Summary:
MMR Strategy Group was retained to rebut survey evidence submitted in an action arising from the alleged infringement of guitar shapes in the matter Gibson Brands Inc. v. John Hornby Skewes & Co, Ltd. Dr. Justin Anderson, Senior Vice President of MMR Strategy Group, testified that the plaintiff’s consumer survey expert failed to properly and reliably measure genericness. Did plaintiffs strum up enough evidence to succeed in this trademark infringement action?
Case Facts
Gibson Brands Inc. (“Gibson”) filed a complaint in the Central District of California against John Hornby Skewes & Co, Ltd. (“JHS”), alleging trademark infringement, trademark dilution, trademark counterfeiting, false advertising, and trade dress infringement. Gibson is a leader in the manufacture, sales, and distribution of guitars and has held trademarks for guitar shapes and designs since the 1960s. It holds various trademarks for electric guitars and acoustic guitars. JHS Music released a line of guitars with “vintage” shapes for sale online and in retail locations in the US.
In this action, Gibson alleged that JHS infringed six of its trademarks, including Gibson’s SG Body Shape Design, better known as the Les Paul model; Explorer Body Shape Design; ES Body Shape Design; Flying V Body Shape; Flying V Peghead Design; and Kramer Peghead Design. By doing so, Gibson alleged, JHS also diluted the Gibson brand through its sales in retail and online locations across the United States and online. It further alleged that JHS attended various industry conventions promoting guitars with shapes that allegedly infringe on the trademarks at issue.
In its counterclaim, JHS requested the cancellation of the six Gibson trademarks. According to the counterclaim, numerous other manufacturers began marketing, advertising, and selling guitars and basses with similar shapes in the 1960s and continue to do so in the present, including guitars with similar or identical shapes to those Gibson claimed. The counterclaim cited examples of other manufacturers who produced similar or identically shaped guitars that were commercial successes and made famous not by Gibson marketing, but through their use by famous musicians.
Consumer Survey Design Measuring Genericness and Confusion
Gibson retained a consumer survey expert to design a survey to determine the extent to which certain Gibson guitar body shapes are thought to be generic. The plaintiff’s expert designed a survey based on the Teflon survey format for measuring genericness.
The plaintiff’s survey expert designed a “mini-test” to qualify respondents, in which they were provided with definitions of branded shapes and common shapes, then instructed to categorize guitar shapes as either branded shapes or common shapes. Four test shapes were assigned at random to the qualifying respondents, along with a control shape.
The resulting measures were submitted as evidence that the claimed guitar shapes were not generic, the plaintiff’s survey expert testified. The survey found that:
- An average of 57.8% of respondents identified one of Gibson’s guitar shapes as branded rather than common.
- 26.3% correctly identified the guitar shape as Gibson’s.
MMR Strategy Group’s Rockin’ Rebuttal
Dr. Justin Anderson testified that the plaintiff’s survey evidence was flawed for several reasons: it failed to properly qualify respondents, used a flawed research design, included inappropriate stimuli, did not follow proper data collection and analysis methods, and, as a result of these flaws, drew unsupported conclusions.
One important flaw Dr. Anderson identified was that the opposing expert added questions to the screening questionnaire after the pretest, but that 49 pretest interviews were not removed from the survey’s database. In fact, the survey’s database showed responses to the additional questions from those 49 pretest respondents, even though they had no opportunity to answer the questions. The data for those pretest respondents also showed that the images shown to them were not randomized; all of them saw the same test image. The expert who designed and conducted the survey acknowledged at deposition that the 49 pretest respondents should have been removed, and Dr. Anderson argued that not removing them was an important flaw in the survey.
For a survey to reliably measure genericness, the survey’s universe must accurately represent consumers who have purchased or will purchase a product in the relevant category. In Dr. Anderson’s rebuttal, he testified that Gibson’s expert’s survey included only a subset of guitar purchasers, resulting in an underinclusive group of respondents who were more knowledgeable than the average buyer. This could produce bias in the survey results, Dr. Anderson testified.
Dr. Anderson also pointed to a flaw in the design of the Gibson survey. He said the plaintiff’s survey provided respondents with an improper definition of “common shapes,” because it said common shapes are used to identify a product or service. This is incorrect; common shapes are not source identifiers, and could only indicate a “type of” product or service. The “mini-test” used to measure the difference between brand shapes and common shapes was therefore ineffective and may have admitted respondents who provided unreliable data. Also, in typical Teflon survey designs, respondents are exposed to more than one control stimulus. The plaintiff’s survey included only one control stimulus, which was a common shape; no branded shape was included as a control, as would typically be true in a Teflon survey. As a result, the survey did not allow comparisons between the results for the trademarked shapes and the results for a branded control.
In addition, Dr. Anderson testified that the stimuli presented to respondents did not isolate the body shapes of the guitars. The stimuli included other distinctive features that could be source identifiers, such as control knobs, switches, and fret inlays. In fact, Dr. Anderson opined, responses to Gibson’s survey show that in some cases, respondents did use these other types of source identifiers to determine source. Dr. Anderson further testified that the control shape was inappropriate because it was an acoustic guitar, whereas the body shapes in dispute were electric guitars. As a result, the intended audience of electric guitar purchasers may have classified the control guitars differently from the test guitars because they were a different type of guitar, rather than by body shape.
Finally, Dr. Anderson argued that the survey did not follow its own logic, thereby including or excluding the wrong respondents; the opposing expert failed to report which respondents were disqualified at certain junctures of the survey; and that there was no explanation for the coding of open-ended questions. He also noted that the plaintiff’s expert was not able to answer questions about particular areas of survey design and data collection.
How Did the Court Shape Its Decision?
The court excluded Gibson’s genericness survey, citing several reasons drawn from Dr. Anderson’s testimony. In particular, the court noted that the pretest group in the Gibson survey was included in the final survey database after the pretest criteria were changed. This is a serious flaw, the court said. Survey experts must ensure that their work is free of major oversights that make their data unreliable.
If you are involved in a trademark infringement matter that requires a reliable likelihood of confusion survey, contact MMR Strategy Group.