Defendant challenged plaintiff's surveys based on a failure to include a fictitious control design, the lack of a control cell, the vagueness of the questions, the use of leading questions, the creation of demand effects, the use of an underinclusive universe, the failure to replicate marketplace conditions, and an "order bias." The Northern District of California denied defendant's motion to strike plaintiff's surveys, finding that any flaws in the surveys went to the weight of the surveys, not their admissibility.



Plaintiff Levi Strauss & Co. ("Levi's") owns two trademark registrations for its embroidered pocket design called the "Arcuate Stitching Design Trademark" (the "Arcuate Mark") used on its jeans. Plaintiff sued defendant Abercrombie & Fitch Trading Co. ("Abercrombie"), claiming that Abercrombie's Ruehl 925 jean-pocket design ("Ruehl Design") infringed and diluted the Arcuate Mark.

Plaintiff conducted two surveys. One survey tested consumer recognition of Levi's back-pocket stitching. The other tested likelihood-of-confusion between the Ruehl Design and the Arcuate Mark. Abercrombie moved to exclude the two surveys on a number of grounds.


In one of plaintiff's surveys, Dr. Sanjay Sood measured consumer recognition of the Arcuate Mark. Participants were shown photographs of the back pockets of three brands of jeans: Levi's, Lee, and Lucky. The order of the photographs was rotated across the three brands. Participants were asked a series of questions about each photograph to test their recognition of each brand's pocket design. The first question in the series was: "Have you seen jeans with this style of pocket stitching before?" If a participant answered "yes" or "don't know," he or she was asked, "Do you associate jeans with this style of pocket stitching with one brand or company, more than one brand or company, [or] don't know?"

Abercrombie challenged the validity of the consumer recognition survey on the grounds that the questions asked were vague, the survey did not include a fictitious control design and, thus, did not account for "spurious recognition," and the survey lacked a control cell. The court concluded that the questions were not so inherently ambiguous that they affected the reliability of the survey. The court also found that the lack of a fictitious control design and the lack of a separate control group went to the weight of the survey, rather than its admissibility, apparently crediting Dr. Sood's contention that he did not include a fictional stitching design because of the many stitching designs on the market and because control jeans were used in the survey.

The second survey conducted by Dr. Sood tested for a likelihood-of-confusion. That double-blind, mall-intercept survey tested women who were between 20-45 years of age and who had either purchased jeans that cost at least $75 in the last six months or planned to do so in the next six months. Participants were shown a photograph of a female wearing a pair of Levi's jeans "from the back side such that the person wearing the jeans is posed in a stance as if she were striding or waiting at a street corner." The participants were then shown four photographs in the same perspective of four different brands of jeans: Citizens for Humanity, Seven for All Mankind, Ruehl, and True Religion. The interviewer then asked each participant a series of questions. The first question was: "Do you think that any of these jeans are made, sponsored or endorsed by the same company that made the jeans you saw in the first picture? Or, do you think that none of these jeans are made, sponsored or endorsed by the same company that made the jeans you saw in the first picture?" If the participant answered in the affirmative as to one or more of the jeans, she was asked: "Which one or ones do you think are made, sponsored or endorsed by the same company that made the jeans you saw in the first picture?" For each pair of jeans that a participant identified in response to the second question, she was asked: "What is it about these jeans that makes you say it is made, sponsored or endorsed by the same company that made the jeans that you saw in the first picture?"

Abercrombie argued that this survey was inadmissible because it used leading questions, created demand effects, lacked a proper control, utilized an underinclusive survey universe, lacked marketplace conditions, and suffered from "order bias." The court found that none of the questions were unduly suggestive, noting that in the cases cited by Abercrombie for this proposition, questions were found leading because they identified the trademarked product and the accused product in the question. It also found that Abercrombie's concerns regarding order bias, demand effects, and lack of marketplace conditions went to the weight rather than the admissibility of the survey. Further, it held that because controls were used in the survey and the questions asked were open-ended, any failure to use a separate control group did not render the results of the survey so unreliable that it should be excluded.

Finally, the court considered Abercrombie's argument that the survey universe was underinclusive because it focused on the target market for Ruehl jeans. Abercrombie argued that plaintiff's survey universe should have included girls and women 13 years and older, and should not have been limited to any price point. However, Abercrombie's expert did not explain how or why the persons excluded from Dr. Sood's survey universe would have reacted differently from those included in Dr. Sood's survey. Accordingly, the court concluded that any defects in the survey universe also went to the weight of the survey.

Abercrombie, as part of its motion for summary judgment, also attempted to introduce surveys conducted by plaintiff for other litigations, which it claimed showed that the Arcuate Mark had lost distinctiveness, but offered only an attorney's (rather than an expert's) calculation analysis of the significance of the survey results. The court sustained plaintiff's objection to the surveys' use and did not consider them on summary judgment, finding that "[u]nsubstantiated attorney argument regarding the meaning of technical evidence is no substitute for competent, substantiated expert testimony."


This opinion provides direction regarding the Northern District of California's evaluation of the admissibility of consumer recognition and likelihood-of-confusion surveys, and underscores the importance of expert testimony in analyzing survey results.