It’s been a rough couple of months for use of surveys in the courtroom. First, Judge Richard Posner calls into question the use of consumer surveys conducted by one of the parties in a lawsuit. Then, a district court judge excludes the testimony of an expert who did not explain the connection between a survey he relied on and his conclusion regarding damages.
For as long as I can remember, I—along with many of my family and friends—thought that Cracker Barrel was a restaurant that sold cheese. Or maybe it was a cheese company that had a chain of restaurants. Whatever. It’s pretty confusing. So confusing in fact, that it’s been litigated. And out of that litigation has come a warning to those who rely on consumer surveys in litigation.
For many years Kraft Foods had been selling packaged cheeses with the “Cracker Barrel” brand in grocery stores.
Also for many years, Cracker Barrel Old Country Store has operated a chain of restaurant/gift stores, usually located near highway exits.
More recently, the Old Country Store group announced plans to sell packaged food items—but not cheese—in grocery stores.
Kraft objected to the Old Country Store’s sales in grocery stores, arguing that consumers would be confused by the similarity of the logos. They argued that consumers would think that the Old Country Store’s food products were manufactured by Kraft, and then blame Kraft if they were dissatisfied with a Old Country Store product. Judge Richard Posner seems to agree, concluding:
A consumer who thinks Kraft makes bad hams may think it probably makes bad cheeses as well.
Judge Posner agreed with the district court’s decision of a substantial likelihood of confusion between the labels, and that the resulting risk of damage to Kraft’s goodwill could be significant and irreparable—thus outweighing any harm to the Old Country Store. Therefore, the grant of a preliminary injunction was affirmed.
Posner’s problem with surveys
As a bit of a bonus, Judge Posner shared his misgivings about surveys of consumer confusion conducted by expert witnesses hired by one of the parties. He goes so far as to say that such surveys are “prone to bias.” Judge Posner says the problems are:
- Survey questions may be phrased in a way that is intended to elicit the surveyor’s desired response from the survey respondents and may reflect a bias in favor of the party that hired the expert;
- Reactions of consumers to photographs of trademarks shown to them by an expert may differ from their reactions when they encounter the same labels while actually shopping and making purchase decisions; and
- Surveys that produce results contrary to the interest of the party that sponsored the survey may be suppressed and never become a part of the trial record.
Judge Posner identified statistical studies or expert testimony concerning buying habits and consumer psychology as alternatives to the use of surveys.
A completely different matter involving alleged patent infringement sends another caution on the use of survey results in calculating damages.
A patent-holding company, Rembrandt Social Media, sued Facebook for infringing two patents. The plaintiff says its patents came from a social networking pioneer who created an early “online diary” program, and is demanding that Facebook pay it royalties.
Rembrandt claims that Facebook “bears a remarkable resemblance, both in terms of its functionality and technical implementation, to the personal web page diary” technology covered by its patents. For example, Facebook lets its users arrange personal information, as well as third-party content, in a chronological format; it allows the user to share “specific diary entries with a selected group of people, such as the user’s friends, through the use of user-settable privacy levels.” And Facebook is powered by advertising revenue, a business model that’s specifically described in one of the patents.
In formulating a reasonable royalty, one of the steps performed by Rembrandt’s expert, James E. Malackowski, was to rely on customer and advertiser surveys used to rank the importance of various features of Facebook. The survey results were used to exclude the amount of revenue attributable to features not causing Facebook to infringe.
Note—and this is key—the expert did not attempt to determine the revenue attributable to the two features alleged to cause Facebook to infringe on Rembrandt’s patents.
Good surveys going bad
Judge T. S. Ellis expressed some seemingly minor misgiving about the surveys themselves, but found Mr. Malackowski’s application of the survey results were unreliable to the point of excluding his damages calculations. In particular, Judge Ellis concluded:
- The surveys did not test the importance of the two alleged infringing features, nor did they test for the “importance of Facebook’s global presence” (whatever that means) and the number of Facebook users;
- The expert who conducted the surveys admitted that they were not meant to be used to address “the revenue question” and that linking the surveys to the revenue question required a “separate analysis.”
- Mr. Malackowski assumed—without explanation—that that the weighted importance of any given feature is exactly equal to that same percentage of advertising revenue. For example, if 4.0 percent of survey respondents said a feature was important, then the expert assumed that 4.0 percent of Facebook’s revenue is generated by the feature. Because the expert did not explain his weighting scheme and did not perform the “separate analysis,” the methodology is “suspect and unreliable” under Rule 702, Fed. R. Evid. and Daubert.
Cracker Barrel and Facebook cases offer two lessons on the use of surveys:
- Courts are suspicious of survey conducted by parties or their experts. It would be helpful to support survey results with other information such as statistical analysis or expert testimony.
- Experts should explain the connection between survey results and the experts’ conclusions, especially if the surveys were designed to address a different issue.
The cases are:
Kraft Foods Group Brands LLC v. Cracker Barrel Old Country Store, Inc., 2013 WL 6017396 (7th Cir. Nov. 14, 2013).
Rembrandt Social Media, LP v. Facebook, Inc., 2013 U.S. Dist. LEXIS 171127 (E.D. Va. Dec. 3, 2013).