Friday, August 21, 2020

Tiffany's blues: Costco gets a trial on sales of "Tiffany" rings with Tiffany setting

Tiffany & Co. v. Costco Wholesale Corp., Nos. 17-2798-cv, 19-338, 19-404, 2020 WL 4743020, -- F.3d – (2d Cir. Aug. 17, 2020)

The district court found that Costco’s sales of otherwise unbranded diamond engagement rings with the solitaire setting known as a Tiffany setting, identified by point-of-sale signs containing the word “Tiffany,” constituted infringement and counterfeiting. It awarded Tiffany trebled profits, prejudgment interest, and punitive damages (under NY law) totaling over $21 million. The court of appeals held that summary judgment should not have been granted to Tiffany on liability and remanded for trial, including on Costco’s descriptive fair use defense.

Of note: (1) “Use as a mark” continues its wild and not particularly coherent ride in the US. It’s not a requirement for infringement liability in the Second Circuit but it is, we learn, a requirement for counterfeiting liability. (2) Consistent with Booking.com, the court isn’t particularly interested in the difference between generic and descriptive—among other things, the court indicates that generic use can constitute descriptive fair use. This makes sense given the justification for the defense (as well as the historic use “common descriptive name” to identify generic terms). (3) But on the flipside, the court seems relatively uninterested in what I would have thought was the very important fact that, according to the facts recited, there is no other name to identify the setting at issue other than “Tiffany setting.” Even so, the court indicates that there could be infringement if consumers are confused. (4) Relatedly, the court’s suggestion at the end of its fair use discussion that the descriptive fair use defense could fail if consumers are confused seems inconsistent with KP Permanent. Descriptive fair use, in its current form, supposedly weighs something else against likely confusion—perhaps whether the competitor is objectively behaving reasonably given the descriptiveness of the term at issue, even if an ordinarily sufficient number of consumers are confused.

Some other relevant facts: Tiffany has a number of relevant trademark registrations. Costco used “Tiffany” only in connection with rings with Tiffany settings; it sold several other styles identified by similar point-of-sale signs, each of which indicated the name of the corresponding ring’s setting style and none of which used the word “Tiffany.” Also:

At some point in the late nineteenth century, Charles Lewis Tiffany developed and sold an engagement ring incorporating a particular style of six-prong diamond setting. Since that time, numerous advertisements, dictionaries, trade publications, and other documents have referred to diamond settings reminiscent of that style as “Tiffany settings.”

Costco used small, white point-of-sale signs that displayed information about the rings “(ostensibly including setting style)” in uniform black text, along with their prices in larger, bold text. “In general format, but not necessarily in size, these signs resembled other point-of-sale signs that identified virtually all items for sale throughout Costco stores.” Costco argued that it copied these signs directly from descriptions supplied by its vendors. Its signs for rings with a Tiffany setting “at times identified the ring setting using the phrases ‘Tiffany setting,’ ‘Tiffany set,’ or ‘Tiffany style,’ and at other times used only the word ‘Tiffany’ for that purpose,” and it “apparently identified other setting styles in a similarly variable manner.” Tiffany purported to challenge only the “Tiffany”-only uses, not those with “style,” “set,” or “setting.”

When Tiffany objected, Costco said that, within one week, it voluntarily removed all uses of the word “Tiffany” from the signs in its jewelry display cases. After Tiffany sued, Costco also sent a letter to all customers who had purchased engagement rings with Tiffany settings to alert them that Tiffany had sued. It explained that Costco’s point-of-sale signs had “used the word ‘Tiffany’ to indicate that [the associated] ring had a Tiffany-style pronged setting,” asserted that Costco “do[es] not believe [its] signs were inaccurate,” and reminded buyers that they could return their rings for a full refund at any time under Costco’s existing return policy. “Approximately 1.3% of customers who received this letter returned their rings to Costco.”

Infringement: A reasonable jury could find for Costco; the district court erred in finding otherwise, based on three factors. [Footnote on the multifactor test: “similarity of marks” is shorthand as defendant need not use a term “as a mark,” that is, “as a symbol to attract public attention,” to infringe.] The summary judgment standard is the same for trademark as for anything else, and also “insofar as the determination of whether one of the Polaroid factors favors one party or another involves a legal judgment—which it often does—we must review that determination de novo.” Thus, “in the majority of cases, we should review de novo both a district court’s determinations as to each Polaroid factor and its ultimate balancing of those factors.”

The key factors: (1) Actual confusion. The district court relied on (a) the deposition testimony of six Costco customers, each of whom alleged that he or she was confused by Costco’s point-of-sale signs and (b) Dr. Jacob Jacoby’s survey, which surveyed 944 people, 606 of whom identified themselves as Costco patrons who thought that “they [or their significant other] would consider buying a diamond engagement ring at Costco that cost at least $2,500.” Of the 606, each of whom was shown a photo of a diamond engagement ring alongside one of Costco’s point-of-sale signs, either in isolation or after seeing photos of other branded items sold by Costco—“more than two out of five ... were likely confused into believing that Tiffany & Co. was the source of the rings.” The district court found this evidence unrebutted, but that was error.

First, Costco argued that 6 out of the 3,349 customers who purchased Tiffany-set rings at Costco during the relevant period was only de minimis evidence, and it submitted a report from its own expert, Dr. Russel S. Winer, criticizing Dr. Jacoby’s survey methodology and results. Costco’s expert opined that Jacoby should have targeted only customers with a “present purchase interest in buying a diamond ring,” and that the survey respondents “could not have been a group whose perceptions provided any valid or reliable predictor of past or future Costco diamond ring purchaser beliefs.” In addition, he contended that the survey was fatally flawed due to “artificial, contrived and biasing” stimuli that “ignore[d] the reality of the customer purchase process.” Rather than showing survey responders a point-of-sale sign as it would have appeared to customers—in a display case surrounded by other rings identified by “other tags ..., some of which would have other setting and style types indicated”—Jacoby’s survey showed them only a single ring and sign in isolation. Indeed, Costco’s expert argued that, though both images (one with only a single point-of-sale sign and the other with a small section of a Costco display case) provided insufficient context, the relatively lower confusion among respondents who saw the latter was “striking.” A jury was entitled to consider “how the percentage [of customers identifying ‘Tiffany’ as a descriptive word rather than a brand] might have changed had the context provided been comparable to the real world experience.”

In addition—comment: highly relevant given the live dispute over the genericness of “Tiffany” for the name of the setting—Costco’s expert contended that the screening questions, which required responders to sort words into brand names and descriptive words, “trained” those responders to “bias[ ] the responses in favor of selecting Tiffany as a brand identifier.” To the extent that this nudged the numbers upward, that’s relevant also to the descriptive fair use defense, as KP Permanent says that the amount of confusion can be relevant.

Costco’s evidence was sufficient to raise a question on actual confusion. The district court had reasoned that Winer’s criticisms went to weight rather than admissibility, and that he didn’t perform his own survey. But the weight to be given to a particular piece of evidence “can be determinative of whether the moving party is entitled to summary judgment or whether a jury could find a material fact favorable to the non-moving party.” A reasonable jury could find that Tiffany failed to present sufficiently persuasive evidence to meet its burden.

(2) Good faith: The question is whether the defendant attempted “to exploit the good will and reputation of a senior user by adopting the mark with the intent to sow confusion between the two companies’ products.” Although “where the allegedly infringing mark is identical to the registered mark, and its use began subsequent to the plaintiff’s trade-mark registration, the defendant must carry the burden of explanation and persuasion,” it is still the case that “[p]rior knowledge of a senior user’s trade mark does not necessarily give rise to an inference of bad faith and may [actually] be consistent with good faith.” Indeed, “the intent to compete by imitating the successful features of another’s product is vastly different from the intent to deceive purchasers as to the source of [one’s own] product.” And “subjective issues such as good faith are singularly inappropriate for determination on summary judgment.”

The district court concluded that “no rational finder of fact could conclude that Costco acted in good faith in adopting the Tiffany mark.” The evidence it cited included an email from a Costco employee indicating that Costco’s jewelry boxes should have a more “Tiffany or upscale look”; the deposition testimony of a Costco inventory control specialist who acknowledged that she took no action in response to two emails ostensibly indicating customer and employee confusion over the source of Costco’s rings; and photographs and emails that suggest efforts by Costco to “copy Tiffany’s designs by making references to Tiffany designs and sharing links to Tiffany’s website [in communications with vendors].”

Both Tiffany and Costco also pointed to a customer email asking whether the word “Tiffany” on the label referred to the “Tiffany setting or Tiffany brand.” It was consistent with good faith that a Costco employee responded to the customer’s email that “[i]t means Tiffany setting.” And importantly, “the jury could also consider this email as evidence of an absence of actual customer confusion,” since inquiries about a potential relationship are not actual confusion and arguably show lack of confusion.

Costco’s contrary evidence indicated that it had never attempted to adopt the Tiffany mark; that its signs actually used the word “Tiffany” as a brand-independent description of a particular style of diamond setting; and that those signs merely reflected information provided by its own suppliers.

The district court thought that no reasonable jury could believe Costco, but the court of appeals disagreed. “We have consistently recognized that intent to copy a product’s useful, nonprotected attributes should not be equated automatically with an intent to deceive. Therefore, Costco’s admitted intent to sell jewelry that looks like Tiffany’s—as opposed to an intent to have its jewelry pass as Tiffany’s—cannot be enough to justify a finding that Costco acted in bad faith.” There was “substantial” evidence favoring Costco, including declarations from a diamond buyer and an assistant general merchandise manager affirming that Costco inventory control personnel took the term “Tiffany” directly from vendor descriptions, that the representatives understood Tiffany as a “generic style name,” and that indeed it was “the only name ... used to denote [that] type of pronged setting.” There was supporting evidence that the term “Tiffany” “has been used as a generic descriptor—both explicitly in conjunction with a word like “setting” and implicitly by itself—in thousands of advertisements, dictionaries, trade publications, and other public documents since the late 1800s.”

Costco also provided evidence that its rings were not branded with Tiffany’s mark (and indeed were branded with its supplier R.B. Diamond’s logo instead); that the rings came in “unbranded containers bearing no resemblance to Tiffany’s distinctive robin’s-egg blue packaging”; that buyers received Costco-branded receipts, appraisal forms, and other sales documents; and that Costco’s return policy permitted customers to return their rings any time after purchase. A reasonable jury could conclude that Costco’s that signs “were the product of a good-faith attempt to communicate to its customers the setting style of certain rings that it sold.” A jury could also infer good faith from Costco’s voluntary cessation and communications to purchasers.

Tiffany’s only rejoinder was that, after Costco voluntarily stopped using the name “Tiffany,” it was still able to describe those rings’ settings as “Solitaire.” “But ‘Solitaire,’ which describes any single gem in a simple setting, is undeniably a less descriptive term than ‘Tiffany,’ which ostensibly describes a specific type of six-prong setting.” If a seller claimed rights in “spoon” for a spoon, we wouldn’t think it sufficed to allow other sellers to advertise their “utensils.”

(3) Consumer sophistication: Although sophistication may usually be proven by direct evidence, including expert opinions or surveys, “in some cases a court is entitled to reach a conclusion about consumer sophistication based solely on the nature of the product or its price.” Jacoby’s study assumed that respondents who said they “would consider buying a diamond engagement ring at Costco” were representative of relevant customers, and Tiffany offered no other evidence. Costco’s expert countered that the purchase of an engagement ring is a “high involvement” transaction, and that actual purchasers—as opposed to those who merely “would consider buying an engagement ring”—possess substantial “subject matter knowledge and familiarity with the relevant vocabulary.”

The district court nevertheless found no factual question, as Costco’s evidence only went to the weight of Tiffany’s evidence and wasn’t affirmative evidence of consumer sophistication. This wrongly shifted the burden to Costco on Tiffany’s motion for summary judgment in its favor, and ignored Costco’s evidence in the form of its expert’s declaration. “A jury could reasonably conclude, by crediting Costco’s evidence and rejecting Tiffany’s, that the relevant population of consumers would be sufficiently attentive and discriminating as to recognize that Tiffany had nothing to do with Costco’s diamond engagement rings.”

Overall, a jury could reasonably conclude that the relevant consumers would know, or learn, that “Tiffany” describes a style of setting not unique to rings manufactured by Tiffany, and recognize that Costco used the term only in that descriptive sense. “Such consumers may also be distinctly capable of recognizing that Costco’s rings were not manufactured by Tiffany—based, for example, on their price, place of purchase, packaging, or paperwork—and consequently be particularly unlikely to be confused by any aspect of Costco’s point-of-sale signs,” especially given that they’d see “a jewelry case full of other unbranded rings, each identified by a sign indicating its own setting type in a similar or identical way.” Although there was clearly a “potential for confusion” inherent in the public association between the Tiffany brand and high-quality engagement rings, that wasn’t enough. In light of Costco’s evidence, as well as “Tiffany’s failure to demonstrate that actual purchasers would not recognize the word ‘Tiffany’ as denoting a commonly used setting style,” summary judgment for Tiffany was improper.

In addition (maybe), Costco was entitled to reach a jury on its descriptive fair use defense “even where the challenged material is likely to cause some confusion.” By “some” the court means “some otherwise actionable,” otherwise KP Permanent would be pointless.

The district court’s good faith finding above led it to reject Costco’s defense on the good faith prong; this was now reversed, and Costco also raised factual issues on the other elements.

(1) Use as a mark: Use as a mark means use “as a symbol to attract public attention,” or “to identify and distinguish ... goods [or services] ... and to indicate [their] source.” Relevant factors include: whether the challenged material appeared on the product “itself, on its packaging, or in any other advertising or promotional materials related to [the] product,” and the degree to which “defendants were trying to create, through repetition ... a[n] association between [themselves] and the [mark].”

A reasonable jury could find no use as a mark. “Tiffany’s own evidence indicates that Costco typically identifies the trademark associated with its branded products as the first word on the product label,” including when it did sell genuine Tiffany merchandise, whereas Costco produced hundreds of examples of signs for its engagement rings, “none of which began with the word ‘Tiffany’ or any other brand name.” Instead, it used the word “in the exact same manner (including typeface, size, color, and relative location on the signs) that it displayed setting information for other engagement rings.” “Tiffany” didn’t appear on any of its rings or ring packaging, and that the rings actually bore the logo of a different manufacturer.

(2) Descriptive use: This includes “words that describe a characteristic of the goods[ ] such as size or quality,” and also “words or images that more abstractly identify some information about the goods in question.” A jury could reasonably credit Costco’s evidence that “Tiffany” has a descriptive meaning independent of Tiffany’s brand and that Costco “intended to and did invoke that meaning when it created its point-of-sale signs.”

Tiffany argued that this result would be absurd because Tiffany is a valid mark for jewelry, and that the court shouldn’t allow “Tiffany” to be a source identifier for rings of other styles, but a descriptive term for rings in the so-called “Tiffany” style. But that’s not absurd; it’s inherent in the idea that a descriptive term can be a trademark. It’s black-letter law that “the public’s right to use descriptive words or images in good faith in their ordinary descriptive sense must prevail over the exclusivity claims of the trademark owner.” It doesn’t matter that Tiffany the company and the Tiffany setting derive their names from a common source; nor did it matter that “Tiffany” didn’t “inherently” describe the setting. For whatever reason, Tiffany didn’t stop “Tiffany setting” from becoming a well-known term [here again I note that the court doesn’t seem particularly concerned that there doesn’t appear to be another name for the setting, which I think is a pretty important issue]. “Indeed, the fact that Tiffany does not here challenge Costco’s use of the phrase ‘Tiffany set’ or ‘Tiffany setting’ may signal an implicit recognition that some uses of its protected mark are indeed descriptive.”

In conclusion, however, the court suggests that the ultimate question is whether the descriptive use worked to avoid confusion—which seems inconsistent with KP Permanent: “To be sure, a reasonable jury could also reject Costco’s evidence and find that customers would not recognize the word ‘Tiffany’ as descriptive even with the context Costco provided” (emphasis added).

The counterfeiting claim also had to be reversed and sent to the jury. In a footnote, the court indicated that if there was no use as a mark, there “likely” could be no counterfeiting even in the presence of infringement liability, since a non-mark use cannot be a “spurious mark.”  “We fail to see how a term can be a ‘fake’ mark if it is not actually used as a mark, or how a term can ‘deceptively suggest an erroneous origin’ if it is not used as a means to indicate origin in the first place,” even though terms not used as marks can still generate confusion as to “affiliation, connection, ... association[,] ... sponsorship or approval,” and thus constitute trademark infringement [court’s rather puzzling citation to §43(a)(1)(B) as a source of trademark infringement liability omitted]. “But because terms not used as marks are not ‘spurious,’ they cannot, as a matter of law, be counterfeit.”

The court declined to reach whether punitive damages under NY law would be available to Tiffany if it prevailed.

No comments: