A Sketchy Federal Circuit Holding on a Converse Trade Dress ITC Proceeding

In 2013, the USPTO issued U.S Trademark Registration 4,398,753 to Nike subsidiary Converse, Inc. The registration covered the Converse Chuck Taylor All Star midsole stripes, toe cap, and multilayered toe bumper featuring diamonds and line patterns, which Converse claims to have been selling since 1946.  In October 2014, Converse filed a complaint with the ITC alleging § 337 violations by various importers. Several respondents, including Skechers and New Balance, defended against infringement and alleged the asserted registration was invalid.

In June 2016, the ITC issued its final determination that Converse’s registration was invalid, and there existed no common law rights in the design because the design lacked secondary meaning. The ITC relied predominantly on evidence of extensive third party use of the same or related designs, and on a consumer survey from the spring of 2015 that purported to show 21.5% of consumers associated the design with a single source, a rate the ALJ found insufficient to establish secondary meaning.

Converse appealed, and a panel of the Federal Circuit vacated and remanded the ITC determination, finding the agency erred both in its determination of invalidity and in its infringement analysis. Judge O’Malley concurred with the ultimate holding, but dissented from a substantial portion of the majority opinion.

The first eyebrow-raising portion of the opinion occurs after the panel articulated six factors the Federal Circuit will consider when determining whether a mark has acquired secondary meaning: (1) association of the trade dress with a particular source by actual purchasers (typically measured by customer surveys); (2) length, degree, and exclusivity of use; (3) amount and manner of advertising; (4) amount of sales and number of customers; (5) intentional copying; and (6) unsolicited media coverage of the product embodying the mark. The court then focused on the second factor, particularly evidence of third party use. Apparently, Converse introduced evidence corresponding to the decades of use of its asserted trade dress. The respondents introduced evidence showing decades of third party use of related trade dress elements.

This is where the court’s analysis becomes interesting. The court held that third-party-use evidence for purposes of evaluating secondary meaning is most relevant for “the recent period before first use or infringement.” Analogizing from § 2(f), the court held that the five year look-back period is the “most relevant” for purposes of the second factor. The court explained, “Consumers are more likely to remember and be impacted in their perceptions by third-party uses within five years and less likely with respect to older uses.”

An issue in the case was whether there was sufficient evidence of secondary meaning to sustain Converse’s federal registration, which was still contestable.  Third party use from five years before registration should be highly relevant for purposes of evaluating whether the dress had sufficient secondary meaning to warrant registration.  But that was not, however, what the majority of the court held. The court’s analysis was for purposes of evaluating infringement. The discussion of relevant third party use was framed in the context of measuring secondary meaning for the five years predating a defendant’s first use of the accused mark.  That is a more suspect holding.

Foremost, if the majority truly meant the relevant measuring period was “the recent period before first use or infringement,” that would eviscerate the relevance of secondary meaning surveys in most cases. How many defendants conduct survey’s of a plaintiff’s mark before the defendant is accused of infringement?  I would estimate the rate is approximately never.

Second, the opinion gives no reason for limiting the relevant time period for reviewing secondary meaning evidence to only third party use evidence. Every portion of the court’s reasoning regarding recency of evidence applies equally to factors 1, 2, 3, 4, 5, and 6. If a mark owner introduces decades of sales and advertising information, then fairness dictates that decades worth of corresponding third-party use should also be relevant. For example, a mark owner’s advertising will be mitigated by contemporaneous third-party use of the same or closely related marks. Further, the recency of all types of secondary-meaning evidence should be more relevant to the current impact on consumer impressions. Certainly, while third-party use from twenty years ago is less relevant than from two years ago, so too are advertisements and unsolicited media coverage from twenty years ago.

Further into the opinion, the majority took critical issue with the intervenors’ consumer survey measuring single-source association. The court instructed the ITC that the survey should be given little weight regarding source association where the defendant’s first use of the accused mark predated the survey by more than five years.  In other words, if the defendant started using its accused mark in 2005, a consumer survey showing the plaintiff’s mark lacked secondary meaning in 2015 would not be relevant.  That cannot be correct.  First, the legal and logical presumption is that marks gain strength, not lose strength, over time through cumulative consumer impressions. If a mark lacks secondary meaning now, the presumption should be it never had secondary meaning. The presumption could be rebutted where the source origination was lost due to the conduct of that defendant, but that does not appear to be the circumstances of this case. Second, even assuming circumstances may have changed, the issue in most trademark cases is whether the defendant should be permanently enjoined based on current and ongoing confusion. Whether a plaintiff’s mark still functions as a mark should always be relevant in trademark litigation.  Finally, the majority’s holding incentivizes owners with questionably distinctive marks to delay five years before bringing suit in order to innoculate themselves against adverse source-association surveys.

The majority opinion also addresses the relevance of the survey for purposes of evaluating the validity of Converse’s registration. The ITC ALJ found the survey’s 21.5% showing of single-source assocation to be insufficient to establish secondary meaning.  The majority opinion states, “We see no error in the conclusion that the survey does not establish secondary meaning, but we are unclear as to the ITC’s reasoning as to why the survey supports the opposite – a lack of secondary meaning.” The majority advises that on remand, “the Board should analyze whether the survey shows lack of secondary meaning as of the date of registration. Unless the survey affirmatively shows a lack of secondary meaning, there is simply a lack of survey evidence of secondary meaning – which is a neutral factor favoring neither party.” The majority’s language and logic are obscure.  It is unclear whether the majority is advising that the survey is not relevant because it did not measure secondary meaning at the actual time of registration (as opposed to two years later) or whether a secondary meaning survey that shows no meaningful consumer association is a neutral lack of evidence, or both. Either way, in this author’s opinion, the majority is wrong. How is a defendant supposed to affirmatively prove a negative? Perhaps, for this majority, it is a matter of semantics. Need the respondents and ALJ only repackage the conclusion “affirmatively,” i.e. “78.5% of respondents do not associate the asserted dress with a single source?” If the majority intended a temporal criticism, why would a survey conducted two years after registration not reflect consumer association at the time of registration?  If anything, shouldn’t association be stronger over time? Relatedly, if we are going to use a temporal starting point for relevance, shouldn’t we use the date of application, particularly since that is the date from which federal registration rights are based?

One of Judge O’Malley’s several concerns in dissent is that the panel overextended itself beyond what was necessary to resolve the appeal, and that the panel indulged in advisory rulings. Judge O’Malley’s criticism of the majority opinion is well placed. The opinion’s suspect legal guidance strays beyond the facts before the panel. The opinion’s broad expanse and generalities in certain areas may cause confusion and give rise to future mischief in this area of trademark law.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s