survey evidence disregarded in another vanilla case

Twohig v. Shop-Rite Supermarkets, Inc., 2021 WL 518021, No. 20-CV-763 (CS) (S.D.N.Y. Feb. 11, 2021)

ShopRite sells organic vanilla soymilk. Plaintiffs brought the now-standard vanilla versus vanillin claims. They argued that the ingredient list, which includes “Organic Natural Flavors” and “Organic Vanilla Extract” “fails to clarify any front label ambiguity” because organic vanilla extract contributes less to the Product’s vanilla taste “than the front label and the ingredient list would have consumers believe.” Their consumer survey allegedly found that over forty-three percent of consumers expected the origin of the Product’s vanilla taste to be “vanilla beans from the vanilla plant” and that almost fifty-five percent of consumers would be less likely to purchase the Product if the taste were due to imitation vanilla flavoring.

Even accepting as true that the product wasn’t predominantly or exclusively flavored by vanilla beans, plaintiffs failed to plausibly allege that a reasonable consumer would in fact conclude that the word “vanilla” on the Product’s front label implies that the Product’s flavoring was derived exclusively or predominantly from vanilla beans. “A reasonable consumer would understand that ‘vanilla’ is merely a flavor designator, not an ingredient claim.”

Why disregard the survey? The survey couldn’t change what reasonable consumers would understand. [Comment: trademark law bounces rather casually between normative and empirical understandings of the reasonable consumer. These cases provide great evidence that false advertising law does so as well.] Anyway, the survey wasn’t great. Although plaintiffs alleged that “over 43% (a plurality) believed the origin of the vanilla taste comes from the vanilla plant,” the survey didn’t show that over 43% of the respondents believed the flavor in the product came predominantly or exclusively from vanilla beans, as alleged. The survey offered options about what the label said about the source of the vanilla taste including “‘That it comes from vanilla beans from the vanilla plant’ and ‘That it comes from both the vanilla plant and non-vanilla sources.’” But it could have asked whether the flavor came 100% from the vanilla plant, so the implicit contrast wasn’t enough. [Is it likely that the results will be significantly different if you do rewrite the first option that way? It’d be interesting to find out.]

Also, the survey asked, “What does the label pictured above convey about the origin of the vanilla taste?” This didn’t give participants the option of stating that they believed that the label conveyed nothing about the origin of the vanilla taste, so it was too unreliable to be plausible. Kicking out a survey on a motion to dismiss is quite a thing, but seems popular in these cases, even as Lanham Act cases say you don't even have to plead a survey even if you will eventually have to provide one.

And even if federal regulations were being violated, the complaint didn’t allege that reasonable consumers were aware of these complex regulations. [Do you have to be aware of them to be guided by them? I have no idea how octane levels in gas are assessed, but I rely on the existence of a standard regardless.]