Properly designed and implemented surveys can provide strong evidence of whether a disputed claim is confusing to consumers. For cases involving trademarks and trade dress infringement, surveys can additionally inform on the strength and recognition of a mark. The utility of surveys in claims pertaining to comparative advertising can be nuanced. For statements that are literally true, courts generally do not find survey evidence that consumers misunderstand the statements to be dispositive. However, survey evidence can inform a fact finder whether those statements are likely to mislead consumers. Finally, in Lanham Act cases involving false advertising more generally—when the issue of materiality is relevant—surveys can help answer the all-important so what question.
Consumer Surveys for Trademark or Trade Dress Infringement
Trademark or trade dress infringement cases (henceforth referred to as trademark infringement) often require establishing that consumers are likely to be confused between the at-issue marks. Although each federal court relies on its own set of rules or factors for establishing likelihood of confusion, there are common themes across the 13 federal courts. For example, the Ninth Circuit considers a list of eight Sleekcraft factors compared to the Federal Circuit’s 13 DuPont factors. Both sets of factors include an assessment of the strength of the mark (e.g., is the mark distinct and/or famous?); the similarity between the marks at issue; incidents of actual confusion; type of goods (e.g., high-involvement vs. low-involvement goods); and marketing channels used by the products at issue. While these factors are not an exhaustive checklist, establishing likelihood of confusion in trademark infringement cases is generally understood to be a multifactor test.
When addressing likelihood of confusion, surveys can be informative on a number of factors, including the strength of the mark, similarity of the marks at issue, and actual confusion. As an initial matter, a necessary—but not sufficient—requirement for infringement is that the allegedly infringed mark has to be well-known to the public. To this end, recognition or fame surveys present respondents with a copy of the relevant mark (as it appears in its real-life setting) and ask them to identify the source of or company associated with the mark. To provide a benchmark for comparison, these surveys can also ask respondents to identify additional well-known and lesser-known marks that are not at issue. In addition to the relative measure of recognition, at least one treatise on trademarks and unfair competition proposes that fame requires a minimum recognition rate of 75% among general U.S. consumers. In the 2014 IP dispute over trade dress infringement between Apple and Samsung, the parties disagreed on the minimum recognition rate for fame. Apple claimed that a 65% recognition rate indicated fame, citing a prior Nissan case, while Samsung argued for a higher 75% threshold. Although the court ultimately sided with Apple, it acknowledged that the famousness of Apple’s trade dresses was a “close question.”
At a high level, surveys used to demonstrate actual confusion among consumers vary across two dimensions: (1) the type of confusion (forward vs. reverse) and (2) survey design (Eveready vs. Squirt). Forward confusion is when a plaintiff claims that consumers mistake goods bearing the junior mark for products coming from the senior mark holder. For example, in Adidas America, Inc. v. Skechers USA, Inc., Adidas alleged that consumers were likely to confuse Skechers’s Onix shoe for the Adidas Stan Smith shoe.
Reverse confusion, on the other hand, implies the opposite, i.e., consumers mistake products bearing the senior (original) mark to be from the junior (newer) mark holder. For example, in Wreal, LLC v. Amazon.com, Inc., Wreal, a Miami-based pornography company that owns “FyreTV,” claimed that consumers mistakenly assumed that its mark was associated with the Amazon brand. In this case, Wreal is the senior mark, having used its mark FyreTV prior to Amazon’s Fire TV.
The distinction between forward and reverse confusion has at least one important implication when designing a customer confusion survey. For forward confusion, the target respondents are actual and potential customers of the product belonging to the junior mark holder. In the above example, Adidas can claim harm from Skechers’s alleged infringement only if a portion of Skechers’s customers purchased Onix shoes while mistaking them to be from Adidas. In the but-for world with no confusion, some of these customers would not have purchased from Skechers and opted for Adidas instead. In this case, a survey that does not include Skechers’s customers cannot evaluate likelihood of forward confusion. In similar fashion, the target respondents for evaluating reverse confusion are the senior mark user’s existing and potential customers. In the Wreal v. Amazon example, Wreal suffered damages from the alleged infringement only if a portion of Amazon’s customers purchased Fire TV while mistaking it for Wreal’s FyreTV.
Irrespective of the direction of confusion, two accepted formats for confusion surveys include the Eveready and Squirt formats. The major difference between these formats is whether respondents are shown the mark or product from one (Eveready) or both (Squirt) disputing parties. In a forward confusion Eveready survey, for example, respondents are shown the junior user’s mark or product to evaluate what portion of the respondents associate it with the senior user. One advantage of using an Eveready format is that it can address the similarity of the two marks and the strength of the mark in addition to likelihood of confusion. A high recognition rate for the mark or product when shown in isolation can be an indicia of the mark’s strength.
In the Squirt format, respondents view the mark or product from both parties. With the availability of both marks or products, the survey can be designed such that one party’s mark or product is shown to the respondents in isolation (depending on whether the claimed confusion is forward or reverse) and the second mark is shown together with other similar marks. Arguably, showing respondents multiple marks in a lineup is a closer replication of real-world conditions where the mark or product may appear in the presence of other similar marks and products, such as on a store shelf. By similar logic, a Squirt format is less appealing if the parties’ marks do not appear in close proximity in the marketplace.
Given that each format for a confusion survey has advantages, the choice between formats depends on the specific facts of the case (e.g., whether the parties’ marks appear in close proximity in the marketplace) and knowledge of which factors are most important to test. The availability of these choices in format and adaptability of the surveys to elicit answers on the factors that determine likelihood of confusion, all from respondents in the right target population, underscores the versatility of survey methodology in trademark infringement cases.
Consumer Surveys in Comparative Advertising
Comparative advertisements—ads that claim the superiority of the advertised products or services with a known competitor—can stimulate competition and be an important source of information for consumers when deciding between products. Comparative ads have been around for over a century. For example, in its recent smartphone advertisement, Samsung directly called out Apple’s competing product—“Buckle up for Apple’s latest launch as you enter a world where heads will turn. . . . Just not in your direction.” The ad proceeded to describe how the latest in camera specifications and form factor are not in the iPhone but are in Samsung devices. For consumers in the market for a new smartphone, these are arguably important pieces of information. To compete with Samsung’s smartphone, it then falls on Apple to demonstrate why consumers should still opt for an iPhone. Not surprisingly, in the U.S., the Federal Trade Commission recognizes and encourages the use of “truthful and nondeceptive” comparative advertisements.
While there are obvious benefits to comparative advertisements, the topic historically has been a controversial one given the potential for advertisers to lose focus and create misleading advertisements. When advertisers include false claims in their comparative ads, the cases that follow are similar to false advertising Lanham Act cases. These cases often use survey evidence. The more interesting phenomenon with comparative advertisements is that even when such ads include literally true statements, at times these ads can be misunderstood or have the potential to mislead consumers.
Select court rulings on prior matters underscore a potential limitation of survey use in comparative advertising cases. Specifically, when the at-issue claims are literally true, courts have ruled that survey evidence is not dispositive if it merely demonstrates that some portion of the respondents misunderstood the true claim. In Mead Johnson & Co. v. Abbott Laboratories, for example, the at-issue claim was the defendant’s use of the phrase “1st Choice of Doctors.” The claim itself was literally truthful in that more doctors recommended the defendant’s product than any other infant formula. The plaintiff, however, conducted a survey and demonstrated that some portion of the respondents misunderstood the claim to mean that a majority of physicians preferred the defendant’s product. The court in this case ruled that the claim made by the defendants was literally true and survey evidence could not be used to determine the meaning of words.
In other instances of comparative advertisements, where companies make claims that are literally true but may mislead consumers, survey evidence continues to be useful and accepted. In Merisant Co. v. McNeil Nutritionals LLC, for example, the ultimate question was that of the defendant’s intent, whether the true statement in its advertisement “Made from Sugar” falsely implied that McNeil’s sweetener, Splenda, contained real sugar and/or was “natural.” The claim itself was literally true in that Splenda is made using a patented, multistep process that begins with sugar but is then converted into a no-calorie, noncarbohydrate sweetener. However, the end product Splenda itself does not contain any sugar, nor is it “natural.” In contrast to the Mead Johnson matter, the court determined that Merisant’s survey evidence was admissible for at least two reasons. First, it questioned whether the “made from sugar” claim was meant to be deliberately misleading or whether it was simply misunderstood; and second, the scope of the survey aimed to determine broader consumer confusion as opposed to just the meaning of the word “from” in the challenged claim. Whether McNeil’s advertisement claim misled consumers into thinking Splenda contained sugar and/or was natural—a positive attribute—can be readily answered using survey methodology. That the plaintiff’s survey evidence was deemed admissible in this matter—in contrast to the Mead Johnson matter—underscores the importance of the advertisers’ intent in determining the admissibility of surveys in comparative advertisement Lanham Act cases.
The aforementioned court rulings have been informative with respect to when survey evidence can be dispositive in Lanham Act cases involving comparative advertisements. Advertisers have a fine balancing act on their hands when it comes to literally true statements in their comparative ads. While a degree of exaggeration and boasting in advertising, otherwise known as puffery, is generally not considered false advertising, the same is not true for statements (true or otherwise) intended to mislead consumers. When evaluating claims of false advertising in comparative advertisements, courts may attempt to determine whether the at-issue claims have the markings of puffery or if there may be an underlying intent to mislead consumers. For the latter, surveys offer a flexible, tried and tested method for assessing whether consumers are likely to be misled by the at-issue claims in an ad.
Materiality Surveys in Lanham Act Cases
The final category consists of claims that are literally false or misrepresentation. False statements or misrepresentations, once established, may be sufficient to result in the removal or modification of the at-issue ads. However, when plaintiffs seek damages resulting from the alleged false advertising, the underlying economic questions must be addressed: Did the plaintiffs suffer economic harm, and did the at-issue advertising cause the harm? In Wall & Associates, Inc. v. Better Business Bureau of Central Virginia Inc., for example, the plaintiff claimed damages from the defendant’s rating system, which it asserted was falsely advertised as “national, uniform, unbiased, and objective.” In dismissing the case, the court opined that while the plaintiff’s injuries may have been caused by the unfavorable ratings it received on the defendant’s system, the plaintiff had failed to draw a causal connection between its injury and the defendant’s characterization of its rating system as national, uniform, unbiased, and objective.
Lanham Act cases with a direct causal chain between the alleged false advertising and economic harm—unlike the particular example in the Wall & Associates case—are well suited for survey analysis. For example, suppose Company A claims to have lost customers because Company B falsely advertised its own product as having some nonexistent features. In addition to demonstrating Company B’s advertisement to be false or misleading, to prove economic damages, Company A has to provide evidence of a causal link between the alleged false claim in Company B’s advertisement and Company A’s injury. To this end, a survey using an A/B approach can be designed to evaluate the causal relationship in question. In its simplest form, survey respondents can be randomly assigned to a test or control group, where the test group views the advertisement with the features at issue and the control group views the same advertisement without the at-issue features. Both groups of respondents are then asked about their likelihood of purchasing the product in question, and any difference in the likelihood can be attributed to the at-issue features. A difference in the likelihood of purchase implies a causal link between the at-issue feature and consumers’ purchase decisions.
Materiality surveys can be used to evaluate a causal connection between the at-issue advertisement and consumers’ purchase decisions. However, additional analyses and/or surveys may be required to complete the causal chain to include the plaintiff’s economic harm. In the above example, establishing harm still requires Company A to show that the consumers going to Company B are those who would have gone to Company A in the but-for world.
The question of reliance, which precedes materiality, can at times be the link that completes the causal connection between the defendant’s at-issue advertising and the plaintiff’s economic harm. Reliance pertains to whether consumers relied on the claimed misrepresentation, or rather the advertisement with the alleged misrepresentation, during the purchasing decision. For example, even if a plaintiff could establish that an ad for a dietary supplement was falsely advertised as “natural,” the question remains as to whether consumers purchasing supplements relied on (or even saw) the particular ad. It may be more reasonable to assume reliance in instances where the claim appeared on the product packaging itself or in advertisements that appeared online or on national television for an extended period. However, if the alleged misrepresentation appeared in a brochure that was not widely circulated and was used for only a short while, it may not be as reasonable to presume reliance. Presuming reliance could be particularly problematic in instances where the alleged feature at issue was an important purchase factor for consumers. In such instances, before determining whether and to what extent the alleged false advertising affects consumers’ likelihood of purchase, surveys can be used to evaluate the extent to which respondents may have relied on the at-issue advertisements when making their purchase decisions.
Surveys can be designed to evaluate the exposure to and reliance on the specific ads at issue, or the question can be broached more generally. For reliance, one approach could survey potential and existing customers of the at-issue product, asking respondents to list what (if any) information sources they relied on when considering or purchasing the at-issue product. Following an open-ended question, respondents could then be asked to report how often (if at all) they rely on specific advertisement media like brochures, online ads, word of mouth, etc. Tabulating the responses on the various information sources can provide valuable information on the most and least common sources. For example, if the at-issue advertisement was in a brochure, survey evidence that only a small percentage of respondents relied on brochures generally for product information may be highly relevant for determining the impact or effectiveness of the advertisement. Such a result may suggest that the at-issue advertisement was not effective in reaching a significant portion of the target audience or that the advertisement was not the main reason for consumers’ decisions to purchase the product. Moreover, such a finding implies that the advertising in question is unlikely to have caused any substantial economic harm to the plaintiff.
Conclusion
Survey methodology offers a powerful tool for assessing economic harm in Lanham Act cases, including addressing issues of liability and quantifying damages. While this article has focused mostly on the former, others have written extensively on the use of conjoint surveys for estimating damages. Further, as with any powerful tool, it is important to exercise proper caution and understand any limitations. Survey experts have to take care to ensure that they reach the correct target population for their survey. For trademark confusion surveys, this means making the relevant adjustments for forward versus reverse confusion. In comparative advertisement cases, experts need to design their survey instruments to evaluate likelihood of the at-issue claims to mislead consumers and not merely be misunderstood by consumers. By contrast, when considering literally false advertising, survey experts need the proper context to determine whether economic harm can be established using a materiality survey alone or only in conjunction with a reliance survey.
