Plaintiff sold cigars under the mark JR CIGARS and the domain name “jrcigars.com.” Defendant GoTo.com, which later changed its name to Overture, operated a “pay-for-priority” search engine that sold search keywords to the highest bidding advertiser. Defendant’s search engine results featured the highest bidder first, followed either by lower bidders or by the “natural” search results. Advertisers paid defendant only when users clicked on their paid listings. Plaintiff claimed that defendant’s sales of its mark as keywords enabled competitors to pass themselves off as plaintiff and to divert internet traffic from plaintiff’s website to competing websites. Defendant countered that, according to its relevancy guidelines for accepting bids, all advertisers represented that their advertisements and corresponding websites did not violate any third-party trademark rights. In particular, defendant did not accept search terms for comparative advertising unless the advertiser’s website presented “actual, significant information about their competitors’ products by comparing them to their own.” Defendant claimed that all of the disputed keyword listings either met this requirement or resulted from the confusion of the letters “JR” in “JR Cigar” for the suffix “Jr.,” presumably for the sale of smaller, junior-sized cigars. Plaintiff sued for infringement, unfair competition, and dilution. Both parties moved for summary judgment. Although plaintiff did not raise the issue of secondary liability, the court initially noted that the Third Circuit has held that “in certain instances, secondary, indirect liability is a legitimate basis for liability under the federal unfair competition statute.” Regarding infringement, the court first focused on whether defendant used plaintiff’s marks in commerce. Relying on the GEICO v. Google decision, the court held that, as a matter of law, defendant made trademark use of plaintiff’s marks. The court distinguished this case from the WhenU pop-up ad cases because, unlike defendant here, WhenU “did not market the protected marks themselves as keywords to which advertisers could directly purchase rights.” The court reasoned that giving prominence to the highest bidder is “qualitatively different” from pop-up advertisements where the internal computer coding is non-public and not for sale. The court held that defendant used plaintiff’s marks in three ways: (a) defendant traded on the value of plaintiff’s marks by accepting bids from plaintiff’s competitors, (b) defendant ranked its paid advertisers by their bids before providing any “natural” listings and thus “injected itself into the marketplace, acting as a conduit to steer potential customers away from JR to JR’s customers,” and (c) defendant suggested plaintiff’s marks to prospective advertisers as search terms and marketed them to plaintiff’s competitors. Turning to likelihood of confusion, the similarity-of-the-marks factor favored plaintiff because JR’s mark and the search terms sold by defendant were similar if not identical. The strength-of-the-mark factor also favored based on plaintiff’s use of the mark for nearly thirty years, plaintiff’s substantial advertising expenditures and sales revenues, and extensive unsolicited third-party recognition. The power of the JR mark in the cigar industry was further reflected by the 20+ competitors that bid on plaintiff’s mark. The absence of actual confusion in the form of mistaken purchasing decisions favored defendant, but evidence regarding “temporary diversion of potential customers away from [plaintiff’s] website” created an issue of fact on this factor. Regarding intent, defendant’s relevancy guidelines and its application of them created another issue of fact. Even though the target consumers of each business were different–cigar smokers and search engine users–the marketing efforts of the companies overlapped because both advertised to consumers seeking to buy cigars. Despite defendant’s contention that retail cigar services and search engine services were unrelated, the court held that “the goods to which Internet users are ultimately directed in GoTo’s search results are similar to JR’s products.” Finally, the court examined the issue of initial-interest confusion and found a genuine issue of fact based on defendant’s intent, discussed above, and on plaintiff’s statistical evidence on the diversion of consumers away from its site. For example, during a two-month period, plaintiff averaged a 23.98% “diversion rate” in click-throughs and paid listings to plaintiff’s competitors. Defendant argued that this evidence of diversion was insufficient to prove initial-interest confusion absent proof that any consumers were actually confused. The court, however, found that this statistical evidence was “arguably indicative” of a likelihood of confusion. Because defendant did not contradict this evidence of diversion, the “state of the record favors [plaintiff],” but it was up to the trier of fact whether “to credit this evidence.” The court thus denied summary judgment on the issue of infringement. Finally, the court also denied summary judgment on the federal and state dilution claims. Although the court found that plaintiff’s marks were famous, there were issues of fact on whether there was actual dilution.