Rosetta Stone, a leading language-learning software producer, sued Google for trademark infringement and dilution based on Google’s sale of Rosetta Stone’s marks as keywords. In 2010, the District Court for the Eastern District of Virginia entered summary judgment against Rosetta Stone’s trademark claims (direct, contributory, and vicarious trademark infringement; dilution), and dismissed Rosetta Stone’s unjust enrichment claim. See Rosetta Stone Ltd. v. Google, Inc., 730 F. Supp. 2d 531 and 732 F. Supp. 2d 628 (E.D. Va. 2010).
Rosetta Stone appealed. On April 9, 2012, the Fourth Circuit issued its long-awaited decision. See Rosetta Stone Ltd. v. Google, Inc., — F.3d –, No. 10-2007, 2012 WL 1155143 (4th Cir. April 9, 2012). The Court reversed the district court’s summary judgment rulings in Google’s favor on Rosetta Stone’s three most important claims – for direct infringement, contributory infringement, and dilution. The decision is notable for the Court’s frank criticism directed at the district court’s orders, finding substantial flaws in the district court’s analysis of both the factual record and the applicable legal doctrine. The Fourth Circuit’s analysis contains a number of important assessments of actual confusion evidence, a defendant’s level of knowledge necessary to prove contributory infringement, the possibility of a plaintiff’s proving dilution when recognition of the plaintiff’s mark actually increased during the relevant time period, and several other matters. In light of these determinations, the Fourth Circuit’s affirmance of summary judgment on Rosetta Stone’s claims for vicarious infringement and dismissal of Rosetta Stone’s unjust enrichment claim provides scant comfort for Google.
The Fourth Circuit’s decision may well influence the approach to keyword advertising cases under U.S. trademark law. With this decision, the law would seem to have become more favorable to trademark owners whose marks are being sold and used as keywords. Search engines may reconsider some of their keyword advertising practices. Parties who use others’ trademarks as keywords for their own sponsored links may wish to assess whether any of their practices may be affected by the Fourth Circuit’s analysis. This is particularly true for situations in which the trademarks used as keywords also appear in the text of a sponsored link. In the meantime (and barring a settlement), the fate of Google’s current keyword advertising model stands to be determined by a jury.
- A Circuit Split in the Making?
The Fourth Circuit’s evaluation of the evidence of actual and likely consumer confusion stands in contrast to two recent decisions from the Ninth Circuit, Toyota Motor Sales, USA, Inc., v. Tabari, 610 F.3d. 1171 (9th Cir. 2010) and Network Automation, Inc., v. Advanced System Concepts, Inc., 638 F.3d 1137 (9th Cir. 2011). The Ninth Circuit premised these decisions on its finding that consumers have become sophisticated in exploring search engine results, including sponsored links. According to the Ninth Circuit, consumers understand what sponsored links are, recognize them by their labels and graphic set-offs on search results pages, and are “ready to hit the back button whenever they’re not satisfied with a site’s contents.” Tabari, 610 F.3d at 1179; see also Network Automation, 638 F.3d at 1152. Moreover, “consumers don’t form any firm expectations about the sponsorship of a website until they’ve seen the landing page – if then. This is sensible agnosticism, not consumer confusion.” Tabari, 610 F.3d at 1179.
Contrast this with the Fourth Circuit’s observation in Rosetta Stone that “even well-educated, seasoned Internet consumers are confused by the nature of Google’s sponsored links and are sometimes even unaware that sponsored links are, in actuality, advertisements. At the summary judgment stage, we cannot say on this record that the consumer sophistication factor favors Google as a matter of law.” In fact, the Court noted, such uncertainty constitutes “quintessential actual confusion evidence.” Rosetta Stone at *10.
“Sensible agnosticism” versus “quintessential actual confusion evidence” – these are rather different, and potentially outcome-determinative, evaluations of rather similar states of mind. It remains to be seen how these two different approaches will play out in the evolution of trademark law as it relates to keywords and sponsored links.
- Background – Keyword Advertising
“Keywords” are words or word combinations sold by search engines – such as Google, Yahoo, and Bing – to advertisers. The keywords are used by the search engine to trigger sponsored links on the search results page that are relevant to a user’s search terms. Most advertisers pay a predetermined amount whenever a search engine user clicks on a sponsored link; some keywords are also sold at auctions, and the highest bidder obtains the most favorable placement on the search results page. Keyword advertising is big business. In the last four years, Google earned in excess of $110 billion worldwide with its keywords-driven AdWords program. Other search engines, such as Yahoo and Bing, likewise earn billions of dollars in revenue through their keywords-based online advertisements.
Many keywords are trademarks. When keyword advertising gained in popularity in the early 2000′s, a number of trademark owners brought infringement actions against both keyword purchasers and the search engines that sell keywords. The trademark owners argued that using their marks as keywords causes consumer confusion by triggering sponsored links directing consumers to competitors or even to sellers of counterfeit articles.
For many years, however, keyword advertising was largely insulated from legal attack. First, the courts ruled that the Lanham Act, the federal trademark law, simply did not apply at all to keywords because selling and using keywords for the purpose of triggering sponsored links does not qualify as “use in commerce,” a necessary precondition to applying the Lanham Act. That impediment was largely overcome in 2009, when the Second Circuit held in Rescuecom Corp. v. Google, Inc., 562 F.3d 123 (2d Cir. 2009), that the Lanham Act does apply to keywords. Other Circuits have since followed suit, and lack of “use in commerce” is no longer an impediment to bringing keyword-related infringement actions.
But a second impediment to a successful trademark infringement action remained – the widespread assumption among many courts that sponsored links are not very likely to cause confusion among consumers. As noted above, the Ninth Circuit found in the Tabari and Network Automation decisions that Internet users have over the last decade become sophisticated in using search engines and have a solid understanding of sponsored links, so that the labeling and graphic set-off of sponsored links from the “natural” search results serves to minimize any possible likelihood of consumer confusion.
Given these baseline assumptions, proving likelihood of confusion in a keyword case was a heavy burden. With the Fourth Circuit’s decision, the burden may have eased.
- The Fourth Circuit’s Analysis
- Direct Infringement
On Rosetta Stone’s direct infringement claim, the Fourth Circuit held that the district court had improperly discounted several crucial pieces of evidence:
- Rosetta Stone had presented deposition testimony of five consumers who mistakenly purchased what turned out to be counterfeit Rosetta Stone software. The district court had dismissed this testimony as de minimis. The Fourth Circuit noted, however, that Rosetta Stone also had received more than 250 complaints from consumers who were misled into purchasing counterfeit or pirated Rosetta Stone software between April 2009 and March 2010. This was sufficient for the Fourth Circuit to make the evidence more than de minimis. See Rosetta Stone at *8 – *9.
- Google had changed its AdWords policy in 2009 to permit the use of trademarks not only as keywords but also, in certain situations, in the sponsored links triggered by those keywords. The Fourth Circuit noted that this change in policy coincided with a “substantial proliferation of sponsored links to pirate/counterfeit sites.” Id. at *9 (citation and quotation marks omitted). this change in Google’s policy, from which Google “expected a substantial boost in revenue,” could be interpreted by the trier of fact as evidence of intent to cause confusion Id. at *7.
- An internal Google study analyzing consumer confusion in connection with sponsored links had concluded that “the likelihood of confusion remains high” and “94% of consumers were confused at least once.” Id. at *9.
- Google’s in-house attorneys were unable, in their Fed. R. Civ. Pro. 30(b)(6) depositions, to determine whether a given sponsored link was by an authorized Rosetta Stone reseller. Id. at 10. This uncertainty, the Court found, is “quintessential actual confusion evidence.” Id.
- Rosetta Stone’s expert report showed significant consumer confusion about whether sponsored links were authorized or approved by Rosetta Stone. Such confusion “as to affiliation, connection, or sponsorship” was sufficient to create a genuine factual issue for trial. Id. at *10.
An important takeaway from these findings is that, contrary to the district court’s position, consumers’ uncertainty as to the sponsorship, approval, or authorization of a third-party sponsored link by the trademark owner may constitute sufficient actionable confusion to overcome a summary judgment challenge. As noted above, this position contrasts with the Ninth Circuit’s interpretation of similar consumer uncertainty as “sensible agnosticism.” Tabari, 610 F.3d at 1179.
- Contributory Infringement
The Fourth Circuit likewise reversed the district court’s grant of summary judgment against Rosetta Stone’s contributory infringement claim. For Google to be liable for contributory infringement, it must have knowledge that “identified individuals” are using sponsored links to infringe Rosetta Stone’s marks. Id. at *14. Rosetta Stone had provided Google with a spreadsheet of “approximately 200 instances of sponsored links advertising counterfeit Rosetta Stone products.” Id. at 14. The Fourth Circuit found that this showing was enough to overcome Google’s summary judgment motion, and Rosetta Stone should have the right to bring its contributory infringement claim to the jury. Id. at 15.
The Fourth Circuit reversed summary judgment against Rosetta Stone on its dilution claim. The Court disagreed with the district court’s analysis of Google’s fair use defense, finding that the district court mistakenly collapsed an inquiry into Google’s use in commerce of Rosetta Stone’s marks with a fair use inquiry. Id. at 19. This improper analysis caused the district court unduly to shift the entire burden of proof to Rosetta Stone. Id. “In our view, once the owner of a famous mark establishes a prima facie case of dilution by blurring or tarnishment, it falls to the defendant to demonstrate that its use constituted a fair use . . . .” Id. (citation and quotation marks omitted).
The district court had granted summary judgment against Rosetta Stone’s dilution claim on a second ground, finding that Google’s use of Rosetta Stone’s marks could not have diluted these marks because their public recognition increased greatly in the relevant time period. Such a per se rule, the Fourth Circuit held, “was error.” Id. at *21. The district court had placed too heavy an emphasis “upon whether there had been any actual injury suffered by Rosetta Stone’s brand,” failing to take into account the other likelihood of dilution factors enumerated in the Lanham Act. Id. at 21.
This may, however, turn out to be a Pyrrhic victory for Rosetta Stone. The Fourth Circuit spotted a fact issue that may be fatal to the dilution claim on remand, namely, when Rosetta Stone’s mark became famous. Id. at 22. Under the federal dilution act, fame must be established prior to the diluter’s first use of the mark. Here the district court found that Rosetta Stone’s mark had become famous by 2009. Yet Google’s sale of the mark as a keyword had, according to Rosetta Stone’s allegations, caused a likelihood of dilution since 2004. The Fourth Circuit remanded for a more precise determination when the mark became famous. If that date is after Google’s first allegedly dilutive use, Rosetta Stone’s dilution claim must fail. Id. at *23.
- Further Thoughts
The legal principles underlying the Fourth Circuit’s decision are not necessarily limited to claims against search engines. They may also apply to claims against individual advertisers who use a party’s trademarks as keywords to trigger sponsored links. Thus, we may well see an uptick in litigation brought by trademark owners not only against search engines, but also against competitors, counterfeiters, and others using sponsored links to sell their goods or services.
Search engines and individual advertisers may choose to reshape their online advertising practices as a result of this decision.