The FTC Rule on Fake Reviews and Testimonials

Why this matters

Most shops never write a fake review. They do something smaller: the office manager's brother posts five stars, a tech tells a customer the trip fee goes away "if you can do five stars for me", the website widget shows the average of the good reviews only. Each of those sits inside a federal trade regulation that has carried civil penalties since October 21, 2024, and none of them felt like fraud to the person doing it. This card is about the review program a real shop runs, held up against the rule, rather than about the rule's text.

The rule in one paragraph, and who else is watching

The FTC's Trade Regulation Rule on the Use of Consumer Reviews and Testimonials, 16 CFR Part 465, is a federal rule that reaches businesses of every size. Its prohibitions, in brief: fake or AI-generated reviews about people or experiences that do not exist (465.2), incentives conditioned on the review being positive or negative (465.4), undisclosed reviews and testimonials from insiders (465.5), review sites the business controls presented as independent (465.6), review suppression by threat or by misrepresenting the reviews you display (465.7), and bought followers or views (465.8). A sibling card walks each prohibition and the AI angle in detail; see related: AI-Written Reviews and Review Responses: The Rules.

Two other layers sit on top. State law: every state has an unfair and deceptive practices act, and attorneys general in several states, New York among the most active, have pursued fake-review cases under their own statutes, before the federal rule existed and independently of it. Platform policy: the review sites have their own terms, which are often stricter than the rule (some ban any incentive at all), and a platform can remove reviews or flag a listing without any regulator involved. A practice can pass the federal rule and still fail a platform or a state.

This is not legal advice. Where this card says a practice is exposed, the question of how exposed is one for your own attorney.

The standard that catches small shops: "knew or should have known"

Several of the prohibitions turn on whether the business knew or should have known about the conduct. That phrase matters more to a ten-person shop than to a large company, because in a small shop the owner usually does know. The office manager mentions her brother "helped us out". The tech's bonus sheet lists reviews by name. If the conduct was visible in the shop's own records, "I did not tell anyone to do that" is a weak position. The practical consequence: a review program needs a written rule that staff have seen, and someone needs to look at what is actually being posted, because the rule reaches what the business should have noticed.

What the rule does not touch

Shops over-correct as often as they under-correct, and stop asking for reviews entirely. The rule leaves a lot alone:

• Asking every customer for an honest review. Fully allowed, and the core of a healthy program.
• An incentive for leaving any review, good or bad. Not prohibited by 465.4, which reaches only incentives tied to sentiment. It should be disclosed in the review, and many platforms forbid incentivised reviews outright, so the federal answer is not the last word. Read the platform's terms before running a drawing.
• A curated testimonials page. Choosing your favorite customer quotes is advertising, not suppression, as long as the page does not claim to show all or most of your reviews and each quote is from a real customer who agreed to the wording.
• Replying to reviews. Your reply is your own speech as the business. Section 5 of the FTC Act still forbids it to misstate facts, but nothing in Part 465 limits it.
• Reporting a genuinely fake or abusive review to the platform. Using the platform's own process for a review that violates its terms is not suppression. Threatening the reviewer with a lawsuit you have no basis for is.

Insider reviews: where the edges are

Section 465.5 is the one that most often catches a shop by surprise, because the people involved are trying to help.

• Your own staff reviewing the business. An officer's or manager's own review, and a review the business asks an employee or agent to post, are covered unless the relationship is clearly disclosed. An employee who posts unprompted sits outside 465.5's text as written, but is still a Section 5 and platform problem. The fix is not a ban on staff ever posting; it is a rule that any staff review says so in the text ("I work here, but...") or does not get written.
• Relatives of staff. Soliciting reviews from the immediate relatives of staff, or having staff solicit their relatives, without disclosure is covered too. The rule treats a general request sent to your whole customer list, which happens to reach a relative who is a real customer, differently from a targeted ask; read 465.5 for how. A relative who never hired you writing about a job they did not have is not an insider problem at all; it is a fake review under 465.2.
• Testimonials from insiders. A website quote from the owner's neighbor who is also a part-time estimator needs the relationship stated.

What changes the answer: a former employee. The rule's insider categories are written around current relationships, so a review from someone who left two years ago and then hired you as a customer is closer to an ordinary customer review. Someone who left last week and posts about the business is a closer call, and state law does not necessarily draw the line where the federal rule does.

Incentives: the customer side versus the staff side

Read 465.4 as a rule about what the reviewer is offered. That makes a common shop practice turn on a detail:

• A tech bonus that pays per review naming the tech is a staff incentive. Part 465 does not prohibit it. It creates pressure, though, and pressure produces the next item.
• A tech who tells the customer "I'll waive the trip charge if you leave five stars" has now offered the customer something of value conditioned on a positive review. That is the prohibited act, and the shop's bonus program is what made it likely.

So if you pay for reviews internally, pay for the ask (the review request sent, logged in the job record) or for honest reviews of any star count, and write the script the tech may use. A bonus counted only on five-star reviews teaches the exact behavior the rule prohibits.

Review gating is the related trap. Some review tools send a "how did we do?" survey first and route only happy customers to the public site, sending unhappy ones to a private form. Part 465 as finalized does not contain a standalone gating ban, but the FTC's revised Endorsement Guides (16 CFR Part 255) treat selective solicitation that distorts the overall picture as a deception risk under Section 5, and Google's own policy prohibits gating outright. Turn the routing off; send everyone the same public link and handle the unhappy ones by calling them.

Worked example: a nine-person shop audits its program

An owner runs a nine-person plumbing shop: owner, office manager, dispatcher, five techs, one apprentice. A competitor's attorney-general complaint made the rounds, and the owner spends an afternoon reviewing the shop's last 60 public reviews and every practice that touches them. Six findings:

1. The office manager's husband posted five stars describing a water heater install. He is not a customer; the install was at their own house, done by a tech on a weekend at no charge. That is both an undisclosed insider-relative review and, because no customer relationship existed, close to a fake one. Action: he deletes it. The owner does not ask the platform to keep the star count.
2. Two techs posted reviews under personal accounts in their first month. Action: both edit the text to state they work at the shop, or delete. The written rule goes into the handbook.
3. The tech bonus pays a fixed amount per five-star review naming the tech. In the 60 reviews, 41 name a tech, and the owner finds one written note in a job file ("cust asked if 5 stars gets the discount - told her yes") from a tech who waived a trip charge. That one job is a 465.4 problem the bonus created. Action: the bonus now pays on every logged review request sent through the job record, regardless of whether a review follows or what it says, and the waiver practice stops by written rule.
4. The review tool was gating. Customers who answered the survey with 1 to 3 stars were sent to a private form. Over the period, the tool's own log shows 9 of 72 surveys routed privately, so 63 of 72, about 88 percent, were sent the public link. Action: routing turned off; all 72 would now get the same link.
5. The website widget reads "4.9 average" and pulls only reviews of 4 stars and up. That is the displayed-reviews misrepresentation 465.7 reaches, because the widget presents a filtered set as the shop's rating. Action: the widget either shows the platform's own unfiltered rating or is relabelled "Customer stories" with no average.
6. A monthly drawing for anyone who leaves a review. Allowed under 465.4 because it is not tied to sentiment, but the owner reads the platform terms, which forbid incentivised reviews. Action: the drawing ends. Federal permission did not survive the platform layer.

Count the outcome: of six practices, five were rule findings (1 through 5), of which one (finding 2) was fixed by disclosure rather than removal, and one (finding 6) was allowed by the rule and ended for platform reasons alone. The star average on the platform drops slightly once the husband's review is gone. That is the honest number, and it is the one a regulator, a platform or a competitor's complaint would have been measured against anyway.

The failure mode the owner nearly committed: after finding the gated surveys, the instinct was to ask the 9 privately-routed customers to "go ahead and post" now. Re-soliciting only the customers whose complaints were resolved is gating with a delay. If those customers get a public link, it is because everyone does.

How to check your own program

• Pull your last 50 or so public reviews and mark any written by staff, relatives, or people who never hired you. The target is zero undisclosed.
• Read your review tool's settings for any routing by score.
• Read your bonus sheet: does it count stars? If yes, it counts the wrong thing.
• Read every incentive against the platform's terms, not just the rule.
• Look at your website and every ad for a rating or average and confirm what set it is drawn from.
• Put the rule in writing, have each employee read it, and keep the signed copy. That record is the answer to "should have known".

References

• FTC, Trade Regulation Rule on the Use of Consumer Reviews and Testimonials, 16 CFR Part 465 (effective October 21, 2024)
• FTC, Guides Concerning the Use of Endorsements and Testimonials in Advertising, 16 CFR Part 255
• Federal Trade Commission Act, Section 5 (15 U.S.C. 45)
• See related: AI-Written Reviews and Review Responses: The Rules; Asking for the Review Without a Bribe
• See related: Unfair and Deceptive Practices Laws Behind Every Complaint