Every review, with its papers in order.
Review collection is crowded — proof is not. ReviewProof imports your existing reviews, surfaces the patterns the FTC fake-review rule punishes (gating, incentives, bursts, duplicates, AI-generation signals), and publishes a widget that only shows reviews linked to a verified transaction.
Run the free audit → See pricingFree: one location · full risk report · verified-experience widget. No card.
§1 — The free job, complete in one sitting
Heuristics are labelled signals, not verdicts. AI-generation signals indicate patterns associated with generated text; they do not prove a review was AI-written.
Provenance, not collection
Every review gets a chain of custody: source, verified-transaction link, edit and moderation history — including negative reviews kept, which is your proof of no suppression.
Verified-experience widget
The embeddable widget displays only reviews linked to a verified transaction record you attest or import. Any AI summary inside it is visibly labelled as AI-generated.
Evidence packs (Pro)
One export that answers a dispute or an FTC inquiry: the full provenance chain, moderation log and risk report, hash-stamped as a certificate.
§2 — What the audit checks
| Pattern | Why it matters |
|---|---|
| Gating language | Asking only happy customers to review — banned review suppression/gating under 16 CFR Part 465. |
| Incentive markers | Discounts, freebies or refunds tied to reviews must be disclosed; undisclosed incentives are a violation. |
| Burst / timing anomalies | A flood of reviews in a short window is the classic bought-review fingerprint. |
| Duplicate text | The same wording across different reviewers suggests a template — human or generated. |
| AI-generation signals | Phrase and structure heuristics, clearly labelled as signals. The FTC bans AI-generated fake reviews outright. |