Wednesday, April 26, 2017

Internet surveys are admissible (but may raise IRB concerns)

Bimbo Bakeries USA, Inc. v. Sycamore, No. 13-cv-00749, 2017 WL 1377991 (D. Utah Mar. 2, 2017)

Bimbo charged that defendants misappropriated its trade secret for making Grandma Sycamore’s Home-Maid bread, and infringed on its trade dress related to the packaging of its bread. Here, the court resolved challenges to experts, as relevant here in favor of admission.

Defendant U.S. Bakery sought to exclude the expert testimony of Dr. Glenn L. Christensen; the court found him qualified to testify to quantitative surveys, of which he had prepared three. He conducted his surveys over the Internet using pre-screened panels of respondents provided by a third-party vendor, using digital images of the parties’ respective products. The court held that the internet was a proper method for conducting surveys, despite defendant’s argument that they didn’t effectively recreate the consumer experience of buying bread and screened out responses from those who buy bread, but who do not use the internet.  Defendant didn’t cite authority holding internet surveys unreliable. It was true that most consumers don’t buy bread online, so online surveys might not the best way to simulate the bread buying experience. But defendant merely speculated that the results might be different if the surveys were conducted in person or among people who buy bread but don’t use the internet.  Where the overall look of the product was at issue, online surveys could be relevant; other arguments could be addressed to the jury.  “To prove that a survey technique is unreliable the party must do more than speculate that there may have been a better way of completing the survey.”

The surveys also chose a representative enough sample for the jury to weigh them.  The survey looked for respondents in Utah and southern Idaho, the area Bimbo’s trade dress allegedly had secondary meaning, so that was okay. Using online panels was okay; defendant failed to explain how people who participate in surveys on a regular basis may skew the results. Screening out people who completed the survey on a smartphone was also okay because of the smaller screen size shrinking the visual stimuli.

Defendant also challenged the survey questions, arguing that the survey showed respondents the trade dress of Grandma Sycamore’s bread with the words “Grandma Sycamore’s removed,” but didn’t remove the unique spelling of the words “HomeMaid” from the image, thus making the package identifiable by means other than the trade dress.  But defendants didn’t explain how that made the results unreliable, though the jury could weigh it.  Defendant also challenged two surveys because only respondents who answered the questions in a particular way were asked follow-up questions, and that the surveyor also would repeat the respondents’ answer back to the respondent when asking them to substantiate their answer, which increased the likelihood of confirmation bias. Furthermore, Dr. Christensen screened out respondents who completed the surveys too quickly. None of these were fatal; defendant didn’t show how the questions rendered the underlying method unreliable, since the questions themselves were open-ended and not leading. Defendant also didn’t show that screening of results of those who answered too quickly had a disparate impact on those respondents who answered a particular way; if it did so, then exclusion of the survey might have been proper.

Finally, defendant argued that failure to ask whether the respondents would have bought the bread if it wasn’t made locally made the survey unreliable.  Defendant used the tagline “Fresh. Local. Quality.” Dr. Christensen attempted to test whether these advertisements created a false or misleading impression that these were local products and whether this impression was material to whether the respondent purchased bread. But he didn’t ask “Would you have bought the bread if it wasn’t made locally?”  That didn’t make the preceding questions unreliable.
The court also refused to exclude defendant’s survey expert Himanshu Mishra, offered in rebuttal to Dr. Christensen’s surveys.  It didn’t matter that he didn’t conduct surveys of his own.  “Rebuttal experts need not produce extrinsic evidence to be able to testify to perceived surveying flaws…. Dr. Mishra’s testimony is more speculative and theoretical than Dr. Christensen’s actual surveys because Dr. Mishra did not produce surveys of his own. But the rule does not require the exclusion of expert testimony that lacks one hundred percent certainty.”

Also, Bimbo argued that Dr. Mishra shouldn’t be allowed to testify that Dr. Christensen’s failure to secure Institutional Review Board approval prior to conducting his surveys violates the law.  The court held that neither party had adequately briefed the law on the issue.  “If lack of approval does not violate the law then Dr. Mishra cannot testify that it does.”

My final expert note: defendant offered Larry Soter as an expert in the “baking industry”to testify that Bimbo’s ingredients that are used to manufacture Grandma Sycamore’s Home Maid Bread didn’t constitute trade secrets.  Bimbo argued that it would be improper for Mr. Soter to testify that each individual element didn’t comprise a trade secret because the alleged trade secret is the combination of all the steps and ingredients. “[A] trade secret can exist in a combination of characteristics and components, each of which, by itself, is in the public domain, but the unified process, design and operation of which, in unique combination, affords a competitive advantage and is a protectable secret.” However, this rule does not mean that analyzing the individual processes is irrelevant. Finding that some of the components are secret may aid the fact finder in determining whether the combination of the individual processes is a trade secret, and it may be relevant to know how common the individual components of the claimed trade secret are.  Mr. Soter wouldn’t be allowed to claim that the combination of individually publicly known components was not protectable because such a statement would be informing the jury of the wrong legal standard, but there was no indication he would so testify.

No comments: