How do you audit your own reviews, step by step?
By the Ingenious AI team · Published 14 November 2026 · Reviewed 14 November 2026
Run the free review audit See pricing · See a sample provenance report
Short answer
A self-audit runs in one sitting: gather the location's reviews into a single file with dates, sources, handles and any transaction references; import into ReviewProof (CSV or paste); read the report in order — totals, risk score, per-signal breakdown, verified versus UNLINKED counts; open the flagged rows and judge them in context; record moderation decisions in the chain; fix the solicitation habits the signals point to; and export the evidence pack (Pro) so the audit exists as a dated document, not a memory.
What is a self-audit of reviews?
A review self-audit is a business examining its own review profile with the same instruments an investigator would: the complete set of rows, pattern analysis over those rows, provenance for each row, and a decision record. Its defining discipline is completeness — all reviews, including negatives and the embarrassing ones, in one file, judged by rules set before the results were seen. A partial audit is a marketing exercise with extra steps.
Assemble the file like you mean it
The audit's quality is decided before any software runs, in the gathering. Pull every source the location appears on; include the fields that make analysis possible (reviewer, rating, title, body, date, source) and the fields that make provenance possible (verified flag, transaction reference) wherever your records hold them. Resist the two quiet edits: dropping the review you know is from your cousin, and dropping the one-star rant about the manager. ReviewProof will default everything to kept and log what it finds — but it can only audit the file you give it, and the file's completeness is your signature on the exercise.
Enforcement and platform action both turn on records, not intentions. A business that can produce the underlying row, its source, its date, its transaction link or the honest absence of one, and every moderation decision taken about it, is in a different position from a business holding only a star average. That is why ReviewProof never silently deletes: imported reviews default to kept, moderation is limited to kept, published or flagged internally, and every decision is written into the chain of custody. The evidence pack (a Pro and Agency feature) assembles that chain so it can be handed over as one document.
Decide with rules you wrote before you saw names
Flagged rows land differently when a name is familiar. Protect the audit from that by fixing the decision rules first: incentive or gating rows are flagged internally with the solicitation stopped at source; duplicate-and-burst clusters are investigated as clusters and their conclusion logged; AI-signal rows are noted, clustered, never individually accused; UNLINKED rows are worked for genuine matches and otherwise left honestly labelled. Then apply the rules uniformly — including to the owner's favourite review. The moderation log that results, exported with the pack, is the audit's real output: not a cleaner profile, but a defensible one.
ReviewProof's import is forgiving about format and strict about meaning. Column names are matched flexibly — body, review or text for the review itself; rating or stars; reviewer, author or name; a verified flag and a transaction reference when you hold them. Rows are capped per import by plan (500 on Free, 5,000 on Pro, 50,000 on Agency), which is generous for a single location's history. What the importer will not do is upgrade a review's status: a row that arrives without a transaction link is analysed fully and still labelled UNLINKED, because the label is the product.
Mistakes that undo the work
Announcing the audit's conclusions before running it ('we're confident we're clean') pressures every judgement that follows. Auditing only the current platform profile misses reviews on the sources you forgot you had. And scheduling the audit as an annual event wastes it: the monthly import rhythm is what turns an audit from an excavation into a smoke alarm.
The FTC fake-review rule, 16 CFR Part 465, has been in force since October 2024. It bans fake and AI-generated reviews, buying positive or negative sentiment, insider reviews without disclosure, review suppression, and misrepresenting review sites as independent. ReviewProof was built as the evidence layer for that rule: imported reviews carry a provenance chain — import, transaction link, analysis, moderation decisions — and the Pro evidence export includes the moderation log with negative reviews kept, which is how a business shows suppression did not happen. Proof, in this design, is a record anyone can inspect, not a badge a business awards itself.
Worked example: a one-afternoon audit, start to finish
A fictional worked example, written to show the numbers and the decisions in practice.
A physiotherapy practice audited 342 reviews in an afternoon. Gathering took longest: three sources, one spreadsheet with drifted dates (corrected from the source export, noted). The report showed a risk score of 12, with 19 gating rows quoting the practice's old feedback SMS, 3 incentive rows, no burst, and a verified share of 54%. Decisions followed the pre-written rules: solicitation wording replaced the same day, 22 rows flagged internally with reasons, 31 UNLINKED reviews matched to real bookings and linked, the rest left labelled. Pack exported and dated. The partner's verdict: 'Nothing about our rating changed. Everything about our ability to defend it did.'
Your five-action checklist
- Gather every source into one file, negatives and awkward rows included
- Import by CSV or paste and confirm the imported count matches your file
- Write the decision rules before opening any flagged row
- Read the report in order: totals, score, signals, verified share
- Moderate by the rules, log reasons, fix the solicitation at source, export the pack
Frequently asked questions
How long does a self-audit take?
The analysis runs on import. For most single locations the human work — gathering the file and judging flagged rows — fits in an afternoon; the monthly rhythm after that is minutes.
What file format does the import need?
A CSV with a header row (body/review/text, rating/stars, reviewer, date, source, and verified flag plus transaction reference where held), or paste the CSV text directly.
What plan do I need to audit?
Free covers one location with the full risk analysis, report, certificates and widget. Pro adds the evidence export, full chain detail, reply drafts and the fraud graph.
What if the audit finds real problems?
That is the audit working. Stop the practice, keep the rows, flag them internally with the correction logged — a found-and-fixed record is the strongest outcome a self-audit can produce.
Keep reading
Related guides from the ReviewProof library:
- What does the FTC fake-review rule actually ban?
- Why does keeping negative reviews protect your business?
- What should an evidence pack for a review dispute contain?
- What is a chain of custody for a customer review?
- What is a review provenance certificate?
- What does transaction-backed review provenance mean?
- How should you moderate reviews without suppressing them?
Start with the tool itself: open the free audit desk, see how ReviewProof works, compare Free, Pro and Agency plans, inspect the sample provenance report, or browse all ReviewProof guides.