–Dan Kelly, Attorney

Last month, the United States Court of Appeals for the Ninth Circuit issued an opinion in the case of Toyota Motor Sales, U.S.A. v. Tabari that asked whether the domain names buy-a-lexus.com and buyorleaselexus.com used in connection with automobile brokerage services infringed Toyota’s trademark rights in LEXUS.

Conventional legal wisdom is that only the owner of a trademark has a right to use its trademark in a domain name in connection with related goods or services.  The back-of-the-envelope legal calculus is not difficult:  the domain names incorporate LEXUS in its entirety, and they are used in connection with auto brokerage services–services that are closely related to automobiles.  There is only one catch:  the brokers legally deal in genuine LEXUS vehicles.  Astute readers will recognize this fact as raising the issue of nominative fair use.

The Court articulated its nominative fair use test this way:

In cases where a nominative fair use defense is raised, we ask whether (1) the product was “readily identifiable” without use of the mark; (2) defendant used more of the mark than necessary; or (3) defendant falsely suggested he was sponsored or endorsed by the trademark holder.

While correct, I prefer how Steve has articulated the nominative fair use defense, which is established when:

  1. The product cannot be readily identified without using the trademark;
  2. Only so much of the trademark is used as is necessary for the identification; and
  3. No sponsorship or endorsement of the trademark owner is suggested by the use.

I bring this up, because the Court’s analysis began “by asking whether the Tabari’s use of the mark was ‘necessary’ to describe their business.”  In my initial read, I thought that the Court was beginning with the second element of the test, keying off of the word “necessary,” but this is a restatement of the first element, which addresses the issue of necessity in using the mark.  The second element, while using the word “necessary,” really addresses the scope of the use.

The Court recognized that it was not necessary in an absolute philosophical sense for the Tabaris to use buy-a-lexus.com or buyorleaselexus.com; it observed that they could have just as easily used autobroker.com or fastimports.com (Fast Imports being the Tabaris’ d/b/a).  But here’s the clincher:  “One way or the other, the Tabaris need to let consumers know that they are brokers of Lexus cars, and that’s nearly impossible to do without mentioning Lexus, . . . be it via domain name, metatag, radio jingle, telephone solicitation or blimp.”  (It is fair to characterize this last quip as dicta.)

Lawyers and fans of legal minutiae can read after the jump for one other legal issue raised by this case that has me puzzled.

The Court pointed out several times that under 9th Circuit law, the traditional multi-factor likelihood of confusion test does not apply when a defendant raises the nominative fair use defense.  “[N]ominative fair use ‘replaces’ Sleekcraft as the proper test for likely consumer confusion whenever defendant asserts to have referred to the trademarked good itself.”  I found this surprising, and it strikes me as wrong.

Unless I am missing some nuance (and I have not undertaken more extensive research), the nominative fair use elements test only that defense, not likely confusion.  A failure to establish nominative fair use does not thereby establish likely confusion, yet according to this case, that is the result.  It seems to me that there are cases where a defendant’s use would fail the nominative fair use defense (say, for instance, because the product can be readily identified without using the trademark), and yet still be unlikely to cause confusion.

Perhaps one of our colleagues more familiar with 9th Circuit law can shed some light on this issue.

  • Anonymous

    There is no statutory “nominative fair use” defense – the only statutory fair use defense is 1117(b)(4), into which nominative fair use clearly does not fall. Instead, nominative fair use is a name given to a particular character of use of the plaintiff’s mark, that is, where the mark is used to refer to the plaintiff’s own goods and services, not the defendant’s. Since nominative fair use is not an affirmative defense there is only one legal issue, thus only one test, which is whether there is likelihood of confusion. So in a nominative fair use fact pattern, rather than using the Sleekcraft factors (which don’t function well in the nominative context), confusion is tested instead under the nominative fair use test. It’s just an alternative test for confusion in a particular use case.
    I don’t think a court has characterized is as such, but it is as if confusion is presumed if the nominative fair use test isn’t met. A nominative fair use will always lose under the Sleekcraft factors. Similarity of the marks, proximity of the goods, type of goods, intent, will always have to be in the plaintiff’s favor, since you’re talking about the plaintiff’s own mark and goods and services. So an alternative test arose that is more apt for these types of uses, identifying parameters within which there is unlikely to be confusion – necessity of use of the mark, using as little as necessary, and no suggestion of sponsorship or endorsement. Where those conditions are satisfied, consumers aren’t going to be confused. And in the absence of one of these elements, it’s not adequately clear to consumers that the use is nominative so instead they will just be confused.

  • Anonymous,
    Thanks for the assistance–makes sense. The key premise to get to the NFU analysis is that the defendant’s goods are in fact genuine (or the defendant is providing goods or services related to genuine goods or services). If not, then as you point out, it is likely a question of traditional infringement or passing off.
    I have a feeling that if I follow the cases cited in the Tabari opinion upstream, I’ll eventually get to a thorough analysis by the 9th Cir.