Rosetta Stone, maker of the well known language-learning software, sued Google for Google’s use of the ROSETTA STONE trademark in its AdWords program. The AdWords program allows companies to bid on key words, including a trademark, to cause advertisements known as “Sponsored Links” to appear when a Google user enters a query including the key word. The result of the AdWords program was to allow companies that were not affiliated with Rosetta Stone to purchase the right to have a Sponsored Link in a user’s search results when that user of Google searched for the words (and plaintiff’s trademark) “Rosetta Stone.”
The district court found that the material facts were undisputed and granted summary judgment for Google denying various theories of liability pursued by Rosetta Stone. This decision was one of the first to reach the merits regarding Google’s advertising practices. Rosetta Stone appealed the loss. Over forty amici (“friend of the court”) filed briefs weighing in on the issues.
The court of appeals vacated the district court’s decision, in key parts, and ruled, instead, that the material facts were disputed such that one could rule favorably for Rosetta Stone. The case was thus remanded to the district court for a trial on the merits.
The key defenses by Google at issue were directed to theories of liability based on (A) direct infringement, (B) contributory infringement and (C) trademark dilution. Google argued that the direct infringement was only as a nominative use of the subject trademark; that is, unlike classic source confusion, the defendant is using the plaintiff’s trademark to indicate plaintiff’s goods or services, not to indicate those made by defendant or others.
As to direct infringement, the court of appeals examined three factors from the typical likelihood-of-confusion analysis and found disputed facts; namely, (1) Google’s intent, (2) evidence of actual confusion, and (3) the purchaser’s level of sophistication. Google’s advertising policy had shifted over the years in allowing companies to use another’s trademark and, at some point in time, Google itself recognized that confusion may result as to this practice. Thus, there was disputed evidence regarding Google’s intent. The court also found that there was evidence that purchasers had been actually confused which evidence the district court had incorrectly discounted. Even though the purchasers knew that the purchased software was not made by Rosetta Stone, these purchasers believed that Rosetta Stone had in some way endorsed or sponsored this software by others. Importantly, trademark confusion can stem not only from confusion as to the source of goods/services, but also as to any consumer’s erroneous belief as to sponsorship, affiliation or connection between the trademark owner and the source offering the goods/services. This evidence of actual confusion also showed that the purchaser’s level of sophistication could not be regarded as undisputed since even sophisticated purchasers exhibited confusion as to sponsorship. Note that while Google intended a nominative use of the subject trademark (for example, the sponsored links would point to resellers of the plaintiff’s product), this was not always the situation as shown by the disputed facts when the issue of incorrect sponsorship was correctly weighed.
As to contributory infringement, the court of appeals found that the district court’s decision erred in two respects: (1) that the decision incorrectly framed the issue as whether Rosetta Stone had shown it was entitled to summary judgment based on undisputed facts when, in fact, Google had the burden to show that the undisputed facts favored its defense; and (2) that the decision relied on another court case decided, significantly, after a full trial on the merits, and thus that reliance was inappropriate for the summary judgment stage. Summary judgment must be based on undisputed material facts, not a weighing after trial of the disputed evidence.
As to trademark dilution, the court of appeals ruled that the district court’s decision also erred in two respects. First, the district court held that Google’s use of the trademark was a fair use that did not point to Google’s goods/services. The court of appeals, however, ruled that the analysis was incomplete for failing to assess Google’s good faith in using the trademark; an analysis of good faith being required for any finding of fair use. Second, the district court’s decision held that the fame of the trademark Rosetta Stone had actually increased and therefore the plaintiff was not damaged. The court of appeals, however, faulted that analysis as inappropriately relying on another court decision that was focused on a defendant’s parody of the plaintiff’s trademark. Further, the lower court incorrectly focused on the wrong time period of 2009 when Google changed its advertising practices. Specifically, Google’s advertising practices began in 2004 and thus Rosetta Stone had to prove its trademark was famous at that time. The facts as to this fame in 2004 were disputed rendering summary judgment inappropriate. As this case goes forward to trial, this required showing of fame in 2004 may prove to be a higher hurdle for Rosetta Stone.
Returning to the issue of direct infringement, one aspect of the district court’s decision had generated much commentary and speculation. As an alternative basis for finding no direct infringement, the lower court ruled that Google made a “functional” use of the plaintiff’s trademark. The issue of functionality has long been associated with analyzing the validity of the plaintiff’s trademark, not the nature of any defense to infringement by the defendant. It was unclear whether a new doctrine of functionality would arise from the lower court decision. The court of appeals held to the traditional doctrine of functionality and rejected the district court’s deviation in theory. While a new theory of functionality has been squelched by this appeal, it is believed this case will continue to generate interest and controversy since Google’s advertising practices at issue are very much the lifeblood of the company’s revenue and an integral part of the interface between the public and its access to the Internet.