Excluded "Cybersecurity" Trademark Survey Provides Population Pointers

Welcome to?Lanham Act Surveys for Lawyers, your resource devoted exclusively to making survey evidence discussions fun and informative.?We provide insights and timely updates to help guide trial counsel in developing consumer surveys for use at the TTAB, in federal courts, and other ports of call where trademark and advertising disputes are routinely litigated.

Excluded "Cybersecurity" Trademark Survey Provides Population Pointers,?Op. 1, No. 21

Identifying the appropriate population is a critical task on the road to designing a valid survey in a trademark infringement action. Sometimes this can be a (relatively) straightforward proposition, like when the researcher seeks the opinions of those that use and consume common, everyday products and services. But what about when the population is a “niche” audience, say, professionals that offer specialized expertise in a narrow field? Read on to learn more about the thorny issues to keep in mind when dealing with such circumstances.

Today's discussion features a trademark dispute where "Darkowl" allegedly infringed "Arkowl." Darkowl, LLC v. Arkowl LLC, 2023 U.S. Dist. Lexis 133367 (D. Colo. Aug. 1, 2023)(Daubert Order ECF. No. 73). The parties' respective services in this dustup are described as follows:

Both parties provide their services to those operating in the field of "cybersecurity." Arkowl's expert conducted a likelihood of confusion survey (using a Squirt design). Because Arkowl alleged a claim of "forward confusion" the appropriate population was Darkowl's prospective customers, i.e., those companies that are likely to purchase Darkowl's "dark web" database services.

But Arkowl did not settle on this survey population. Instead, Arkowl's expert defined the survey population more broadly as likely purchasers of "cybersecurity software services." Then, the expert used a "funnel" of screening questions to allegedly arrive at "the narrow Darkowl universe." Daubert Order, p. 17.

The Court found this population overbroad and the screening questions did not help to sufficiently target the survey population. Why? Because notwithstanding the screening questions, a survey respondent still qualified to take the survey if he or she:

"was part of the decision-making process for a number of services. . .[t]hose individuals and the organizations they worked for thus could have been involved only with data for customer verification or cybersecurity software and services, with no involvement in dark web data services or hiring vendors for such services." Id. at 18 (emphasis added).

Further, the Court found that there was no way to discern: (1) whether potential customers of DarkOwl’s dark web services actually participated in the survey; (2) the type or size of organization that the respondents worked for relevant to DarkOwl’s prospective customer base; or (3) if the respondents or their employers actually used dark web data. Id.

Another problem noted by the Court was the expert's decision to not use a "specialty" internet panel. Id. at 28-30. Allow me to explain. Online surveys (like the one conducted by Arkowl's expert) are conducted using "internet panels." These panels are comprised of individuals who have either signed up or have otherwise been recruited to participate in online surveys.

Panels come in different shapes and sizes. There can be large "general" panels with millions of "rank-and-file" consumers. These panels are typically used when the goods at issue in the trademark dispute are commonplace (think pizza, coffee, batteries, and scads of other everyday consumables). There can also be "specialty" panels. These are typically used when the goods or services at issue are only relevant to a "niche" or specialized industry (think heart surgeons, nuclear fission scientists, and the like). The Court found that given the sophisticated and specialized nature of Darkowl's services, a "general" panel was improper. Id. at 30.

Thus, the Court excluded the survey.

The Takeaway. Much care and attention should be given to identifying the correct population in any trademark litigation survey. If the dispute at issue involves everyday goods or services, identification will likely be more straightforward. But if the goods and services at issue are more specialized "niche" products or services, proceed with caution. You could end up being a bit owly if you don't. The Court's Daubert order is set forth below, as is an additional reference about internet panels from the American Association of Public Opinion Research (AAPOR).