What should you do in the first 48 hours of a one-star attack?
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
Preserve, import, separate, answer once, file. In the first hours: do not delete anything, do not reply in anger, and do not buy anything. Import the new reviews alongside your full file and split the pile-on into three groups — transaction-linked reviews (real customers, however angry: answer them well, individually), UNLINKED reviews matching your real customer pattern (probably real, unmatched: look harder for their transactions), and rows carrying cluster patterns (tight timing, similar or AI wording, flagged fingerprints: flag internally as a group). Then file the platform dispute for the third group with the evidence pack, and let the record, not your adrenaline, set the pace.
What is a one-star attack?
A one-star attack is a sudden concentration of negative reviews arriving faster than the business's normal rate, comprising some mix of fabricated reviews, coordinated genuine grievances (a viral complaint, a service failure affecting many customers at once), and opportunistic pile-ons. The triage problem is that the three kinds look identical at the level of the star and entirely different at the level of the row — which is why import-and-separate precedes every other act.
Hour by hour: the order that protects you
Hours 0–2 are preservation: screenshot nothing as your system of record (the import is the record), change nothing in your solicitation, and brief the team on the three prohibitions — no replies yet, no deletions anywhere, no 'offsetting' purchases, which a panicked manager can commission faster than any attack can damage you. Hours 2–12: import, reconcile counts, run the separation. The linked group gets individual, human, safe-grammar replies drafted the same day (the AI draft's apology frame is a good floor for tone when your own is shaking). Hours 12–48: the cluster group's internal flags are logged with the combination named, the dispute is filed with pack, and one public-facing statement — if the attack is visible enough to need one — states facts: volume, the business's records shared with the platform, genuine customers invited to make contact.
Keep the negative reviews. That instruction surprises people mid-crisis, but suppression is itself a violation under the FTC rule, and a profile scrubbed clean during an attack reads as manipulation to anyone who later examines it. ReviewProof has no silent-delete path: moderation decisions are kept, published or flagged internally, each one is logged in the chain of custody, and the evidence export counts the negatives you kept. The strongest response to a suspicious review is a boring, complete record — the row, the dates, the signals, the decision, all preserved while you dispute through the platform's own process.
The mixed attack is the normal attack
Pure fabrications are rarer than mixed events: a real failure (a terrible Saturday, a viral post) produces genuine one-stars, and fabricators ride the wave because cover is perfect. Your separation must therefore survive the discovery that some of the pile-on is deserved. That discovery is not a defeat — it is the triage working. The deserved rows get operational answers (what changed, by when, logged in the record against each row), the unmatched get a harder provenance search, and the cluster gets disputed as a cluster with its combination documented. A business that can stand up on day three and show all three handlings, each evidenced, has converted an attack into the strongest provenance story it owns.
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
The instant public claim 'these are all fake' is falsified by the first genuine row a journalist or the platform finds, and your credibility for the real cluster dies with it. Replying to each review defensively in the first hours writes tomorrow's screenshots. And incentivising happy customers to 'balance' the profile mid-attack creates a dated, linked burst of bought sentiment on top of the attack — the single most self-destructive response available.
Triage starts from your own data, not from the public profile. Import the reviews, look at which rows carry which signals, and separate three questions that are often mashed together: is this review about a real transaction, was it solicited in a way the FTC rule prohibits, and what record do we hold if anyone asks? ReviewProof answers the first two from the imported rows — transaction link present or UNLINKED, gating or incentive language present or absent — and answers the third with the provenance chain. What it will not do is declare a review fake: signals are heuristics for human review, and acting as if they were verdicts creates a new problem while solving none.
Worked example: a salon pile-on that was one-third real
A fictional worked example, written to show the numbers and the decisions in practice.
A salon woke to 23 one-star reviews after a colourist's disastrous Saturday and a local group post about it. Import split them: 8 transaction-linked (Saturday's actual clients — each got an individual reply and a remediation offer logged in the record), 6 UNLINKED but locally specific, mentioning real staff and real events (judged genuine-but-unmatched; 4 were linked within the week when appointment records were searched by date), and 9 forming a pattern cluster — generic wording, AI phrases, two flagged fingerprints, timestamps concentrated between 2am and 4am. Those 9 were flagged internally as a group and disputed with the pack; the platform removed 8. The salon's public posture never claimed 'fake reviews' wholesale — it fixed Saturday publicly and disputed the cluster quietly, which is why both halves were believed.
Your five-action checklist
- Freeze: no deletions, no replies, no purchases in the first two hours
- Import everything and reconcile the count against the platform
- Separate linked, unmatched-genuine and pattern-cluster rows
- Answer genuine reviews individually the same day, in safe grammar
- Flag the cluster with its combination logged; dispute with the pack within 48 hours
Frequently asked questions
How fast should we reply to attack reviews?
Genuine linked reviews: same day, individually. Suspected cluster rows: after the internal flagging and dispute filing, if at all — and only with calm fact.
What if the attack follows a real failure?
Then the triage separates, it does not deny. Fix and answer the real part publicly; dispute only the documented cluster. Mixed attacks are the normal case.
Should we post a public statement?
Only if the attack is publicly visible, and only facts: the volume, that records have been shared with the platform, and an invitation for genuine customers to make contact.
Will the platform remove the cluster?
Platforms decide on their own evidence plus yours. The pack — timing, wording, links, flags, your own house in order — is the strongest input you control.
Keep reading
Related guides from the ReviewProof library:
- What should you do if a competitor is leaving fake reviews?
- How do you clean up a profile with bought reviews in its past?
- Why are employee reviews of your own business a violation risk?
- What does it mean when one reviewer appears at many businesses?
- What should you do when a customer threatens a bad review?
- Can having too few negative reviews look suspicious?
- How do you explain a seasonal spike in your reviews?
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.