Keratindose hair products may be misleading about keratin content even without a survey

Price v. L’Oréal USA, Inc., 2020 WL 4937464, No. 17 Civ. 614 (LGS) (S.D.N.Y. Aug. 24, 2020)

Plaintiffs brought California and New York claims against L’Oréal based on its Matrix Biolage Advanced (MBA) haircare product. This includes the Keratindose system of three products: the Pro-Keratin + Silk Shampoo, the Pro-Keratin + Silk Conditioner and the Pro-Keratin + Silk Renewal Spray “Keratindose” appears on the front with the first seven letters of the word in bold. But:

Keratin is a protein that is found in human hair and is also a treatment that customers administer to their hair. The Products do not and have never contained keratin as an ingredient, and the ingredients lists on the back labels of the Products do not include keratin.

The court partially granted and partially denied L’Oréal’s motion to exclude the testimony of Bruce Silverman as a marketing, advertising and branding expert. He opined, inter alia, that “a reasonable consumer would fully expect a family of hair care products named Keratindose to include keratin, just as they would expect of any product that includes a well-known ingredient as part of its name.” L’Oréal argued that his expertise was inadequate because his “career has provided him with no experience in in-salon hair care products or the salon channel generally.” However, “Silverman has an impressive fifty years of experience in advertising,” including reviewing thousands of studies, interviewing thousands of consumers, and attending thousands of focus group sessions, many devoted to health/beauty aids. His opinions premised on his own experience were admissible, including that a reasonable consumer “would expect [ ] any product that includes a well-known ingredient as part of its name” to include that ingredient, and “consumers would see the Challenged Claims as branded ingredient(s) that differentiate the Challenged Products from competitive products.” However, his opinion that keratin is a well-known ingredient in hair products, and that “many women ... are already aware of (or have ample opportunity to be aware of) the restorative properties of keratin” in hair products was not based on his experience, nor is it based on a reliable methodology, just on an internet search and a review of emails and testimony from L’Oréal employees and the PTO ruling on L’Oréal’s trademark application.  Thus, his opinion on consumers’ awareness of keratin as an ingredient in haircare products and consumers’ resulting perception of the challenged terms were excluded as unreliable, including his opinion that “many shampoos actually contain keratin and feature that word on their labels,” and the subsequent conclusion that “[c]learly, keratin is a desirable ingredient that many consumers would believe to be a valuable component in a shampoo, and if they see that word on a label, they would expect it to be in the product.”

The court also excluded plaintiffs’ expert’s damages calculation. The presently unchallenged model concluded that 21% of the price reflected keratin content claims or a price premium of 7% of the price charged. Defendants do not seek to exclude this opinion in their current motion. But the calculation of aggregate economic damages was excluded. The expert used the formula:

Aggregate Economic Damages = (# bottles) × ($ price) × (% damages)

But the actual retail price of the products was not produced, so the expert used the low end of the range of the manufacturer suggested retail prices (MSRP). And L’Oréal doesn’t report bottle sales by state, so he used the percentage of conjoint survey respondents who reported that they had purchased a Matrix Biolage Advanced Keratindose “shampoo/conditioning product,” and who also reported that they were residents of California or New York, to apportion bottle sales to the two states. Also, because L’Oréal only provided wholesale data, and because in practice not all wholesale bottle sales will be resold to consumers, he proposed either a shrinkage rate of 5% or, alternatively, using only wholesale orders that have a corresponding re-sale order at a later date, on the assumption that re-sale orders occur only when customers have sold their previous inventory.

This was excluded because certain of the assumptions about quantity and price were are unreliable. The evidence about MSRP was that, though L’Oréal believed that "the MSRP lists represent good proxies for the prices that salons charge consumers for L’Oréal Products," non-salon retailers charged widely varying and unpredictable prices. Since the class here wasn’t restricted to salon purchasers, MSRP could not be used as an accurate substitute for the actual prices paid by Class members. 

This unreliable methodology for calculating price was exacerbated by questionable assumptions about number of bottles. The proposed shrinkage rate of 5%, “more than two to three times the average of reported shrinkage rates in the United States,” wasn’t sufficiently justified; just being “conservative” wasn’t enough. Nor was there enough information about reorders for that measure to be accurate. In addition, the conjoint survey was initially designed to measure consumer preferences and economic damages as a percentage of actual price charged, and for that purpose had a sample size of one thousand respondents. It wasn’t appropriate to retro-fit that to determine the percentage of the total sales that occurred in California and New York, especially given the small sample size (105) for the damages calculation, and there was no confidence interval provided, which was significant “since the addition or subtraction of a single survey respondent living in California or New York stating they had purchased the shampoo could account for hundreds of thousands of dollars in class damages, and millions in statutory damages under the GBL.”

However, defendants weren’t entitled to summary judgment based on the claim that reasonable consumers wouldn’t be fooled. Plaintiffs offered an expert opinion, and there was also anecdotal evidence showing that keratin was a known ingredient among consumers; the named plaintiffs each testified that they believed, at the time of purchase, that the products contained keratin based on the labeling. This was enough to go forward. (However, emails between L’Oréal employees and the PTO ruling weren’t evidence of deception because they didn’t show the perspective of a reasonable consumer—which would make the PTO sad to hear, I think.)

“In essence, Defendants are arguing that, because that the Challenged Terms are not false on their face, Plaintiffs must proffer extrinsic evidence to show consumers’ understanding of the terms in the form of consumer data or a survey.” But California courts have expressly rejected a survey requirement for misleadingness, and there was no authority that an expert opinion was insufficient under the GBL.

L’Oréal also wasn’t entitled to summary judgment for failure to show damages. The expert formula for  calculating aggregate class-wide damages wasn’t excluded, and non-expert evidence might be applied to this formula to prove aggregate class-wide damages (citing precedent that classwide damages calculations under California law are “particularly forgiving” and require only “some reasonable basis of computation”).

Equally, plaintiffs weren’t entitled to summary judgment on deceptiveness. The meaning of the claims was ambiguous. “ ‘Pro-Keratin’ has no clear meaning, and consumers could understand “Keratindose” to mean that the product is designed to deliver ‘a daily dose of keratin’ as Plaintiffs allege, or understand it to mean that the product is designed to treat hair that has undergone a keratin treatment as proposed by Defendants, or understand it to mean something else entirely.” A reasonable jury could reject Silverman’s expert testimony and agree with L’Oréal.