Funky Ninth Circuit Opinion on Domain Names and Nominative Use–Toyota v. Tabari

By Eric Goldman

Toyota Motor Sales, U.S.A., Inc. v. Tabari, 2010 WL 2680891 (9th Cir. July 8, 2010)

Every time I see a federal appellate opinion on domain names, I’m vaguely reminded of the Country Joe song I-Feel-Like-I’m-Fixin’-To-Die Rag, whose chorus goes “And it’s one, two, three, what are we fighting for?” Fortunately, domain name disputes do not lead to the senseless loss of life we experienced from the Vietnam War. Unfortunately, lengthy domain name litigation usually has little more strategic value. Invariably, the domain name litigation has less to do with rational economic decision-making and more to do with chest-beating and posturing.

I bring this up because the Ninth Circuit’s latest domain name opinion involves litigation that makes no financial sense for either side. The Tabaris are independent auto brokers that help their customers find and buy Lexus vehicles from an authorized Lexus dealer. They run a business called Fast Imports from the domains and

What is Lexus’ problem with those domain names? The Tabaris are helping people buy Lexuses, so Lexus is going to get its fair share no matter what. The appellate opinion did not indicate that the Tabaris are crooks or trying to divert Lexus customers to other brands. So Lexus, why sue your friends? The opinion hints that Lexus was trying to improve dealer relations by squelching a broker who plays dealers off each other, but hey, that’s fair competition.

From the Tabaris’ perspective, losing these domain names should not be intrinsically fatal to their business. The Tabaris could set up shop at any number of other domain names, in which case they would lose only the built-up clicks from existing links to the site (I wonder how many of those there were in this case) and any extra Google juice from having a seasoned domain name with the trademark in it. I always find it weird when appellate courts treat a defendant’s domain name as the dispositive linchpin of communication between interested parties rather than just one of many SEO tools.

Refreshingly, this opinion does not overestimate the domain name’s value. However, it doesn’t see any reason to consider a switch either: “the Tabaris needed to communicate that they specialize in Lexus vehicles, and using the Lexus mark in their domain names accomplished this goal. While using Lexus in their domain names wasn’t the only way to communicate the nature of their business, the same could be said of virtually any choice the Tabaris made about how to convey their message.”

While the opinion focuses on domain names, the Tabaris’ websites also, at some point, used copyrighted Lexus photos and displayed the Big L logo. Normally, a photo rip and unauthorized logo display will get a district court judge to rule in favor of the IP owner. Before Lexus sued, the Tabaris cleaned up those issues, so the Ninth Circuit panel focuses solely on the two domain names (because an injunction was the only remedy at issue). This is a logical move by the Ninth Circuit, but most courts will not be so forgiving of sites that borrow the official logo and copyrighted photos.

With the Tabaris’ use of the two domain names in their auto brokerage business the only issue on appeal, this should be an easy call per the nominative use doctrine. However, the words “easy” and “nominative use doctrine” go together like peanut butter and artichokes. Personally, I still have no idea when businesses outside a manufacturer’s authorized channel can legally include the manufacturer’s trademark in their name. Each case seems to be sui generis.

To segregate legitimate from illegitimate uses of third party trademarks in domain names, the opinion lays out a surprisingly lucid taxonomy with 3 categories of presumptively illegitimate domain names:

1) “When a domain name consists only of the trademark followed by .com, or some other suffix like .org or .net, it will typically suggest sponsorship or endorsement by the trademark holder.” This makes sense intuitively, but (A) the court doesn’t address the seemingly contradictory Lamparello case, and (B) the opinion’s reasoning remains predicated on dicey assumptions about consumer search behavior, such as consumers typing in into their web browser address bar—an assumption that has grown dicier with the rise of omniboxes.

2) “Sites like, or will also generally suggest sponsorship or endorsement by the trademark holder.”

3) “domains like or affirmatively suggest sponsorship or endorsement by the trademark holder and are not nominative fair use”

By implication, other domain names generally should be eligible for nominative use. At minimum, and should be fair game for resellers and related parties like buying agents. In support of this, the court rejects Lexus’ argument that there was something untoward about the Tabaris brokering other auto manufacturers if their customers decided they didn’t want a Lexus. For more on this, see my Brand Spillovers article.

The opinion suggests that the following domain names should qualify for nominative use or otherwise be permissible as well:






* [although I wonder about dilution with these two]


Procedurally, the opinion addresses several key issues about the interaction between the nominative use test and the likelihood of consumer confusion test. The opinion says that an evaluation of consumer confusion is implicitly built into the New Kids on the Block nominative use test. Therefore, “if the nominative use satisfies the three-factor New Kids test, it doesn’t infringe” without needing to consider the likelihood of consumer confusion test at all. Thus, “nominative fair use ‘replaces’ Sleekcraft as the proper test for likely consumer confusion whenever defendant asserts to have referred to the trademarked good itself.” Further, once a “defendant seeking to assert nominative fair use as a defense…show[s] that it used the mark to refer to the trademarked good,” the trademark owner bears the burden of disproving nominative use. All of these procedural points have been hotly contested in prior cases.

The court concludes that the district court’s injunction against the Tabaris using “Lexus” in domain names was too broad and remands the case to the district court to try again. Although the court doesn’t tell the district court exactly what to do, it does indicate: “At the very least, the injunction must be modified to allow some use of the Lexus mark in domain names by the Tabaris.”

This is a rich and multi-faceted opinion written in a confident and emphatic style…perhaps too emphatically, as the opinion swings around like a bull in a china shop, breezily overturning or sidestepping numerous 9th Circuit precedents on both domain names and nominative use. Were this opinion to become the definitive 9th Circuit statement on either domain names or nominative use, this case would be a landmark opinion. However, the 9th Circuit’s Internet trademark jurisprudence has awkwardly accreted on a case-by-case basis for more than a decade, and I doubt this opinion will meaningfully affect the next 9th Circuit panel’s considerations.

Even so, this case has to be good news for shopbots. Although the Tabaris were “manual” shopping agents, the case’s reasoning should apply equally well to all shopbots comparison search engines and review sites that use third party trademarks as part of their taxonomy. These sites regularly get nastygrams from trademark owners. It will be interesting to see if this case helps turn that tide.

A final oddity: Judge Kozinski wrote both this opinion and the recent eVisa decision. Although the opinions involve different trademark doctrines applicable to domain names (a nominative use defense instead of dilution), their spirit couldn’t be more different. The eVisa case was decidedly pro-plaintiff, while this opinion is very defense-favorable. I wonder if Kozinski bent over backwards to help a pro se litigant (the Tabaris represented themselves), or perhaps Lexus’ anti-competitive intent set him off. Otherwise, although the split opinions in theory can be harmonized on numerous bases, they struck me as schizophrenic.

More comments from Rebecca Tushnet (smart and challenging, as always—especially about the numerous empirical deficiencies in the opinion), Ryan Gile and Tom O’Toole.