor and possibly switch over. We cannot ascribe any other interpretation to Groeneveld’s rather startling claim that evidence of diverted sales and declining revenues, which are the normal signs of a market opening up to competition, create “a reasonable inference of confusion and its likelihood.” 80 " Promatek Indus., Ltd. v. Equitrac Corp., 300 F.3d 808, 812 (7th Cir. 2002). Even if the initial interest confusion is an expression that derives from US trademark law, the principle affirmed by the CJEU in Google France, supra note 23, para. 90, may be regarded as the acknowledgment of the same protection in the EU. Indeed, according to Google France, internet users must be able to recognize the origin of the goods or services on the basis of the advertisement, without having to visit the advertiser’s website. See also UK High Court of Justice, Och-Ziff Management Europe Ltd. v. OCH Capital LLP, [2010] EWHC, 2599 (Ch), paras. 79–101, arguing that initial interest confusion is actionable under EU trademark rules. 81 " See, e.g., AMF Inc. v. Sleekcraft Boats, 559 F.2d 314, 348–49 (9th Cir. 1979). The eight factors of the Sleekcraft test are: strength of the mark; proximity of the goods; similarity of the marks; evidence of actual confusion; marketing channels used; type of goods and the degree of care likely to be exercised by the purchaser; defendant’s intent in selecting the mark; and likelihood of expansion of the product lines. 82 " Brookfield Communications, Inc. v. West Coast Entertainment Corp., 174 F.3d 1036, 1054 (9th Cir. 1999). 83 " Network Automation, Inc. v. Advanced Systems Concepts, Inc., 638 F.3d 1137, 1153 (9th Cir. 2011). 84 " Id., 1152. 85 " Rosetta Stone, supra note 68. 86 " 1–800 Contacts, Inc. v.Lens.com, Inc., 722 F.3d 1229 (10th Cir. 2013). 87 " Id., 1245. 88 " Id., 1245. 89 " Id., 1250. 90 " Multi Time Machine, Inc. v.Amazon.com, Inc., 792 F.3d 1070 (9th Cir. 2015). In the same vein, see German Federal Supreme Court, Ortlieb I, para. 55, supra note 34. 91 " Alzheimer’s Disease and Related Disorders Association, Inc. v. Alzheimer’s Foundation of America, Inc., 307 F. Supp. 3d 260, 284 (S.D.N.Y. 2018). By the same token, Greenberg v. Perfect Body Image, LLC, 2019 WL 3485700 (E.D.N.Y. 2019). 92 " See, e.g., Greenberg, supra note 91; Alzheimer’s Disease and Related Disorders Association, supra note 89; Passport Health, LCC v. Avance Health System, Inc., 2018 WL 6620914 (E.D.N.C. 2018); JIVE Commerce, LLC v. Wine Racks America, Inc., 2018 WL 3873675 (D. Utah 2018); EarthCam, Inc. v. OxBlue Corp., 2014 WL 4702200 (N.D. Georgia 2014); 1–800 Contacts, supra note 86; General Steel Domestic Sales, LLC v. Chumley, 2013 WL 1900562 (D. Colo. 2013); CollegeSource, Inc. v. AcademyOne, Inc., 2012 WL 5269213 (E.D. Pa. 2012); Network Automation, supra note 83; Fair Isaac Corp. v. Experian Information Solutions, Inc., 2009 WL 4263699 (D. Minn. 2009). 93 " Goldman, supra note 78, 521. 94 " D.J. Franklyn & D.A. Hyman, Trademarks as Search Engine Keywords: Much Ado About Something? 26 Harvard Journal of Law & Technology 481 (2013). 95 " S. Bechtold & C. Tucker, Trademarks, Triggers, and Online Search, 11 Journal of Empirical Legal Studies 718 (2014). After the Google France ruling, Google relaxed its AdWords policy allowing third parties to register keywords without the approval of the trademark owner, with only a limited complaint procedure for trademark owners. 96 " J.P. Dotson, R.R. Fan, E. McDonnell Feit, J.D. Oldham, Y.-H. Yeh, Brand Attitudes and Search Engine Queries, 37 Journal of Interactive Marketing 105 (2016). 97 " A. Simonov & S. Hill, Competitive Advertising on Brand Search: Traffic Stealing and Consumer Selection, Columbia Business School Research Paper No. 18–59, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3204394 (accessed February 5, 2020). 98 " See also A. Simonov, C. Nosko, and J.M. Rao, Competition and crowd-out for brand keywords in sponsored search, 37 Marketing Science 200 (2018). 99 " L. Coviello, U. Gneezy, and L. Goette, A large-scale field experiment to evaluate the effectiveness of paid search advertising, CESifo Working Paper Series No. 6684 (2017) https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3061698 (accessed February 5, 2020). 100 " T. Blake, C. Nosko, and S. Tadelis, Consumer heterogeneity and paid search effectiveness: A large-scale field experiment, 83 Econometrica 155 (2015). 101 " J.M. Golden & J.J Horton, The effects of search advertising on competitors: An experiment before a merger, Working paper (2018), http://john-joseph-horton.com/papers/search_ad_exp.pdf (accessed February 5, 2020). 102 " European Commission, 17 December 2018, Case AT.40428, Guess, C(2018) 8455 final. 103 " UK Competition and Markets Authority, supra note 18. 104 " Netherlands Authority for Consumers & Markets, supra note 19. 105 " In the Matter of 1–800 Contacts, supra note 20. 106 " European Commission, supra note 17, para. 632. 107 " CJEU, 13 October 2011, Case C-439/09, Pierre Fabre Dermo-Cosmetique SAS v. President de l’Autorite de la Concurrence. 108 " CJEU, 6 December 2017, Case C-230/16, Coty Germany GmbH v. Parfümerie Akzente GmbH. See also General Court, 12 December 1996, Cases T-19/92 and T-88/92, Groupement d’achat Edouard Leclerc v. Commission, arguing that the luxury image or aura of luxury of products may justify a selective distribution system. 109 " For a critical analysis, see G. Colangelo & V. Torti, Selective Distribution and Online Marketplace Restrictions under EU Competition Rules after Coty Prestige, 14 European Competition Journal 81 (2018). 110 " CJEU, 23 April 2009, Case C-59/08, Copad SA v. Christian Dior Couture SA. 111 " European Commission, supra note 102. See also German Competition Authority (Bundeskartellamt), 26 August 2015, Case No. B2–98/11, ASICS Deutschland GmbH, confirmed by the Düsseldorf Higher Regional Court (Oberlandesgericht), 5 April 2017, Case No. VI-Kart 13/15(V). 112 " Guess, supra note 102, paras. 48–52. 113 " Id., para. 121. 114 " Id., paras. 115, 116, 117, and 122. 115 " UK Competition and Markets Authority, supra note 18. See also UK Competition and Markets Authority, Competitive landscape and effectiveness of competition, supporting paper for the Digital comparison tools market study, 2017, https://www.gov.uk/cma-cases/digital-comparison-tools-market-study (accessed February 5, 2020). 116 " UK Competition and Markets Authority, supra note 115, para. 4.62. 117 " Id., para. 4.62. 118 " UK Competition and Markets Authority, supra note 18, 64; UK Competition and Markets Authority, supra note 115, paras. 4.88 and 4.89. 119 " Netherlands Authority for Consumers & Markets, supra note 19. 120 " Tichy v. Hyatt Hotels Corp., 2019 WL 1318674 (N.D. Ill. 2019). 121 " Netherlands Authority for Consumers & Markets, supra note 19, 2. 122 " The price increase is higher (around 5 percent) for hotels who violate the price parity clause OTAs impose. In the Netherlands, OTAs are allowed to impose price parity clauses which imply that hotels cannot post a lower price on their own website than on the OTA. 123 " In the Matter of 1–800 Contacts, supra note 20. 124 " 1–800 Contacts, supra note 86. 125 " In the Matter of 1–800 Contacts, supra note 20, 14. 126 " Id., 6. 127 " Id., 18. 128 " In re Polygram Holding, Inc., 136 F.T.C. 310, 344 (2003); California Dental Association n v. FTC, 526 U.S. 756, 770 (1999). 129 " In the Matter of 1–800 Contacts, supra note 20, 37. 130 " Id., 23. 131 " Id., 38. 132 " Id., 23. 133 " Id., 41. 134 " See Intellectual Property, Internet Law, and Antitrust Professors, Brief for Amici Curiae in support of Respondent, Stanford Law and Economics Olin Working Paper No. 538 (2019) https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3452884 (accessed February 5, 2020). 135 " In the Matter of 1–800 Contacts, supra note 20, 42. 136 " Id., 47. 137 " Id., 50. 138 " Dissenting Statement of Commissioner Noah Joshua Phillips, 1–800 Contacts, FTC Docket No. 9372 (2018), https://www.ftc.gov/system/files/documents/public_statements/1421309/docket_no_9372_dissenting_statement_of_commissioner_%20phillips_redacted_public_version.pdf (accessed February 5, 2020). See also G.A. Manne, H. Singer, and J.D. Wright, Antitrust Out of Focus: The FTC’s Myopic Pursuit of 1–800 Contacts’ Trademark Settlements, The Antitrust Source 1 (April, 2019). 139 " Phillips’ Dissenting Statement, supra note 138, 2–5. 140 " Id., 7. 141 " Id., 8. 142 " Id., 43. 143 " Id., 9 and 43. 144 " Id., 11. 145 " Id., 37. 146 " Id., 26. 147 " Id., 31–32. 148 " See Manne, Singer, and Wright, supra note 138, 7, arguing that the relevant question from an economic perspective is whether and to what extent any portion of 1–800 Contacts’ premium can be attributed to the allegedly anticompetitive trademark settlements. 149 " See CJEU, 30 January 2020, Case C-307/18, Generics (UK) Ltd and others v. Competition and Markets Authority; General Court, 12 December 2018, Case T-691/14, Servier SAS v. European Commission; General Court, 8 September 2016, Case T-472/13, Lundbeck A/S v. EFPIA and Associations and European Commission; European Commission, Guidelines on the application of Article 101 of the Treaty on the Functioning of the European Union to Technology Transfer Agreements, OJ C 89 (2014), §239; FTC v. Actavis, 570 U.S. 136 (2013). 150 " FTC v. Watson Pharmaceuticals, 677 F.3d 1298 (11th Cir. 2012); In re Ciprofloxacin Hydrochloride Antitrust Litigation, 544 F.3d 1323 (Fed. Cir. 2008); In re Tamoxifen Citrate Antitrust Litigation, 466 F.3d 187 (2nd Cir. 2006); Schering-Plough v. FTC, 402 F.3d 1056, 1066 (11th Cir. 2005); Valley Drug v. Geneva Pharmaceuticals, 344 F.3d 1294 (11th Cir. 2003). 151 " In re Cardizem CD Antitrust Litigation, 332 F.3d 896 (6th Cir. 2003). 152 " In re K-Dur Antitrust Litigation, 686 F.3d 197 (3rd Cir. 2012). 153 " Actavis, supra note 149, 158. 154 " M.A. Carrier, Three Challenges for Pharmaceutical Antitrust, 59 Santa Clara Law Review 613, 631 (2020). However, see the dissenting opinion written by Chief Justice Roberts and concurred in by Justices Scalia and Thomas downplaying the connection between payment and patent weakness and supporting the defense based on risk aversion: “A patent holder may be 95% sure about the validity of its patent, but particularly risk averse or litigation averse, and willing to pay a good deal of money to rid itself of the 5% chance of a finding of invalidity” (Actavis, supra note 149, 172). See also In Re Wellbutrin XL Antitrust Litigation, 868 F.3d 132, 168–169 (3rd Cir. 2017), agreeing with the argument raised by a group of antitrust economists that explains why risk aversion makes it difficult to use the size of a settlement as a proxy for the brand-name’s likelihood of success in litigation. 155 " European Commission, supra note 149, §239. 156 " Generics, supra note 149, paras. 85 and 94. 157 " By the same token S.N. Weinstein, Rigged Results? Antitrust Lessons from Keyword Auctions, 91 Tulane Law Review 1 (2017). 158 " Lemley & Shapiro, supra note 8, 76. 159 " S.L. Dogan & M.A. Lemley, Antitrust Law and Regulatory Gaming, 87 Texas Law Review 685, 689 (2009). 160 " Actavis, supra note 149, 164. 161 " Id., 172. 162 " Phillips’ Dissenting Statement, supra note 138, 7 and 37. 163 " Interflora, supra note 24, paras. 57 and 64; Google France, supra note 23, paras. 93 and 95. 164 " Interflora, supra note 24, paras. 90 and 91. 165 " Google France, supra note 23, paras. 83, 84, 89 and 90. 166 " CJEU, 30 January 1985, Case C-35/83, BAT Cigaretten-Fabriken GmbH v. Commission. 167 " Id., para. 33. 168 " European Commission, supra note 149, §235. 169 " See, e.g., CJEU, 6 October 1982, Case C-262/81, Coditel SA and others v. Ciné-Vog Films SA and others (Coditel II), paras. 17 and 19. 170 " For an economic analysis of the different legal standards applied to sham or vexatious litigation in EU and US competition law, see I. Lianos & P. Regibeau, “Sham” Litigation: When Can It Arise and How Can It Be Reduced?, 62 Antitrust Bulletin 643 (2017). Notably, the authors investigate the economic effects of the two different approaches adopted by courts to identify sham claims: according to a narrow view, sham litigation is defined as a pattern of objectively baseless claims, manifestly unfounded or without probable cause; according to a second and broader approach, the fact that the claim is not baseless does not preclude the finding that the use of litigation constitutes an antitrust violation, but rather the existence of sham litigation is evaluated by a purely cost-benefit analysis which leads in characterizing as sham even claims that are filed with probable cause, if the benefits of the litigation, discounted by the probability of winning, would be too low to repay the costs. 171 " Beebe & Scott Hemphill, supra note 14. 172 " See R.A. Epstein, K.N. Hylton, T.A. Lambert, G.A. Manne, H. Singer, and Washington Legal Foundation, Brief for Amici Curiae in support of Petitioner, (2019) https://www.wlf.org/wp-content/uploads/2019/06/06-14-2019-Scholars-WLF-Amicus-Brief-ISO-1-800-Contacts.pdf (accessed February 5, 2020), arguing that the FTC failed to consider the settlements’ procompetitive power to stanch advertisement free riding. See also UK Competition and Markets Authority, supra note 115, paras. 4.65–4.67, evaluating the free-riding justification for brand-bidding agreements. 173 " A different and provocative perspective is offered by R.A. Woodcock, The Obsolescence of Advertising in the Information Age, 127 Yale Law Journal 2270 (2018), which reports the demise of advertising’s information function, thereby arguing that consumers no longer need to have advertising thrust upon them in order to learn about the marketplace. According to the author, advertising in the information age is necessarily anticompetitive: since advertising no longer serves an information function, the only thing advertising adds to markets is its manipulative function, which distorts consumer preferences, rolls back gains associated with consumers’ greater access to information in the information age, and places more efficient and innovative competitors at a competitive disadvantage. 174 " See Goldman, supra note 78, arguing that trademark law must step aside when searchers receive relevant content they may want. 175 " E.V. Mariscal & D.S. Evans, The Role of Keyword Advertising in Competition among Rival Brands, Coase Sandor Institute for Law & Economics Working Paper No. 619, 2012, 3, https://chicagounbound.uchicago.edu/law_and_economics/570/ (accessed February 5, 2020). 176 " See, e.g., Senftleben, supra note 55, 141–143, arguing that trademark rights have progressively lost their defensive nature getting close to exploitation rights, despite trademark owners, unlike inventors and authors, cannot claim to have created IP that furthers science or art. 177 " Mariscal & Evans, supra note 172, 3. Author notes " Jean Monnet Professor of EU Innovation Policy; Associate Professor of Law and Economics, University of Basilicata; TTLF Fellow, Stanford Law School; https://orcid.org/0000-0002-0089-3545; giuseppe.colangelo1975@gmail.com. I would like to thank anonymous referees and participants to the MaCCI Annual Conference at the University of Mannheim for their valuable comments and suggestions. The usual disclaimer applies. © The Author(s) 2020. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permission@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - COMPETING THROUGH KEYWORD ADVERTISING
JF - Journal of Competition Law & Economics
DO - 10.1093/joclec/nhaa011
DA - 2009-03-01
UR - https://www.deepdyve.com/lp/oxford-university-press/competing-through-keyword-advertising-Pj0Xrtj0su
SP - 1
VL - Advance Article
IS - 
DP - DeepDyve
ER -