June 28, 2005
Important 2d Circuit Adware Case--1-800 Contacts v. WhenU
1-800 Contacts, Inc. v. WhenU.com, Inc., Docket Nos. 04-0026-cv and 04-0446-cv (2d Cir. June 27, 2005).
Overshadowed by yesterday’s Grokster mania, the Second Circuit finally issued an important ruling about WhenU's liability for trademark infringement.
The court found that WhenU was not liable for trademark infringement as a matter of law: "We hold that, as a matter of law, WhenU does not 'use' 1-800's trademarks within the meaning of the Lanham Act, 15 U.S.C. 1127, when it (1) includes 1-800's website address, which is almost identical to 1-800's trademark, in an unpublished directory of terms that trigger delivery of WhenU's contextually relevant advertising to C-users; or (2) causes separate, branded pop-up ads to appear on a C-user's computer screen either above, below, or along the bottom edge of the 1-800 website window."
By rejecting 1-800 Contact's case for lack of a trademark "use," which is a precondition for any trademark infringement liability, all other aspects of the trademark case (likelihood of confusion, defenses) were moot. In other words, if the plaintiff can't establish trademark use, then trademark infringement defendants are entitled to summary judgment regardless of any other alleged facts. Thus, the appellate court did not remand the case to the district court for further fact-finding; instead, 1-800 Contacts loses as a matter of law, and 1-800 Contact's trademark claims are completely dead unless they want to appeal the case to the Supreme Court (which would strike me as an odd move here). Because 1-800 Contacts already lost its copyright case, I think their case is effectively dead as well.
What Does This Case Mean?
I think the opinion is generally great. The lower court opinion was truly awful, and the Second Circuit clearly and unambiguously rejected that opinion. In particular, the court gave us lots of insights into what constitutes trademark "use" in the Internet keyword context. There has been considerable confusion on this very question, and the Second Circuit's opinion will be persuasive precedent in all future cases throughout the nation.
The court also understood that this case involved important issues about how consumers look for wanted information and pick preferred tools to do so. Specifically, the court derisively rejects WhenU’s repeated exhortations that WhenU’s pop-up ads are “unauthorized.” The court emphatically states that contemporaneous displays of software windows on a user’s computer screen do not need to be authorized by a trademark owner. The court then notes that WhenU’s pop-up ads are authorized because the user downloaded the software. The court fully appreciates that 1-800 Contacts’ arguments had the potential to take choices away from consumers about how they manage their computer desktop and windows, and the court correctly shuts down this anti-consumer effort.
Amidst the abundance of good news, I do have a few minor twinges of disappointment. Most notably, the court did not discuss the "initial interest confusion" doctrine at all. This doctrine has junked up Internet jurisprudence since 1999, and the lower court opinion exemplified exactly how courts misuse the doctrine. Thus, I was hoping that the Second Circuit would confront this issue and, optimistically, give us a clear statement of the doctrine. That clear statement will have to wait until another day.
Also, the opinion is plagued by several odd arguments that mostly seemed designed to limit the case to WhenU's facts. As a result, this opinion may offer very limited utility to search engines regarding their practices of selling keyword-triggered ads. This is unfortunate because we desperately need clarity on this topic, and the court's reasoning easily could have extended to search engines.
Finally, I fear this case will only exacerbate the trend where plaintiffs like 1-800 Contacts use their legislative influence to convince state legislators to pass protectionist/anti-consumer anti-adware laws like Utah and Alaska have done. Under the Second Circuit’s reasoning, the Utah anti-adware law is currently ineffective against adware vendors (at least, as applied to WhenU) because it requires trademark infringement as an essential element of the claim, and the lack of trademark use means that WhenU is not committing trademark infringement as a matter of law. However, Alaska's anti-adware law has no such requirement--an omission that, I think, reinforces both its unconstitutionality and preemption by federal law—which may also lead other states to replicate Alaska's law as a "model."
Whether this battle is fought in courts or the legislatures, this opinion does not end the battle over keywords, adware or how law can help (or hurt) consumers in making choices. Nevertheless, this case will have a major impact in the Second Circuit and beyond, and fortunately the opinion reaches the right result and makes a number of great points in doing so.
What Constitutes Trademark “Use”?
The court unambiguously says that WhenU did not engage in trademark use. In doing so, it points to two previous cases, Wells Fargo v. WhenU and U-Haul v. WhenU, where district courts had found that WhenU had not engaged in trademark use. The lower court simply ignored this precedent, and the Second Circuit properly chided the judge for cutting this corner. As the court says: “the district court’s consideration of these two comprehensive decisions on the precise issue at hand was confined to a footnote in which it cited the cases, summarized their holdings in parantheticals, and concluded, without discussion, that it ‘disagree[d] with and [was] not bound by these findings.’….Unlike the district court, we find the thorough analyses set forth in both U-Haul and Wells Fargo to be persuasive and compelling.”
The court then discusses how keyword triggers and displaying pop-up ads do not constitute a trademark use:
Trademarked Keywords in the Database Used as Triggers
The court says that WhenU does not use 1-800 Contact’s trademarks when it puts the term in its database of keywords that trigger ads. The court properly notes that WhenU “does not ‘place’ 1-800 trademarks on any goods or services in order to pass them off as emanating from or authorized by 1-800.” Later, the court correctly says that including the term into WhenU’s database “does not create a possibility of visual confusion with 1-800’s mark.” Thus, the court understands that if consumers do not see or perceive the trademark in association with the defendant’s goods or services, there is no trademark use. This is a point that many courts simply overlook or misunderstand, and the Second Circuit gets it 100% correct here.
However, in support of this point, the court makes an odd distinction between “www.1800contacts.com” and the trademark “1-800Contacts.” The court says that www.1800contacts.com “functions more or less as a public key to 1-800’s website.” I have no idea what the court means by this. The court could have said that the term www.1800contacts.com is functional and thus is not entitled to trademark protection, but the court doesn’t say that. If it did, we would have a hard time understanding why the trademark “1-800Contacts” isn’t also functional (it’s the “public key” to calling the company).
I think the court is trying to distinguish between using words as trademarks and using words for other cognitive or semantic meanings. I support such a distinction, but it is a difficult distinction to make. This brings to mind the attempted distinctions made by the district court in Playboy v. Netscape. In that case, the search engine was triggering ads on the words “playboy” and “playmate,” and the lower court said that the search engine was using the words as dictionary words, not for their trademark meaning. I think this subjective inquiry is dangerous (I would rather focus on what consumers understand the words to mean), and the distinction only applies to trademarks that are also dictionary words (i.e., not fanciful terms).
Is the Second Circuit trying to revitalize that distinction? It seems so. The court says that WhenU is using the term www.1800contacts.com “precisely because it is a website address” while for WhenU to have capitalized on the trademark’s fame/recognition, WhenU would have had to include the exact trademark (1-800Contacts) in the database. Thus, arguably the court could, in the future, distinguish any other keyword in the WhenU database that aren’t a URL. The court recognizes this potential but sidesteps it. In FN 11, the court says that it doesn’t opine on whether including the exact trademark in the database is necessarily infringing. Too bad—we’d like to know the answer to that question!
The court then further limits its reasoning to WhenU-specific facts when it notes that WhenU customers cannot buy specific keywords. Instead, WhenU sells topical categories, so customers cannot “request or purchase specified keywords to add to the directory.” This distinguishes WhenU from other adware companies, but it also distinguishes WhenU from all search engines that sell individual keywords. I don’t think it’s fair to read the opinion for the converse proposition—that selling individual trademarked keywords is a trademark use—but the opinion arguably leaves that possibility open.
Finally, in a footnote, the court takes a dig at 1-800 Contacts for duplicity. The court notes that 1-800 Contacts has bought keyword-triggered ads in adware using trademarks of its competitors (including the advertiser defendant in this case, Vision Direct). Note to plaintiffs: courts don’t like it when you engage in the same behavior that you claim caused you harm.
Placing Pop-Up Ads Over Plaintiff’s Website
The court says that placing pop-ads on 1-800 Contacts’ website is not a trademark use because the ads do not display the 1-800 Contacts’ trademarks. This is a confusing statement because some ad copy could have displayed 1-800 Contacts’ trademarks (if only to make comparative advertising, or perhaps more confusingly). It’s not entirely clear how the court can make a sweeping factual statement about the ads’ contents.
However, the court might not care about the ad copy, at least for purposes of WhenU’s liability. WhenU’s ads were labeled as coming from WhenU, so the court might be saying that WhenU didn’t display the trademarks itself and therefore the court cares less what was in the advertiser’s copy. Note that Google’s contributory liability for ads that contained the GEICO trademark is the remaining issue in the GEICO v. Google case.
The court continues by arguing that the pop-up ads are not directly triggered by 1-800 Contacts’ trademarks. In support of this argument, it again makes the distinction between a URL and the trademark, but it also says that the pop-up ads could have triggered by a number of terms other than 1-800 Contacts’ trademarks, including “contacts” and “eye care.”
In my mind, there are lots of reasons why the pop-up ad could have appeared, and we can’t make any assumptions about what happened, why it happened, or what consumers expected. The court fully gets this point. It notes that WhenU users receive ads “in a myriad of contexts,” and then in FN 14 it fully embraces the point: “1-800’s claim that C-users will likely be confused into thinking that 1-800 has sponsored its competitor’s pop-up ads is fairly incredulous given that C-users who have downloaded the SaveNow software receive numerous WhenU pop-up ads – each displaying the WhenU brand – in varying contexts and for a broad range of products.”
Reinforcing that simply overlaying pop-up ads on plaintiff’s website isn’t trademark use, the court analogizes to offline trademark “adjacencies”—such as how retail stores put their house brand next to branded products on store shelves. I have a lot more to say about adjacencies in a future paper. However, I think the court is 100% correct that adjacencies should not be a trademark “use,” and I further think the court understood that the pop-up ad is just another form of adjacency.
Finally, the court again notes that WhenU’s practices differ from other adware companies and search engines because WhenU sells categories, not individual keywords. As with the prior discussion, I would not read this opinion for the converse proposition (i.e., selling individual keywords is a use), but the liability seems to be left open.
I worked with the Electronic Frontier Foundation to file an amicus brief in favor of WhenU. Although the brief didn't get cited, it may have been helpful to the court to recognize the important social issues implicated by the case and to see how the lower court opinion was a travesty that needed to be soundly rejected. I have also written a lengthy law review article criticizing the lower court opinion (among others) and making several arguments that dovetail well with this ruling.