Retargeting Keyword Ads for Potential New Uses

Whatever the final resolution of still-unsettled law of search engine sale of keyword advertising, the current muddle may only be prelude to confronting new and more surreptitious ways of using trademarks to drive Internet commerce. This may involve unseen uses of brand names in search engine optimization strategies or in novel forms of behavioral advertising. As users spend more time on sites other than Google, such as social networking sites, the ways in which these site operators and their advertising affiliates gather and use data for targeted advertising may have greater consequences for trademark owners than search engine sales of keywords now ' and may present greater legal challenges about which to wrap our hands.

Jonathan E. Moskin is a partner in the New York office of Foley & Lardner LLP. He can be contacted at [email protected].

At a recent Practising Law Institute presentation, attended by a solid sampling of intellectual property attorneys, many of whom acknowledged representing major brand-name companies, the author sought a show of hands of those who would recommend that a client bring suit against a third party buying keyword ads based on one of those brands. Only one wavering hand said “maybe.” Not one hand endorsed such action hands down. Despite a great body of precedents and an even greater body of commentary, the heads present, who undoubtedly knew more than most about the subject, seemed more inclined simply to throw up their hands than take up arms. Indeed, in more than a decade since the first decision in Playboy Ent., Inc. v. Netscape Commun's Corp. , 55 F. Supp. 2d 1070, 1086-88 (C.D. Cal. 1999), rev'd , 354 F. 3d Cir. 1020 (9th Cir. 2004), there have been at most a small handful of final decisions giving meaningful guidance as to when or if keyword ads are permitted or precluded. Playboy itself was simply a reversal and remand of a decision granting summary judgment dismissing the plaintiff's infringement claim.

Keyword advertising should be seen as simply one band on the spectrum of online advertising by franchisors and others to promote their brands. Much of that spectrum is now occupied by various types of behavioral advertising, in which advertisers can track conduct of users on the Internet to deliver advertising targeted to a user profile or recent Internet activity. Viewed in this light, a perhaps greater concern to brand owners than the types of sponsored links served by search engines are new opportunities for use of keywords to retarget advertising based on behavioral patterns of Internet users and in manners almost impossible to monitor. For instance, one recent article reports on a retargeting service that can track users' online activity; if that activity reveals an interest in a particular branded product or service (say Toyota), it is not only possible to retarget to the user advertising for that brand but also to retarget advertising for a competing brand based on the use of keywords tied to the first brand. See Magnetic Brings Search Re-Targeting to the Masses, http://searchengineland.com/magnetic-brings-search-re-targeting-to-the-masses-38535 (Mar. 22, 2010).

Thus, what matters most in resolving the cases that continue to be brought regarding use by search engines of such marketing tools is that the legal doctrine not be skewed based on incorrect assumptions about the operation of the Internet or incorrect analogies to the bricks-and-mortar world. Only such a clear-eyed focus will ensure that the law develops in tandem with new technologies.

The Case Against Keywords

The cases that have held (or suggested) that use of a keyword tied to a competitor's mark is (or in theory might be) an infringement have tended to be decided based on three grounds: 1) that it is inherently wrong to obtain a “free ride” by trading on the reputation of a competitor; 2) that even if the Internet user ultimately recognizes before making a purchase that a keyword ad is simply that, and not a posting by the trademark owner, even a fleeting initial misunderstanding constitutes actionable “initial interest confusion”; and 3) application of a unique, Internet-only test of infringement under which the analysis is confined to three factors: similarity of the marks, relatedness of the goods or services and the simultaneity of use on the Internet. Storus Corp. v. Aroa Marketing, Inc., 2008 WL 449835 (N.D.Cal. Feb. 15, 2008).

However, while “bad intent” is certainly a factor that may need to be considered (and almost always is in all intellectual property cases), relying on intent alone can easily become circular, as what is “good” or “bad” in turn depends on what the law allows. And although some courts and practitioners remain enamored of the initial interest confusion theory espoused in Brookfield Commun's, Inc. v. West Coast Ent. Corp. , 174 F.3d 1036 (9th Cir. 1999), drawn from a real-world metaphor of traffic being diverted from a highway by a misleading road sign, others have become increasingly skeptical that the analogy translates to the Internet. Google, Inc. v. American Blind & Wallpaper Factory, Inc., 2007 WL 1159950 (N.D.Cal. 2007) thus noted that it is “not reasonable to find initial interest confusion when a consumer is never confused as to source or affiliation, but instead knows or should know, from the outset that a product or web link is not related to that of the trademark holder because the list produced by the search engine so informs him.” (see also Accord, JG Wentworth, SSC v. Settlement Funding LLC, 2007 WL 30115 at * 7 (E.D.Pa. 2007)). Finally, the logic behind the opaque notion that only three factors need be considered (creating in effect per se liability), is derived from a short-form order without analysis in Comp Examiner Agency, Inc. v. Juris, 1996 WL 376600 (C.D.Cal. 1996), and has never been explained.

The Case to Allow Keywords

The cases that have held (or suggested) that confusion is not likely, have tended to rely on a simple analogy to familiar purchasing experiences, such as in a drug store, where generic store brands are shelved with branded products in a manner arguably similar to the way in which sponsored keyword ads appear alongside search results for familiar branded products or services. 1-800 Contacts, Inc. v. WhenU.com, Inc. , 414 F.3d 400 (2d Cir. 2005). There was also a brief diversion when a series of district court cases (all in New York), relying on 1-800 Contacts, concluded that keyword ads could not entail use in commerce because an “[entirely veiled machine-linking function was not a trademark use, since it did not entail having the mark placed ' on the goods or their containers or the displays associated therewith.” As this author first noted in Virtual Trademark Use ' The Parallel World of Keyword Advertising, 98 Trademark Rptr. 873, 880 (May-June 2008), this narrow view of what constituted infringing trademark use relied on the wrong section of the Lanham Act ' the provision defining how a trademark owner acquires rights in a mark, not the definition of infringement. In Rescuecom Corp v. Google, Inc. , 562 F.3d 123 (2d Cir. 2009), the Second Circuit agreed.

Rescuecom also observed, however, that reasoning by analogy has its limits, such that a bare abstract comparison to a grocer's shelves need not automatically answer whether, when or if Internet users are confused. And this author did recently oversee a survey (in a case settled on a confidential basis) in which significant confusion was shown when a generic store brand and a nationally known brand were shelved together in a checkerboard fashion. The analogy itself thus may need to be shelved.

Although the “use-in-commerce” test has now been rejected as a basis to unfetter keyword advertising, a similar test recently surfaced as a basis to preclude liability against at least the search engines themselves that are providing the technical platform for such advertising. Rosetta Stone Ltd. v. Google, Inc., 2010 WL 3063152 (E.D.Va. Aug. 3, 2010), thus held that Google's AdWords program was simply a functional tool for indexing information relevant to consumer search queries: “The keywords have an essential indexing function because they enable Google to readily identify in its databases relevant information in response to the web user's inquiry.” (Although the decision has been appealed, laying aside the effect of the district court ruling on direct suits against advertisers purchasing keyword ads, it does present an enormous obstacle to claims against search engines for contributory liability in serving up such ads.)

Law in the Balance

Another unique test, more common in Constitutional analysis than infringement litigation, has been a balancing test, weighing the risks of confusion against the functional benefits to consumers from the search technologies financed by advertising. In Hearts on Fire Co. v. Blue Nile, Inc. , 603 F. Supp. 2d 274 (D.Mass. 2009), the court thus expressly noted that “The choice enhancing properties of internet advertising should not be stifled on account of fleeting confusion among competing products.” A similar sentiment was echoed in Mary Kay, Inc. v. Weber , 601 F. Supp. 2d 839 (N.D. Tex 2009), and there is, perhaps, no accident that a similar balancing approach has been endorsed by the Federal Trade Commission in its initial reports assessing the risks of behavioral advertising ( i.e. , the tracking by IP address of computer user activity so as to deliver advertising targeted to the user's interests; see FTC Staff Report: Self-regulatory Principles For Online Behavioral Advertising , February 2009). That logic continues to be reflected in the FTC's more recent report on online advertising, Protecting Consumer Privacy in an Era of Rapid Change, published Dec. 1, 2010; and it may be fitting, given the broader perspective of keyword advertising as simply one point on the spectrum of Internet advertising.

Transcending the broad generalities of most of the reported decisions, Hearts on Fire also set forth seven empirical factors to consider in assessing whether confusion is or is not likely in any given keyword case. These include the mechanics of Web browsing, which of course permits users to toggle back and forth between sites with ease; the sophistication of the users; the downstream content of the Web site being advertised; the duration of any confusion; and the specific content of searches. In merely denying a motion to dismiss, the court did not provide its own final analysis. However, by definition, such an empirical focus surely will yield results more targeted to actual experience than will the decision-by-analogy model. No doubt empirical survey evidence will need to play a bigger role than it has to date in deciding these cases.

Whatever the final resolution of still-unsettled law of search engine sale of keyword advertising, the current muddle may only be prelude to confronting new and more surreptitious ways of using trademarks to drive Internet commerce. This may involve unseen uses of brand names in search engine optimization strategies or in novel forms of behavioral advertising. As users spend more time on sites other than Google, such as social networking sites, the ways in which these site operators and their advertising affiliates gather and use data for targeted advertising may have greater consequences for trademark owners than search engine sales of keywords now ' and may present greater legal challenges about which to wrap our hands.

Jonathan E. Moskin is a partner in the New York office of Foley & Lardner LLP. He can be contacted at [email protected].