“[A] trademark holder has no right to police ‘unnecessary’ use of its mark. Whether necessary or not, a defendant's use of a mark must be confusing in the relevant statutory sense for a plaintiff to raise a viable infringement claim.” It’s sad that this requires reiteration, but it’s nice to have a clear statement.
OK, the shiny details: Swarovski brought a terrible lawsuit because it didn’t like the idea that a downscale, off-price retailer like Building # 19 was selling genuine Swarovski goods in its low-rent digs, and convinced a district court that suppressing Building # 19’s truthful advertising was somehow a legitimate use of trademark law. A year later, the First Circuit reversed.
Initially, Building # 19, which seems like a retailer with a sense of humor about its position as a seller of salvage/overstock/discontinued merchandise, lawfully obtained a number of Swarovski figurines and promoted their sale with an ad using “Swarovski” in large print font. The figurines came from an insurer’s salvage sale after a severe storm damaged the warehouse where they’d been stored; the figurines themselves were apparently free from damage and came with Swarovski Certificates of Authenticity. The salvor placed no restrictions on Building # 19’s ability to use the Swarovski name or label. After a hearing, the district court orally granted a preliminary injunction limiting the use to a much smaller font.
After Swarovski filed its lawsuit, Building # 19 agreed to voluntarily refrain from further advertising or sale of its Swarovski crystal, but then proposed to run an ad reading “ONE DAY EVENT 11AM to 8PM, Tornado Hits Warehouse containing GENUINE SWAROVSKI(R) CRYSTAL Collectibles,” again with “Swarovski” in extra-large, capitalized, bold and distinctive font. A smaller disclaimer read: “Disclaimer: Building # 19 is selling GENUINE SWAROVSKI(R) CRYSTAL products BUT Building # 19 is NOT an authorized dealer, has no affiliation, connection or association with SWAROVSKI(R) and the standard SWAROVSKI(R) limited warranty is not available to our customers. (The lawyers told us we should add that big ‘but’....).”
Swarovski moved for a preliminary injunction, which the court granted only to the extent that the capitalized word “Swarovski” at the top of the proposed advertisement could be no larger than the font used for the name “Swarovski” in the disclaimer at the bottom of the proposed advertisement. (Bonus question: how good was Building #19's designing around the injunction?)
Building # 19 appealed on the ground that the district court didn’t find that its use was likely to confuse consumers or that Swarovski would suffer irreparable harm. The court of appeals agreed and remanded for further findings.
This was a nominative use case: Building # 19 was using the mark to refer to the product. (In a footnote, the court stated that nominative use can also include “situations where the defendant ultimately describes his own product, so long as his use of the mark is in reference to the plaintiff's product,” such as a mechanic who specializes in Volkswagens.) “The potential for confusion in a nominative use case is not one of source—here, the crystal really was manufactured by Swarovski—but rather one of endorsement or affiliation. The fear is that a consumer glancing at Building # 19's proposed advertisement might mistakenly believe that Swarovski had some official association with the sale; perhaps that Swarovski sponsored the sale and so stood behind the goods as a direct seller, or that it had partnered with Building # 19 in a way that might detract from its luxury status.” NB: this is why we need a materiality requirement. Whose fear is this and why should we care?
In what I hope doesn’t create another epicycle in the trademark universe, the court continued, “[b]ecause this kind of confusion does not implicate the traditional ‘source-identification function’ of a trademark, other circuits, most notably the Ninth and the Third, have developed a distinct ‘nominative fair use’ analysis to identify unlawful infringement in cases like this one.” Details differ, but the factors are: (1) whether the plaintiff's product was identifiable without use of the mark; (2) whether the defendant used more of the mark than necessary; and (3) whether the defendant accurately portrayed the relationship between itself and the plaintiff. The First Circuit recognizes the underlying principle, but has never endorsed any particular version of the doctrine.
The district court didn’t make any finding on likely confusion, and its reasoning didn’t allow the court of appeals to infer such a finding. More was required: “Swarovski may not charge infringement against all unauthorized uses of the ‘Swarovski’ name, but only those uses likely to cause consumer confusion, mistake or deception.” The burden of such a showing is “firmly” on the plaintiff. The district court, however, first noted that the traditional multifactor confusion test was a poor fit and then found that only three factors favored Swarovski (Building #19 was using the Swarovski name, though not the logo; it was selling authentic Swarovski products; and the mark was strong—how authenticity could favor Swarovski is an exercise left for the reader, because G-d knows I have no idea). It concluded that it was “not so sure that Plaintiff met its burden under the Pignons test.” (Marketing channels favored Building #19; there was only one piece of disputed, arguably hearsay, confusion evidence; and in terms of intent to adopt the mark, Building #19 had no such intent, since it was simply describing the items it was selling.)
The district court then turned to nominative fair use, but it didn’t explain “if it was using those factors to measure the likelihood of consumer confusion.” It described the factors as (1) whether the product was readily identifiable without use of the mark, (2) whether the defendant used more of the mark than necessary, and (3) whether the defendant falsely suggested sponsorship or endorsement. The first and third factors weighed in favor of Building #19; as the court of appeals summarized, “the proposed ad made clear the product's origin, how it came into the hands of Building # 19 and that the sale was not sponsored by Swarovski.”
Thus, the decision against Building #19 turned on the second factor. The district court thought the font was too large: “the name Swarovski is larger than any other font in that ad. It's larger than the name Building # 19. It's larger than the words ‘one-day event.’ It is clearly, I think, more use of the mark than is necessary to make the point.”
This was insufficient:
This analysis does not mention consumer confusion, but merely decided that Building # 19 used “more of the mark than necessary” to effectively communicate its message. But as we have explained, a trademark holder has no right to police “unnecessary” use of its mark. Whether necessary or not, a defendant's use of a mark must be confusing in the relevant statutory sense for a plaintiff to raise a viable infringement claim.
Unfortunately, the court of appeals then proceeded to speculate that the mere use of the word Swarovski might have created confusion: “it might have made the ad look like one from Swarovski rather than Building # 19; it might have lead viewers to believe that Swarovski endorsed or sponsored the sale; or it might have suggested some official affiliation between Swarovski and Building # 19.” But the district court didn’t suggest that any of those things had happened. It might have been led astray by the different versions of nominative use applied by different circuits, given the Ninth Circuit’s holding that the test “replaces” traditional likely confusion analysis contrasted to the Third Circuit’s characterization of the test as an affirmative defense. The district court stated that Building # 19 “has some issues with its burden under the nominative fair use test,” suggesting the Third Circuit approach, but that means nothing if there’s not first a showing of likely confusion.
The court of appeals declined to pick an approach, but reiterated that “a trademark defendant has no burden to prove anything until the plaintiff has first met its responsibility to show infringement by demonstrating that the defendant's use of its mark is likely to confuse consumers.” Thus the entry of a preliminary injunction was in error.
This problem was related to the absence of an irreparable injury finding, which is required even if there’s likely success on the merits. Swarovski pointed to evidence at the hearing “that its luxury brand reputation would suffer as a result of its association with Building # 19.” But there was no formal finding of irreparable harm, as you’d expect in the absence of a confusion finding. Swarovski argued that trademark plaintiffs who demonstrate likely confusion get a presumption of irreparable harm. But eBay “threaten[s] the continuing viability of this presumption.” The eBay concurrences, joined by a total of seven Justices, suggested that such rules of thumb could survive as “lesson [s] of ... historical practice” that might inform the district courts' equitable discretion “when the circumstances of a case bear substantial parallels to litigation the courts have confronted before.” Still, there’s a bar on “general” or “categorical” rules, and there’s a remaining question whether that applies to a presumption of irreparable harm in trademark cases, as other circuits have held with respect to patent and copyright cases.
The court of appeals declined to decide the question here. No matter what the general rule, it’s hard to see irreparable harm without a finding of confusion. In further proceedings, the district court might conclude “that Swarovski has presented sufficient evidence to establish irreparable harm to its luxury reputation without the benefit of the presumption, or that the facts of this case are so similar to past cases ‘that a presumption of irreparable injury is an appropriate exercise of its discretion in light of the historical traditions.’” (Ugh on the first, but that’s really a problem descending from the ridiculous concept of confusion in play here.)