FTC’s Confusing Guidance on How Merchants Should Manage Their Consumer Reviews

This blog post covers an FTC closing letter. A few words about FTC closing letters if you’re not familiar with them. When the FTC staff open an investigation but then decide not to take action, staff issues a “closing letter” explaining why they didn’t act. Often, the investigated company helps to secure a closing letter by making some changes or promises to the FTC’s satisfaction. While closing letters don’t have precedential effect, they provide a window into the FTC staff’s thinking–with the caveat that no one is pushing back on the FTC’s views of the law, because the investigated company is relieved to hear the FTC is dropping the matter. Thus, I view closing letters as previews of what might be coming from the FTC, but not definitive statements of the law.

The recipient of this closing letter is Yotpo, a company I’d not heard of before. Yotpo describes itself with a remarkable buzzword salad:

We’re on a mission to provide brands with everything they need to win in a customer-centric world.

With the most advanced solutions for user-generated content marketing, referrals, and loyalty programs, Yotpo helps brands accelerate growth by enabling advocacy and maximizing customer lifetime value. With Yotpo, brands can effectively leverage social proof to increase trust and sales, cultivate loyal customer advocates, and make better business decisions based on customer feedback.

The FTC focused on how Yotpo helps merchants aggregate and present reviews written by their customers. The closing letter says: “Our investigation concerned whether Yotpo’s star-rating and sentiment filters provided its clients with the means and instrumentalities to easily and deceptively suppress negative product reviews on their websites and mislead consumers that the reviews displayed accurately reflected the views of all purchasers who submitted reviews.”

Hmm. I’ve not seen evidence on this point, but I assume that consumers don’t expect that merchants publish every review submitted to them. Consumers want merchants to screen fake reviews, off-topic reviews, and offensive reviews. At the same time, many consumers likely assume that merchants don’t cherry-pick only the good reviews. (Plus, consumers are suspicious of uniformly positive reviews, so that strategy is probably a bad business decision).

At the same time, merchants might make marketing representations about their review corpus to increase consumer confidence. Any false affirmative representations about the review corpus could be actionable; and it may be actionable if merchants take advantage of consumer expectations about review corpuses even without making any affirmative false representations.

As typical for closing letters, the FTC explains the factors that persuaded it not to act. The FTC favorably cites:

Yotpo’s commitment to implement measures to protect against the misuse of its review management services to suppress or delay the posting of negative product reviews. To this end, Yotpo is implementing clear and prominent guidance to its clients on their need to promptly post reviews, including negative reviews.[FN] Yotpo will also automatically post negative reviews that have not been promptly reviewed and acted upon by its clients.

So Yotpo promises to tell merchants not to let negative reviews fester in the queue and to automatically post any reviews that festered. Easy concessions for Yotpo to make.

(Note: the FTC targeted Yotpo, a review service vendor, not any individual merchant. The FTC loves pursuing intermediaries rather than individual bad actors. However, Yotpo’s “concessions” may reflect Yotpo’s lack of investment in the issue).

Unfortunately, the FN highlights why it’s impossible for the FTC to regulate the presentation and characterization of consumer review corpuses. The FN says:

FTC staff does not believe that sellers are required to display customer reviews that contain unlawful, harassing, abusive, obscene, vulgar, or sexually explicit content or content that is inappropriate with respect to race, gender, sexuality, or ethnicity, so long as the criteria for withholding reviews is applied uniformly to all reviews submitted. We also believe that sellers are not required to display reviews that are unrelated to their products or services. A seller’s “services” include its customer service, delivery, returns, and exchanges. Sellers should not withhold reviews relating to such services

This footnote essentially tries to reverse-engineer content moderation policies for merchant-managed review corpuses. However, it raises many questions.

Some of the footnote language comes from the Consumer Review Fairness Act, which says that review site TOSes don’t raise CRFA problems if they permit removal of any consumer review that:

(i) contains the personal information or likeness of another person, or is libelous, harassing, abusive, obscene, vulgar, sexually explicit, or is inappropriate with respect to race, gender, sexuality, ethnicity, or other intrinsic characteristic;
(ii) is unrelated to the goods or services offered by or available at such party’s Internet website or webpage; or
(iii) is clearly false or misleading

Why didn’t the FTC track this language identically? The gaps/omissions don’t make sense. For example, the closing letter uses a catchall, “unlawful,” that might cover several omitted categories, such as libelous, false/misleading, or “containing the personal information or likeness of another person.” However, “unlawful” is narrower than the materials that might fit into those more specific categories. The closing letter also omitted the reference to “other [inappropriate] intrinsic characteristics.”

The footnote also added some screening criteria beyond what the CRFA mentioned. The footnote contemplates that the “criteria for withholding reviews is applied uniformly to all reviews submitted.” However, as every content moderation expert knows, “uniformity” in content moderation is impossible. Indeed, the footnote permits filtering for “inappropriate,” “vulgar,” and “abusive” reviews, and each term is so intrinsically amorphous that no one could ever implement them consistently.

There is a second-order problem with the FTC’s standards. How will the FTC validate the merchant’s content review standards to determine if, in fact, the merchant applied the standards consistently? The FTC snooping into the content moderation decisions of review services sounds like a serious incursion into editorial practices. This verification issue will be my next big paper project for 2021.

The closing letter also added to the CRFA language a definition of what constitutes a seller’s “services” (customer service, delivery, returns, and exchanges) and says “Sellers should not withhold reviews relating to such services.” Is the FTC thinking that merchants must carry all submitted reviews “relating to such services”?

The closing letter sidestepped one other obvious scenario. Many merchants post customer “testimonials,” including praise drawn from consumer reviews. If the merchant cherry-picked only the positive reviews and called them testimonials (in a way that complied with the FTC’s endorsement and testimonials guidelines), what outcome?

Consumer reviews are a net win for consumers, but review corpuses need active management to ensure they actually help. The closing letter is a reminder to merchants that the FTC is watching how they moderate consumer reviews–and that the FTC expects merchants to play fair. Ultimately, I believe the FTC is trying to say that merchants offering a consumer review function can’t screen out negative reviews just because the criticism stings. Unfortunately, translating that principle into specifics isn’t so easy.

Citation: Re: Yotpo, Ltd., FTC File No. 202-3039, letter from Serena Viswanathan to Amy R. Mudge and Randall M. Shaheen dated Nov. 17, 2020.

BONUS: Beyond 79, LLC v. Express Gold Cash, Inc., 2020 WL 7352545 (W.D.N.Y. Dec. 15, 2020). This is a false advertising lawsuit between competitors over allegedly merchant-controlled review sites that portray themselves as independent but really bash the competition. The court lets the Lanham Act claims and a few others proceed beyond a motion to dismiss. Professor Tushnet’s coverage.