Proving Trademark Dilution: Special Challenges in Survey Evidence

By Ian B.

Over the last decade of Intellectual Property litigation, trademark dilution by blurring has become an increasingly popular plaintiff’s claim for a multiplicity of reasons: the non-necessity of showing confusion, the increased judicial recognition of pre-sale and post-sale consumer impressions, the enactment of the Federal Trademark Dilution Act and a more recent Supreme Court ruling on its construction all contribute to the attractiveness of making such a dilution claim.

However, the increasing popularity of simply ‘tossing in’ a blurring claim along with more structured assertions of confusion, has also created a vast amount of uncertainty as to what needs to be proven to sustain such claims of dilution and how the blurring claim should be proven.

What follows is a brief discussion of some of the relevant issues involved when attempting to present a successful empirical showing of trademark dilution by blurring.

THE BASIC REQUISITE SHOWINGS

At the outset it is worth noting that there are two necessary elements which must be effectively proven to sustain any claim of trademark dilution. The first necessary element is fame of the mark. In many cases a showing of fame is a mere formality because the concerned plaintiff is a widely recognized entity with vast commercial exposure.

In some cases, however, the concerned plaintiff’s mark may not be wholly distinct from other famous marks of a similar nature or its fame may be relegated to a certain region or demographic. It is generally established that nationwide fame is the requisite showing, but with the wide accessibility of internet commerce and exposure, plaintiffs frequently argue that marks which might otherwise be famous only in a certain region, are in fact recognizable on a national if not worldwide level. This, along with the fact that many marks may only be famous in a given set of consumer-demographics, has led to the use of survey evidence, even at this most basic threshold stage, to establish a dilution claim.

While creation and use of such surveys is not an uncertain practice itself, the presence of such surveys (particularly where the chosen route is to establish secondary meaning) can complicate the design and defense of surveys which may later be necessary to make a showing of ‘actual dilution’ as opposed to simple association. Frequently a strong empirical showing of secondary meaning necessary to establishing fame can end up weakening an assertion of an actual loss of distinctiveness in the concerned mark where such a showing will be necessary to meeting a court’s requirement to demonstrate dilution via loss of selling power.

The issue of fame aside, all claims of trademark dilution must also be based upon a second necessary element: a senior and a junior mark which are ‘similar.’ For dilution, only the marks themselves must be similar. This presents a stark contrast to establishing similarity when making a confusion claim (confusing similarity) which usually requires not only similar marks, but also similar products and/or business channels or other showings under the standard Sleekcraft Factors.

In short, claims of trademark dilution require only a showing of fame and a showing similarity of the concerned marks and the threshold for these showings, while sometimes a bit complicated, is not especially high. For this reason, dilution claims, and particularly claims based upon blurring, are frequently made upon a wide range of actual consumer associations and the existence of actual harm may remain speculative. Hence, the courts have applied a variety of standards in determining whether or not a prima facie assertion of dilution is in fact indicative of actual blurring. The only certainties which have been offered to date are a handful of elements which the Supreme Court has deemed as not necessary to actually establish a dilution claim.

UNNECESSARY ELEMENTS

Prior to the enactment of the 1996 Federal Trademark Dilution Act (FTDA), standards for trademark dilution were governed entirely by state laws. After the implementation of the FTDA, trademark dilution effectively became the province of federal law, and while the certain standards reflected in prior state laws such as the requisite elements of fame and similarity were retained, other aspects of the federal law were uncertain because rulings were inconsistent.

In response to these inconsistencies, the Supreme Court eventually heard Moseley v. Victoria’s Secret Catalogue, Inc., 537 U.S. 807. The Court’s ruling in Moseley clarified the following standards and limitations in regard to sustaining a claim for trademark dilution:
1. Dilution may only be based upon a claim of actual dilution, likelihood of dilution is insufficient.
2. Showing confusion, either a likelihood or actual, is not necessary in proving dilution.
3. Whether or not the senior and junior mark holders are competitors is not a limiting factor.
4. Showing actual loss of sales is not a requisite element to proving dilution.

The Court in Moseley did an excellent job in outlining what claims and elements of proof could not and did not need to be used to sustain a claim of dilution, but as to blurring it offered precious little guidance in terms of what a plaintiff should show in order to prove a loss of distinctiveness or how any requisite showings could be properly supported. This presents us with the current post-Moseley landscape, in which, as will now be discussed, there is vast uncertainty as to what standard judges will expect or apply when assessing a claim for trademark dilution by blurring. It has largely become an ‘anything goes’ atmosphere and many famous mark holders and plaintiff’s attorneys are struggling to find any sort of predictability to the manner in which they must prove their claims of trademark dilution. In brief, the present picture is as follows.

A FREE-FOR-ALL OF ASSOCIATION

In so far as marketplace association is concerned, the traditional model of dilution has historically been that at the point of sale, a consumer would see the defendant’s mark and by its similarity to the plaintiff’s famous mark, would make an association with the plaintiff’s mark thereby diluting its distinctiveness. This model is no longer the standard.

Judicial recognition of actionable commercial association either prior to or after the point of sale, coupled with increasingly effective arguments in consumer psychology studies based upon an aggregate marketplace experience, have led courts to accept a variety of associations and respective effects as being potentially indicative of dilution by blurring. Today it is just as probable that blurring may be proven by showing the traditional association ‘defendant’s mark recalls plaintiff’s mark’ as it is by the reverse.

This broadened scope of acceptable types of association to substantiate dilution is due in part to the aforementioned factors but is also a result of the Court’s observation in Moseley that dilution does not require an actual showing of loss of sales. Upon this assumption, attorneys have increasingly turned their attention to making convincing and creative showings of association alone, arguing that this in and of itself can evidence actual dilution.

The problem is that many courts still hold that dilution is not proven unless an actual loss of distinctiveness in terms of lost selling power is shown. This raises the question as to how one can show a loss of distinctiveness through a loss of selling power without having to actually show lost sales. It is upon this hurdle that many of the most convincing forms of association surveys get tripped up. The result is a lack of standardization in how to actually construct and execute a dilution survey and an increasing reticence among plaintiff’s attorneys to rely solely on such surveys when trying to formulate a convincing presentation of empirical evidence to prove dilution by blurring.

This relative open-endedness and uncertainty in how to effectively prove dilution has not only reduced the frequency with which surveys are relied upon by plaintiffs, but it has created a variety of approaches in how to conduct effective dilution surveys. Issues such as the potentially diminished value of a survey in the overall association model, persistent problems of cost, decreased attractiveness in the face of alternative empirical methods and predictable characterizations of flaws and bias by opposing experts, are all issues which must be accounted for in survey construction. In lieu of these issues, what follows is a short list of a few of things which should be done when designing and implementing a trademark dilution survey aimed at proving actual dilution by blurring.

WHAT TO DO

1. Use a diverse and size appropriate population: Diversity of the survey population is a crucial element to guarding against the inevitable assertion by opposing counsel that the design is flawed.
2. Implement sufficiently dissimilar controls: Control subjects which are too similar to the marks at issue will uniformly open the design to attack; the risk vs. reward is disproportionate any other way.
3. Use a double-blind method: To prevent allegations of clear bias, both the administrators and the survey participants must be ignorant of the purpose of the survey.
4. Reduction for noise: This is merely a standard methodological practice in proper survey analysis.
5. Focus only on the element to be proven: If the survey is to support blurring, design the survey to optimize the chance of showing reduced distinctiveness and selling power, don’t try to produce evidence of secondary meaning or confusion by the same design.
6. Lead without leading: A clever survey design must increase the chance that participants will indicate the desired response but the approach must be extremely subtle, opposing counsel will be constantly vigilant for bias.
7. Don’t overreach: The strength of survey results or the survey design must never be presumed as unassailable, a prudent approach assumes that flaws will be found and corroborative evidence will be necessary.
8. Account for marketplace realities: Every design must mimic the actual consumer experience as closely as possible; it is better to slightly sacrifice desirable results than to expose to survey to invalidation too easily due to an unrealistic design.
9. Be ready to fight it out: At the end of the day, every survey will be deemed “fatally flawed.”

BEYOND A BASIC BLURRING DESIGN: THE CONUNDRUM OF TRADE DRESS

Surveys to prove dilution of a trademark by blurring are feasible albeit lacking in a standard approach which is predictably known to be effective. Records of case law and various practice guides indicate that certain models are recognized to be optimal given certain situations and there is a reasonable degree of published discussion on the subject of dilution surveys generally. This said, as detailed above, prudent attorneys seldom approach the problem of proving trademark dilution by relying on a single survey as the sole form of empirical evidence to support their case. Increasingly the typical approach employs a survey along with some actual evidence of sales records of the plaintiff both before and after the introduction of a defendant’s mark into the marketplace and/or an extensive opinion by a well-versed consumer psychologist to support the assertion that association alone can and will lead to diminished selling power via a loss of distinctiveness.

In the case of a claim for dilution of trade dress, the picture is far less well developed. In fact, there are essentially no published opinions offering any record of a successfully designed and litigated consumer survey which has actually been deemed to prove dilution of trade dress by blurring. Simply put, trade dress claims are less frequently brought than trademark claims or they are brought in tandem. Claims for dilution of trade dress are rarer still, and there are only a handful of published cases in which a survey to support an assertion of dilution of trade dress by blurring has actually been considered or attempted.

This presents a very interesting caveat in an otherwise reasonably well documented area of legal study: How to design and use a consumer association survey to effectively yield empirical evidence of actual dilution in trade dress by blurring? The cases outlined below have at least touched on the issue.

THE ELUSIVE TRADEDRESS DILUTION SURVEY: CASES WHICH ARE CLOSE

Nabisco, Inc. v. PF Brands, Inc., 50 F. Supp. 2d 188 (S.D.N.Y., 1999): In Nabisco the dispute involved the popular ‘Goldfish crackers’ and a planned product to promote the cartoon ‘Cat-Dog’ featuring a variety of animal shaped crackers including one of a fish which was very similar in appearance to ‘Goldfish.’ The overall circumstances of the case presented an excellent opportunity to design and conduct a trade dress dilution survey which would likely have ended in a winning showing for the plaintiff, but the planned survey in the case was derailed when Nabisco failed to comply with a discovery order to produce the requisite units of ‘Cat-Dog’ crackers for Pepperidge Farm to conduct its proposed analysis. As such the survey was scrapped and its proposed design was never published.

Gibson Guitar v. Paul Reed Smith Guitars, 423 F. 3d 539 (C.A. 6, 2005): In Gibson the dispute was between two guitar makers for a similarly shaped ‘swoop-cleft’ design which constituted a substantial portion of the right side of the front face of a guitar. The basic facts of the case presented a seemingly good opportunity to conduct an effective survey for dilution of trade dress but the survey was never produced as Gibson lost on the issue of confusion and did not pursue the matter further.

Gateway, Inc. v. Companion Products, 2002 WL 34248562 (D. SD., 2003): Gateway involved the famous trade dress of the ‘Gateway Cow Motif’ and a computer accessories designer selling products comprised of several different cow-appearing items which allegedly called the Gateway-Motif to mind. A survey was conducted in the case and published and the judge found the evidence it yielded to be compelling but the survey was designed to show a likelihood of confusion. Even though Gateway prevailed on both the counts of confusion and trade dress dilution, the only published empirical evidence was the confusion assessment and it is well established that proving confusion is an entirely different standard and proposition than proving dilution. Hence, the design of the confusion survey cannot be relied upon as an effective standard for evidencing dilution across the board, even though the designed survey may have worked in this particular case.

I.P. Lund Trading v. Kohler Co., 11 F. Supp. 2d 112 (D. Mass., 1998): This case involved a dispute between two makers of faucets. In holding that there was sufficient evidence of actual dilution the court considered an application of the Mead Data Factors which assess the similarity of the marks, the similarity of the products, the sophistication of consumers, the question of predatory intent, and renown of each of the concerned marks. Through an analysis of these factors the court was comfortable in arriving at the conclusion that actual dilution was occurring. The main problem with this decision is that it offered no real explanation as to how an actual reduction in selling power by loss of distinctiveness was occurring. This case stands for the example that even after Moseley, courts take a wide swath of approaches in deciding what truly needs to be shown to constitute actual dilution and are sometimes prepared to take logical leaps of faith.

Adidas America v. Payless Shoesource, 2008 WL 4279812 (D. OR., 2008): The diversity of approaches to assessing actual dilution is also evident in the Adidas decision which presented a very strong case for dilution of trade dress but did not entail a survey. Instead, the plaintiff’s winning argument utilized an extensive report by a consumer psychology expert who presented a convincing argument that actual dilution could be determined upon the strength of consumer association alone. This case demonstrates how the uncertain standards in proving dilution of tradedress are encouraging attorneys to rely on alternative forms of evidence which may be more cost-effective and less time consuming to produce than surveys.

CONCLUSION

The aforementioned cases nearly constitute the entire universe in which trade dress dilution surveys have been considered or have been a viable option. Their ‘proper construction’ generally remains an open issue and if and when a successful trade dress dilution claim is brought and proven on the strength of a survey, it will likely be a very well studied case. Until such time though, this remains a largely open issue and the existence of a generally appropriate approach to presenting empirical evidence of dilution remains uncertain.

Because of this reality, the continually and rapidly changing consumer experience in an open marketplace and the variety of judicial approaches to what must be shown in substantiating a claim of actual dilution, it can only be said that a savvy approach to the issue must be a flexible one which is not reliant on any one mode of proof.