Trademark issues such as likelihood of confusion or the fame of well-known marks are difficult to prove. In practice, these issues can be very fact-specific and the answers to these issues often depend on the subjective perception of the parties in disputes. Therefore, when a party attempts to prove these issues, the evidence often presented through the form of "market surveys". However, market surveys are not widely accepted under the practice of the Courts in Taiwan. In addition, the probative value of market surveys in trademark disputes is also quite controversial.

In the past, the Courts' attitude was more conservative against market surveys. If a market survey report was not delivered by the designation of the Courts, such report normally would not be accepted as evidence. In recent years, the Courts' determination on the probative value of market surveys started to get clearer. For example, in a recent administrative litigation regarding a trademark opposition, the Court expressed a worth noting view on the probative value of market survey reports.

The plaintiff in this case was engaged in the business of mobile phone accessories. The plaintiff held a trademark registration in Class 9 with the designated goods such as "PDA cases, PDA sleeves, mobile phone sleeves, mobile phone cases, lithium battery, fuel cell, alkaline battery, nickel cadmium battery…" Judging that the registration is confusingly similar to its trademark, a British fashion brand filed an opposition against the plaintiff's registration with Taiwan Intellectual Property Office (TIPO). In the opposition, TIPO held that the British fashion brand's trademark is a well-known mark therefore the plaintiff's registration should be partially cancelled. The Plaintiff appealed the case all the way to the Intellectual Property Court.

In the administrative litigation, the Plaintiff offered a market survey conducted by a research organization in Taiwan in attempt to prove that the British fashion brand's trademark should not be considered as a well-known mark. In determination of the probative value of the market survey, the Court first clarified that, even though the market survey was not delivered by the designation of the Courts, the market survey should not be rejected per say as evidence. Nevertheless, after reviewing the original statistics of the market survey, the Court still decided that the market survey was not probative and should not be accepted.

First of all, the Court declared that the selection of respondents of the market survey should be done in a random manner in order to reduce bias in representation of the larger population. However, in this market survey, there was a large disparity between the numbers of the respondent interviewed by each researcher. In addition, after checking the interview locations with the GOOGLE map, some of the locations were remote and rarely visited by the crowd.

The plaintiff argued that, the most important factor in a market survey is whether the respondents may honestly reflect their actual feeling in the interview. The Court, however, pointed out the issue of whether the respondents may honestly reflect their actual feeling in the interview is not conflicted with the issue of the selection of respondents and the interview location. The Court further indicated that the plaintiff's argument could not resolve the problem of bias in this market survey. Therefore, the Court held that the market survey was not probative and should not be accepted.