The signals that work on every platform
Six signals carried more than 80% of the successful removals in our log. They are platform-agnostic. The submission path changes; the evidence does not.
- Reviewer's account is recent (under 30 days) and thin (under 5 reviews)
- Reviewer has reviewed direct competitors in the same vertical and city within a 30-day window
- Reviewer's profile photo returns matches on Google Lens or TinEye reverse image search
- Review describes a service or product the business does not offer
- Multiple new accounts post similar negative reviews within a 72-hour window (coordinated attack)
- Reviewer is identifiable as an employee, vendor, or person with a financial relationship to a competitor
23% aggregate removal rate across 42,318 reports. The variance between platforms is real; the evidence package that wins is the same on all of them.
Build the evidence package once
Before you open any platform's report form, build the package. A screenshot of the suspect review with a timestamp visible. A screenshot of the reviewer's public profile showing the disqualifying signal (the new-account date, the competitor reviews, the small total review count). A second piece of corroboration when one is available: a reverse-image-search result for an AI or stock profile photo, a LinkedIn screenshot showing the reviewer working for a competitor, a Maps screenshot of the geographically impossible posting pattern.
Across 9,847 successful removals in our log, 81% included at least one screenshot and 53% included a second piece of corroboration. Submissions with no attachment removed at 7%. Submissions with one screenshot removed at 31%. Submissions with two pieces of evidence removed at 47%. The pattern held on every platform we measured.
Google: file the BRF after the in-product flag denies
Start with the in-product flag from the three-dot menu on the review. Pick Conflict of Interest, Off-topic, or Spam to match the strongest signal. If the flag comes back denied, file the Business Redressal Form with the review URL, a one-sentence policy citation, the screenshots, and a note that explains the signal in plain language. In our log, BRF submissions on suspected fakes resolved in a median 11 days at a 31% removal rate, climbing to 44% with a screenshot and 51% with two pieces of evidence.
Do not name the suspected reviewer by their real name in the BRF text. Google treats this as a privacy violation and rejects on principle. Reference them by their public Google profile name and let the attached evidence make the connection.
Amazon: Brand Registry beats Report Abuse for fakes
On Amazon, the standard Report Abuse path resolves in a median 8 days at a 14% removal rate. Brand Registry's Report a Violation path, available to enrolled brands, resolves in a median 6 days at 41%. The Brand Registry queue is staffed by a team that recognizes coordinated competitor sabotage faster than the general queue. For any seller doing more than a few thousand units a month, enrollment is worth the multi-week setup, even just for the fake-review path.
Pair the report with a clear off-topic citation when the review describes a feature your product does not have, or a conflict-of-interest citation when the reviewer profile shows a competitor pattern. Vague reports closed at half the rate of cited reports across 2,841 matched pairs in our 2025 Amazon log.
Trustpilot: the identifiable-experience challenge is the lever
Trustpilot's strongest fake-review path is the not based on a genuine experience flag, which moves the burden to the reviewer to provide proof of the transaction within roughly seven days. Reviewers who cannot supply order numbers, receipts, or other transaction evidence have the review removed automatically. In our 2025 log, this single category removed at 41%, roughly triple the rate of general flag categories.
Add a Compliance team escalation through the Help section if the in-product flag denies and you have screenshot-backed evidence of a competitor or ex-employee pattern. Compliance escalations resolved in a median 9 days at a 31% removal rate.
The submission path changes between platforms. The evidence package does not. Two pieces of evidence remove at 47% aggregate; no attachment removes at 7%.
Facebook, Yelp, and Tripadvisor: the smaller wins
Facebook removed 12% of in-product reports and 24% of Meta Business Suite escalations in our 2025 log. Pattern evidence (the same reviewer leaving similar negatives across unrelated pages, or recently created accounts with no other activity) lifted the rate to 38%. The Meta Business Suite Get Support form is the only place to upload pattern evidence; the in-product reporter has no attachment field.
Yelp is the toughest of the six. The platform's automated filter already suppresses many suspected fakes from public display, which Yelp uses as the rationale for declining most explicit removal requests. We logged 12% removal on filed reports. The filtered reviews remain visible on a Not Recommended page; visibility on the main profile is the goal, and adjusting the filter is largely outside the business owner's control.
Tripadvisor removed 24% of fake-review reports in our log. The platform's Management Center accepts written escalations with screenshots, and the strongest path is documenting reviews that describe a property that does not match the listing (wrong star rating tier, wrong amenity description, wrong location). Reports that name a single signal and attach a screenshot remove at roughly twice the rate of unsupported flags.
The FTC rule that finally has teeth
The FTC's Trade Regulation Rule on the Use of Consumer Reviews and Testimonials took effect in October 2024 and started producing visible enforcement actions through 2025 and into 2026. The rule prohibits buying positive reviews, suppressing negative reviews through threats or compensation, posting reviews from insiders without disclosure, and selling fake reviews. Civil penalties begin at $51,744 per violation in 2026.
For businesses, the rule is most useful as the basis for an escalation when a fake-review service has demonstrably attacked the profile and the platform's standard channels have failed. State attorneys general and the FTC have both opened cases against fake-review brokers identified through victim reports. The FTC's report channel at reportfraud.ftc.gov accepts evidence and feeds the broker enforcement docket. It is not a fast personal-removal path. It is a slow strategic path that has begun to remove some of the largest fake-review networks at the source.
What does not work and what hurts the profile
Three patterns we still see weekly that produce no removal and sometimes damage the profile. Filing the same report repeatedly with no new evidence; review queues correlate this as low-quality reporting. Naming the suspected reviewer by their real name in the report text; every platform treats this as a privacy violation. Buying positive reviews to dilute fake negatives; this triggers algorithmic spam-pattern detection on your own profile across every platform we measured, and now exposes the business to FTC enforcement under the 2024 rule.
The right counter to fake negatives is real positives invited from real customers using each platform's official invitation tool, paired with a public reply on the negatives that addresses the substance without acknowledging the fakery directly. Drawing public attention to the suspected fakes routinely makes them more visible than they would have been if quietly reported.
What to do this week
List every platform where the suspect reviews appear. Build the evidence package once: screenshot of each review, screenshot of each reviewer's profile, reverse-image-search result for any suspicious profile photo, LinkedIn or directory evidence for any suspected employee or competitor pattern. File the in-product flag on each platform with the matching category and a one-sentence signal note. For platforms with an escalation path (Google BRF, Amazon Brand Registry, Trustpilot Compliance, Meta Business Suite, Tripadvisor Management Center), queue the escalation now rather than waiting two weeks. The wrong move is filing a single category-only flag on each platform and refreshing the page for the next two weeks.
