Hershey’s Kisses Go to Court – Federal Court of Appeal Decision in PIM Brands Inc. v. Hershey Chocolate & Confectionery LLC

The Federal Court of Appeal’s recent decision in Promotion in Motion, Inc. (PIM Brands, Inc.) v. Hershey Chocolate & Confectionery LLC offers fresh guidance on how courts evaluate survey evidence in trademark disputes, particularly online surveys. It also sheds light on why descriptive certification marks (think geographic indicators like “Swiss” for Swiss-made goods) have limited value in rendering a trademark distinctive.

Background

In 2013, the appellant, Promotion in Motion, Inc. (PIM) applied to register the trademarks SWISSKISS and SWISSKISS & Design for “chocolate of Swiss origin”. The respondent, Hershey Chocolate & Confectionery LLC (Hershey), opposed the registration, arguing that SWISSKISS was confusingly similar to their iconic trademarks for “chocolate candy”, KISS and KISSES.

The Registrar of Trademarks refused PIM’s application on the basis that there was a reasonable likelihood of confusion with Hershey’s marks. PIM then appealed to the Federal Court, where both parties filed extensive new evidence, including consumer surveys conducted by experts to assess likelihood of confusion. However, the Federal Court upheld the Registrar’s decision and rejected PIM’s application. PIM appealed the Federal Court’s decision to the Federal Court of Appeal (FCA).

FCA Decision

On appeal, PIM argued the Federal Court erred by:

1. focusing on the KISS elements of its marks and holding that the SWISS element was a descriptive certification mark that could not be distinctive; and
2. concluding that the survey evidence filed by PIM was inadmissible.

Certification Marks

PIM argued that, because it had permission to use the registered SWISS certification mark owned by Chocosuisse (the Association of Swiss Chocolate Manufacturers cooperative), the word “Swiss” in SWISSKISS should help distinguish it from Hershey’s KISS trademark.

The FCA rejected this argument, explaining that certification marks like SWISS exist for a different purpose than traditional trademarks: where a traditional trademark indicates the commercial source of a good or service, the certification mark shows that the good or service meets some defined standard. Certification marks are thus inherently descriptive. The SWISS aspect of SWISSKISS was therefore not capable of distinguishing it from KISS.

Survey Evidence

PIM engaged two survey experts to assess whether consumers who purchased Swiss chocolate in retail stores would mistakenly attribute SWISSKISS products to Hershey. Both surveys were conducted online. The results showed that only 2-3% of participants mistakenly attributed the marks to Hershey.

The lower court had found that the surveys suffered from validity and reliability issues that made them inadmissible as expert evidence. In particular, it found that the surveys failed to ask the right questions, in the right circumstances, and in the right way:

1. The “right questions” were not asked: The surveys repeatedly mentioned “Swiss chocolate” in each question, which likely influenced participants to focus only on Swiss chocolate brands and ignore other chocolate manufacturers like Hershey.
2. The questions were not put in the “right circumstances” to participants: Surveys in trademark disputes are designed to simulate the imperfect recollection of a casual consumer in a hurry. For this reason, participants were to see the SWISSKISS mark for only up to 8 seconds. However, the experts were unsure whether participants could simply use the “back button” in Internet browsers to view the image again.
3. The questions were not put to participants in the “right way”: Online surveys are inherently flawed. Among other things, the experts could not confirm that the participants from the online panel were the ones who actually completed the surveys, or that the participants were free of external influence or access to external information.

The FCA agreed with the lower court on the first two issues. However, it disagreed with the suggestion that online surveys are fundamentally unreliable and should only be accepted if someone monitors participants in person or records them on video. The FCA found that, while imperfect, Internet surveys may sometimes provide useful evidence to inform the confusion analysis when properly designed and implemented. The admissibility of online survey evidence, as with any other type of evidence, must be determined on a case-by-case basis.

Nevertheless, the FCA found that this error in the Federal Court judge’s decision did not amount to an overriding error that would warrant overturning the decision. Consequently, the lower court’s findings on both issues were upheld.

Key Takeaways

The FCA’s decision in this case provides valuable insight to trademark litigants on the admissibility of survey evidence in trademark disputes. Online surveys will not be rejected out of hand, but they must be meticulously designed to avoid pitfalls like priming bias and methodological flaws. Courts will scrutinize whether survey questions were phrased neutrally and whether the survey environment adequately controlled for external influences. Parties relying on survey evidence should ensure their experts can address these concerns.

Additionally, incorporating a geographic certification mark (such as “Swiss,” “Canadian,” or “Belgian”) into a traditional trademark does not render that element distinctive. These marks remain descriptive of geographic origin, and the confusion analysis will focus on the other, more distinctive, elements of the trademark. Businesses should consider this limitation when selecting and designing their branding.

For more information, please contact any other member of our Intellectual Property Litigation Group.