The Federal Court’s most recent decision on trade-mark surveys sets a high standard for validity of survey evidence. In Promotion In Motion Inc. v Hershey Chocolate & Confectionery LLC, 2024 FC 556 (“SWISSKISS”), the Federal Court provided a detailed review of the validity and reliability criteria of survey evidence. Although the survey evidence in that case was designed by two leading survey experts in Canada, Dr. Corbin and Mr. Legere, the Court found that there were fundamental issues with the surveys, rendering the evidence inadmissible or of no probative value.

In particular, armed with critical expert testimony from a third survey expert, the Court found that:

1. “The right questions” were not put to the survey participants – the wording of the questions created a priming bias;
2. The questions were not put to the participants “in the right way” – the Internet-based online survey format did not involve the expert or an interviewer who could attest to the participant from the online panel actually being the one who filled out the survey, with no video recording of the participant filling out the survey serving as an alternative method of confirmation; and,
3. The questions were not put “in the right circumstances” to participants – the online survey’s design allowed participants to simply click the “back button”, with the expert unable to provide evidence as to the effect that click would have or its impact on the survey’s results.

Overall, the Court took significant issue with the surveys being conducted online without the presence of the expert or an interviewer who could attest to compliance with the survey requirements by participants. On this basis, the Court found that the criterion of reliability was not satisfied. Due to the serious flaws identified with the survey evidence, the Court found that it was inadmissible.

The Court’s guidance sets a high bar for online surveys, seemingly requiring litigants to ensure meticulous monitoring and a high level of control in online surveys, or alternatively opt for costlier and still-vulnerable in-person surveys instead.

Yet, for those thinking that perhaps such evidence is not worth the expense, the Federal Court in Energizer Brands, LLC v Gillette Company, 2023 FC 804 ("Energizer”), commented critically on the lack of survey evidence in that case. Noting that due to the lack of such evidence, the full real world context was missing from the expert evidence.

The Court hinted in Energizer that survey evidence may be a requirement to succeed on a claim under section 22 of the Trade-Marks Act. Otherwise, the Court is left to rely on its own “common sense” to determine the likelihood of confusion, rather than on the empirical evidence that may be provided by surveys. As noted by the Supreme Court in Masterpiece Inc. v. Alavida Lifestyles Inc., 2011 SCC 27, “[s]urveys … have the potential to provide empirical evidence which demonstrates consumer reactions in the marketplace — exactly the question that the trial judge is addressing in a confusion case. This evidence is not something which would be generally known to a trial judge” (para 93).

As such, even with all the pitfalls that survey evidence presents, litigants may have no choice but to incur the expense. In determining whether survey evidence is necessary for their case, litigants should carefully consider the information that is available to the experts without such evidence. If the evidence is not sufficient to provide the Court with an understanding of the “real world context” of the use of the marks, and thus the likelihood of confusion – survey evidence may be necessary.

Should that be the case, the design and administration of survey evidence should be carefully designed in light of the Court’s comments in SWISSKISS. Online evidence may still be used, however it may require, for example, the participation of an interviewer for each survey. In addition, having a different expert provide commentary and criticism of any proposed survey design would help identify issues before they make it into an expert report.