The Fourth Circuit recently issued its widely anticipated ruling in Rosetta Stone Ltd. v. Google, Inc.,1 more than six months after oral arguments were held in the case. Rosetta Stone, the producer of language-learning products, sued Google based on its AdWords service, which allows third parties to purchase keywords such as “Rosetta Stone” that then trigger the display of third-party advertisements on the Google search results page. Rosetta Stone alleged, among other things, that many of the advertisements triggered by a Google user’s input of the keyword “Rosetta Stone” promote counterfeit Rosetta Stone products.

The U.S. District Court for the Eastern District of Virginia had granted summary judgment in favor of Google, but the United States Court of Appeals for the Fourth Circuit vacated several of the district court’s key rulings, holding that its rulings on direct trademark infringement, contributory trademark infringement and trademark dilution were all in error. Notably, the Fourth Circuit’s ruling was largely procedural in nature and does not preclude Google from ultimately proving that its AdWords program is non-infringing based on a complete factual record presented to the district court.

A federal trademark infringement claim generally has two key elements. A trademark owner must show both that (1) the defendant used a trademark in commerce and that (2) the defendant’s use causes a likelihood of confusion among consumers. Here, the Fourth Circuit assumed, and Google did not contest, that Google’s policy allowing advertisers to employ trademarks owned by others as keywords to trigger advertising constitutes “use in commerce” for the purposes of the first prong of the analysis.2 This is consistent with the Second Circuit’s Rescuecom Corp. v. Google, Inc. decision and is consistent with the general trend in legal challenges to the keyword advertising programs, which has been to focus the inquiry on the likelihood of confusion prong, rather than on use in commerce.

With regard to likelihood of confusion, the Fourth Circuit endorsed the district court’s consideration of only three of the nine traditional confusion factors, namely (1) Google’s intent with regard to the ROSETTA STONE mark, (2) actual confusion and (3) the sophistication of the relevant purchasers.3 In doing so, the Fourth Circuit held that there is no rule requiring a point-by-point analysis of all nine factors, particularly whereas – like in this case – the trademark in question is being used to refer to the trademark owner’s own goods.4 However, the Fourth Circuit also held that the district court committed reversible error in finding no material question of fact in its analysis of each of the three factors that were examined.

First, as to the question of intent, the appeals court noted that Google’s own internal surveys showed that Google knew that confusion would likely result from allowing advertisers to employ third-party trademarks in ad text, but that Google anticipated a boost in revenue from allowing the practice. The Fourth Circuit held that, viewed in the light most favorable to Rosetta Stone, this was sufficient to permit a reasonable trier of fact to conclude that Google intended to cause confusion.5 Second, the appeals court held that the lower court had failed to properly consider Rosetta Stone’s substantial evidence of actual consumer confusion, including deposition testimony, evidence of consumer complaints to Rosetta Stone and a consumer survey. 6 Third, with regard to the sophistication of the relevant purchasers, the court held that the evidence of the relatively high price of Rosetta Stone’s software (up to $579 for certain bundles) was not necessarily dispositive with regard to the sophistication of Rosetta Stone’s customers. 7 Additionally, the Fourth Circuit held that the district court misapplied the functionality doctrine, which prevents trademark plaintiffs from enjoining useful product features.8 The district court had held that, because the term “Rosetta Stone” – when used as a keyword – allowed Google to more usefully index the Internet, the term was functional. The Fourth Circuit reversed, holding that because the mark ROSETTA STONE is not essential for the functioning of its language-learning products, and indeed is a classic source identifier, the functionality doctrine did not apply.9

With regard to the contributory infringement claims, the Fourth Circuit held that the district court inappropriately shifted to Rosetta Stone the burden of showing that Google’s actions in addressing complaints of counterfeiting were inappropriate.10 Finally, the Fourth Circuit held that the lower court erred with regard to the trademark dilution claims, which allege that Google’s conduct will weaken the ability of the ROSETTA STONE mark to identify a source.11 In particular, the Fourth Circuit held that the district court inappropriately collapsed the dilution analysis into the question of whether Google was using the ROSETTA STONE trademark to identify Google’s own goods and services.12

This decision signals that future disputes over keyword advertising will continue to focus heavily on the likelihood of confusion analysis, and, in our view, a question remains concerning what set of facts need to be present for a court to find that Google’s AdWords policies (or other keyword-triggered advertising practices) result in a likelihood of confusion (and ultimate liability).