In trademark litigation, whether involving trademark registration, opposition or invalidation proceedings, or infringement actions, courts frequently refer to the perceptions of the relevant consumers as a basis when adjudicating the scope of trademark rights or whether the other party has committed trademark infringement. For example, a senior rights holder seeking the cancellation of a later-registered trademark must demonstrate that the trademark gives rise to a likelihood of confusion among the relevant consumers. Similarly, a plaintiff asserting trademark infringement must demonstrate that the trademark used by the defendant is likely to cause confusion among the relevant consumers. By further example, a rights holder of a registered trademark asserting its status as a well-known mark must demonstrate that the mark is widely recognized by the relevant industries or consumers.

Consumer perception, however, is highly subjective and difficult to objectively and concretely establish. To empirically demonstrate relevant consumer perception, market surveys have become an important means by which parties in trademark litigation substantiate the subjective perceptions of relevant consumers and support their respective claims.

That said, preparing market surveys is a complex process involving numerous details. Factors such as the surveying institution, survey respondent screening, sampling methods, questionnaire design, and implementation methods may all impact the objectivity and credibility of the results. Whether a market survey possesses evidentiary value or serves as a meaningful reference in court often becomes a key focus of contention between the parties and of judicial reasoning in trademark cases.

Accordingly, in domestic practice, what factors do courts reply upon when determining the credibility of market surveys? This article synthesizes and provides readers with a structured reference of domestic judicial practice from three dimensions: methods of screening relevant consumers, survey questionnaire design, and required content of market survey reports.

I. Methods of Screening Relevant Consumers

As discussed above, when determining whether a trademark is distinctive, qualifies as well-known, or is likely to be confused with another person’s trademark, the core issue in each instance lies in the perceptions of the relevant consumers. This then raises the question: who qualifies as a “relevant consumer”? According to the view of Taiwan’s Supreme Administrative Court,[1] relevant consumers are consumers who have previously purchased or used the goods or services at issue, as well as potential consumers who may, or intend to, purchase or use such goods or services in the future. Thus, when conducting a market survey, the first critical issue is whether an appropriate method is applied to screen survey respondents and whether the resulting sample adequately represents the population of interest.

In a trademark invalidation case involving the supplementary education industry,[2] the plaintiff submitted a market survey report to demonstrate that the later-registered trademark was likely to cause confusion among the relevant consumers. The court found that the plaintiff’s exclusion, when conducting the market survey, of respondents who had not sought supplementary education-related products or services for themselves or for relatives and friends within one year was intended to ensure that respondents surveyed were aware of information relevant to the industry. The court thus found the screening method reasonable and consequently, the market survey of probative value. In addition, in a trademark registration case involving lip balm products,[3] the plaintiff submitted a market survey report to establish that the three-dimensional trademark for which registration was sought had acquired distinctiveness. The court reasoned that, because the relevant consumers of the lip balm products were female, the plaintiff’s decision to limit the survey respondents to female subjects aged between 18 and 30 was reasonable.

Conversely, in another trademark registration case involving cosmetic products,[4] the court found that excluding consumers who did not intend to purchase the plaintiff’s branded cosmetics in the future from the population impacted the probative value of the market survey report. Further, in a trademark infringement case involving the real estate sector,[5] the plaintiff submitted a market survey report to establish that the trademark is a well-known trademark. The court pointed out that the demand for, and purchasing capacity with respect to, residential property vary significantly across different age groups. Yet, the plaintiff failed to apply any age-based stratified screening and instead treated respondents aged between 20 and 30 as constituting an effective sample. This sample was insufficient to objectively reflect the well-known status of the trademark within the real estate sector. The plaintiff furthermore screened respondents on the basis of educational attainment, which was unrelated to the trademark dispute at issue and improperly influenced the results of the market survey.

In summary, when screening survey respondents, enterprises should ground their screening criteria in the inherent characteristics of the goods or services at issue and reasonably set such criteria by, for example, reference to gender, age, or prior interaction with the relevant industry. Conversely, if screening criteria are unrelated to the characteristics of the goods or services, or if respondents who may not be supportive of a particular brand are excluded in advance, courts may find the sampling method lacking representativeness, impacting the probative value of the market survey.

II. Design of Survey Questionnaires

To truthfully and objectively reflect consumers’ perceptions, the design of survey questions and the implementation of surveys must remain neutral and objective. When formulating survey questions and conducting the survey, one should, to the greatest extent possible, avoid incorporating their own subjective views or employing any means that may lead respondents toward specific answers. Otherwise, the market survey may, due to defects in its design or implementation, be deemed by the court as lacking probative value.

By way of example, in a case arising under the Fair Trade Act involving a food processor,[6] the plaintiff alleged that the appearance of the defendant’s food processor was substantially similar to that of their product and submitted a market survey report in support of this claim. However, the court pointed out that the wording of the survey questions employed phrases such as “this product with a similar appearance” and “such a similar product,” directly incorporating the subjective view of “similar appearance” into the questions themselves. This clearly risked steering respondents and undermined the objectivity of the survey. Hence, the market survey report could not be relied upon as evidence to determine whether the defendant’s product design was likely to cause consumer confusion or misidentification.

It is worth noting that, compared to closed-ended or multiple-choice questions, open-ended questions that do not provide any predetermined options are more readily accepted by courts. If open-ended questions are not adopted, due to research objectives or practical considerations, the party should specify reasonable justifications for the design of the questions and the response options.

For example, in a trademark infringement case involving a beverage product,[7] the defendant submitted a market survey report to demonstrate that the alleged trademark design did not cause confusion or misidentification among relevant consumers. The court found that, as the survey presented respondents solely with the alleged trademark design and asked them to mark how the words presented should be segmented by punctuation based on their own understanding without providing any predetermined response options, the survey was able to accurately reflect how relevant consumers perceived and understood the overall textual content of the alleged trademark design. The survey therefore possessed probative value.

On the other hand, in a trademark registration case involving apparel products,[8] the court pointed out that although the market survey report submitted by the applicant did not employ open-ended questions and instead used a multiple-choice format, the seven brand options listed were objectively selected from the EICP consumer profile database identifying the top seven casual sportswear apparel brands. As there was no indication of bias or intentional misleading, the survey questions were accepted as evidence in support of the applied-for trademark’s acquired distinctiveness.

In addition, if open-ended questions are not adopted, the response options should include an “other” option so as to avoid constraining the direction of respondents’ answers and being deemed by the court to have steered respondents. For example, in a case arising under the Fair Trade Act involving the appearance of mobile phone cases,[9] the plaintiff submitted a market survey report to demonstrate that the appearance of their phone case constituted well-known trade dress. The court held, however, that the answer options for the question “How was the product’s brand owner identified?” in the questionnaire not only failed to include an “other” option, but also that more than half of the answer options were structured in a manner favorable to the plaintiff, thus raising concerns of the questionnaire design being suggestive.

In summary, when enterprises administer trademark-related market survey reports, besides ensuring that the question design and survey methodology do not steer respondents, they should give priority to open-ended questions that do not provide predetermined response options. If non-open-ended questions are used, reasonable justifications should be provided and the response options should include “other” to ensure the credibility of the market survey report.

III. Required Content of Market Survey Reports

When a market survey report is presented before a court, what information should it include for the court to regard its contents as truthful and credible? From our review of relevant court judgments, we have identified items that the courts have recognized as necessary components of market survey reports, and break them down into three principal categories as follows:

A. Credibility of the survey institution: the length of time the survey institution has engaged in market survey business activities, the number of market survey cases it has conducted and its experience, and whether its survey reports have previously been accepted by courts.[10]

B. Information on the market survey: the survey period, methodology, qualifications of survey personnel, survey techniques, geographic scope, population, sampling method, definition of the population and total number of respondents, type of questionnaire, question types, question categorization, basic principles, structural organization, causal relationships, relationship between the content and the conclusions, and explanations of margins of error or confidence levels.[11]

C. Information enabling ex post verification by the court: including the original records of respondents’ basic demographic information and the original market survey questionnaires.[12]

Moreover, in some cases, expert opinions issued by scholars with expertise in statistics regarding market survey reports have also served as important references for courts in determining whether to admit such reports. In relevant judgments,[13] courts have admitted affirmative opinions provided by professors from the Department of Statistics at National Chengchi University and the Director General of the Institute for Information Industry’s Science & Technology Law Institute, and in turn accepted the results of the market survey reports.

Conclusion

Preparing a market survey report is time and effort consuming and may also involve substantial costs. Whether the survey results are ultimately accepted by the court is thus of significant concern to the parties of interest. As reflected in the foregoing synthesis of judicial practice, courts in Taiwan do not assess market survey reports solely on the basis of their final statistical results. Rather, courts rigorously examine the design of the questionnaire and how the survey is conducted to determine the report’s credibility and probative value. To avoid rendering invested efforts and expenses futile, enterprises should carefully plan market surveys in advance and ensure that the survey methodology, implementation process, and method of presentation all comply with the standards established by judicial practice. Only then may a market survey report effectively persuade the court and fulfill its intended evidentiary function.

Co-written by Gary Kuo, Partner; Pei-Hsu Wu, Attorney; and Jing-Min Huang, Researcher.

This is a translation of the original article in Chinese, which can be found here. Translation by Emily Lin.

If you would like to learn more about IP law matters, please feel free to contact Gary Kuo at gkuo@winklerpartners.com.

[1] Supreme Administrative Court Judgment No. 104-Pan-792.

[2] Supreme Administrative Court Ruling No. 109-Shang-78.

[3] Intellectual Property Court Judgment No. 104-Xing-Shang-Su-54.

[4] Supreme Administrative Court Judgment No. 104-Pan-792.

[5] Intellectual Property Court Judgment No. 107-Min-Shang-Shang-2.

[6] Intellectual Property Court Judgment No. 100-Min-Gong-Shang-1.

[7] Intellectual Property Court Judgment No. 106-Min-Shang-Shang-11.

[8] Intellectual Property Court Judgment No. 106-Xing-Shang-Su-57.

[9] Intellectual Property and Commercial Court Judgment No. 112-Min-Gong-Shang-2.

[10] See Intellectual Property Court Judgment No. 107-Min-Shang-Shang-2 and Intellectual Property Court Judgment No. 106-Min-Shang-Shang-11.

[11] See Intellectual Property Court Judgment No. 105-Xing-Shang-Su-22 and Intellectual Property Court Judgment No. 103-Xing-Shang-Su-136.

[12] See Intellectual Property and Commercial Court Judgment No. 111-Min-Shang-Su-25 and Intellectual Property Court Judgment No. 101-Min-Gong-Shang-6.

[13] Supreme Administrative Court Ruling No. 109-Shang-78 and Intellectual Property Court Judgment No. 104-Xing-Shang-Su-54.