Step-by-step method
A reproducible path from question to conclusion.
- 1
Set the period and platforms
Define the observed window, affected locations or products and normal review volume. Without a baseline, ordinary variation can look exceptional.
- 2
Preserve the posts
Record URL, date, rating, text, any reply and platform rules. Work from copies and collect only profile data necessary to answer the question.
- 3
Classify the content
Separate detailed experience, general opinion, clear error, apparent duplicate and verifiable claim. Similar wording can arise from a shared event or template.
- 4
Test coordination
Examine timing, unusual repetition, accounts and external relays. Compare these clues with news, service changes and legitimate campaigns.
- 5
Respond and monitor
Correct checkable facts, resolve valid complaints and use reporting tools for rule violations. Track the trend without amplifying marginal posts.
A timeline needs a baseline
Read a surge against prior volume, seasonality, promotions and service incidents. One spike alone cannot establish its cause.
Keep positive reviews and the organisation’s replies too, so the analysis reflects what users actually saw.
Test competing explanations
People may use the same words because an issue became public. Coordination becomes more plausible when several independent clues converge: timing, rare identical phrasing and shared relays.
State confidence and what remains unknown before distinguishing organised action, spontaneous mobilisation and real customer problems.
Reply without escalating the situation
A short public reply can acknowledge established facts, provide a resolution channel and avoid personal data. Unsupported accusations can create a new risk.
Removal requests should identify specific rule violations with URLs and evidence rather than target a merely unfavourable opinion.
What to check and what to record
Use this grid to turn the method into a reviewable case file. A missing item remains an open question. The interpretation limit prevents a finding from becoming an unsupported conclusion.
| Check | Useful record | Interpretation limit |
|---|---|---|
| Baseline | Review volume and cadence over a comparable period. | A legitimate campaign can abruptly change volume. |
| Similarity | Relevant excerpts, dates and documented repetition. | Similar wording does not prove coordination. |
| Handling | Affected reviews, cited rule and platform ticket. | A genuine negative review must not be equated with fraud. |
Common pitfalls
Four shortcuts that weaken the result.
Equating criticism with fraud
A negative review may be legitimate.
Counting copies as sources
Repetition is not independent corroboration.
Publishing profiles
Exposing authors can harm privacy and amplify the issue.
Replying in anger
A premature accusation can displace the facts.
Practical questions
Frequently asked questions.
Do same-day reviews prove coordination?
No. A real event may prompt simultaneous reactions. Look for other clues and context.
Can a false review be removed?
Use the platform’s process with precise evidence; a suspicion alone may be insufficient.
Should every review receive a reply?
Prioritise verifiable issues and resolution needs according to volume and platform rules.
Editorial scope
Published by Internet Intelligence Service on 23 September 2026. Last content update: 6 October 2026. This educational guide describes a lawful, defensive method. It is not legal advice, an emergency service or authority instruction.