No gating. No paying. No fake stars.

The Federal Trade Commission’s Endorsement Guides (updated 2024) lay out clear rules for review-solicitation tools. Revue follows them — not because we have to, but because the only review economy worth building is one customers can trust.

A review is only as valuable as the trust behind it.

1. We do NOT gate by rating.

1-star and 5-star reviewers get the exact same path to Google. No “consider sending us private feedback first.” No hidden submit button for low ratings. No different button color, size, position, or copy.

• The “Post to Google” button is identical at every rating — same size, same color, same position, same word.
• The “Send privately to owner” option exists atevery rating, not only low ones — happy diners can send private kudos too.
• We track submit-rate by rating as a self-audit metric. If 1–3 star submission falls more than 15% below 4–5 star submission, we treat it as an effective-gating alarm and redesign.

2. We do NOT pay for reviews.

No discount codes contingent on a review. No free menu items in exchange for a post. No loyalty points, no raffles, no perks of any kind tied to whether a diner submits to Google. Revue is a writing assistant; it is not an incentive program. If a merchant asks us to enable an incentive feature, the answer is no.

3. We do NOT generate fake reviews.

AI helps the diner write what they actually thought. The diner edits the draft. The diner taps Submit on their own Google account. Their words, their account, their accountability.

• For 1–3 star ratings, we do not generate AI drafts at all. The text box starts empty. The diner writes in their own voice.
• For 4–5 star ratings, AI drafts are bounded to the diner’s actual inputs (stars, highlight chips, dish selection, optional note) plus a strict banned-vocabulary system that prevents day-of-week, clock-time, and meal-segment hallucinations.
• We never auto-submit. The diner must edit at least one character before submitting (a hard front-end check), and must tap Submit on Google’s native interface.

4. We disclose AI assistance to every diner.

Diners see an inline “AI-assisted draft” note below their draft text, with a link to our AI Disclosure page. They know what they are doing before they edit. They know that AI generated the starting words, and that those words are bounded by what they themselves selected.

5. We do not edit, suppress, or filter reviews on Google.

Once a diner posts to Google, the review is theirs and Google’s. Revue does not have administrative access to remove or alter posted reviews. We cannot bury a bad one. We cannot promote a good one. We do not try.

6. We are transparent about our methods.

Every prompt layer, every banned-vocabulary list, every few-shot example pool is documented internally and audited every time we change a prompt. We measure diversity (Self-BLEU 0.002, target ≤ 0.35) and hallucination rate (0% on our canonical test set) before every prompt deploy.

If you see us behaving inconsistently with this policy

Tell us. Email support@revue.today with the subject line “Policy concern” and we will treat it as a critical bug, on the same on-call pager as a production outage. If you would prefer to report to the FTC directly, reportfraud.ftc.gov is the official channel.

If this policy ever changes

This page changes too — with the same date stamp at the top. If we ever loosen a commitment on this page, we will email all active merchants at least thirty (30) days in advance and we will explain why.