It’s a short but meaty decision, providing good guidance for using customer reviews in advertising.

Using NAD’s Fast-Track SWIFT program, PerSé Beauty challenged competitor Function’s claim that Function had “over 110,000 5-star product reviews!”  PerSe argued that the claim was misleading because a review in the “shampoo and conditioner” category was counted by the advertiser as two product reviews, one for shampoo and one for conditioner.  Function defended its double counting on the ground that most users buy both products, so that their reviews in the single category could appropriately count as reviews for each of the products.  However, NAD determined that Function’s substantiation was not a “good fit” with its claim, and recommended that Function modify it.

As an initial matter, NAD recognized the critical importance of product reviews in the e-commerce ecosystem, and noted that “an advertiser [meets] its obligation to provide non-misleading information when it posts product reviews that are what they appear to be -- authentic consumer reviews of the product, representative of the whole range of product reactions, and reflective of the actual opinions of consumers." (emphasis added). NAD noted that, here, Function asked for reviews neutrally (“tell us about your purchase”), didn’t provide incentives to consumers to provide reviews, maintained records in order to verify that reviewers were actual purchasers of the products, and used an independent third party to collected the reviews from users and then post them to Function’s site without editing or alteration.

The problem, NAD determined, was that “ at least one reasonable interpretation of the claim 'over 110,000 5-star product reviews!' is the express message that consumers have submitted 110,000 distinct reviews,” a claim that Function could not support. Rather, Function’s review system, with a forced-choice drop down menu on the review page, required users to review shampoo and conditioner together, as a bundled set. Users could not review the products individually.  And, when users gave the set 5 stars, Function would count that review twice: once, as a 5 star review for the shampoo and once as a five star review for the conditioner. (In those “rare” instances where a consumer had purchased only one of the products, Function would only count the review once towards the aggregate 5 star claim.)  Thus, while Function could truthfully claim that it had received over 55,000 5-star reviews for the shampoo and conditioner category, it could not adequately support a claim that it had 110,000 5 star product reviews. (Would a new user on the site think that 55,000 5-star reviews is that much less compelling than 110,000 5 star reviews?  Inquiring marketing minds want to know.)

Lessons learned?  Here’s an advertiser who appears to have done everything pretty well in its solicitation of reviews.  Real users and real reviews? Check.  A review page that displays the good and the bad?  Check.  No incentives to bias the results?  Check. Unedited reviews?  Check.  But, even so, the advertiser tripped up in how it characterized those reviews.  It got a little greedy in how it counted them up and touted them. The good news, though, is that even though the claim itself will have to change, the advertiser has already instituted a process that will allow it to make other strong claims about how highly its users review its products.

NAD Fast-Track SWIFT Case #6938 (02/09/2021)

(Description of image for visually-impaired readers: drawing of a screen with star ratings and check boxes next to each category.)