Remove fake reviews: prove it and dispute it

You can remove fake reviews when you can show a platform that the reviewer had no genuine experience of your business, and that is a matter of evidence rather than argument. A platform removes a review because it breaks a written rule, so the work is to match what you can prove to the rule it breaks, file one clear report with the proof attached, and use whatever appeal the platform offers.

Most advice on this topic is a list of buttons: find the review, click report, wait. That works for the easy cases and fails on the ones that hurt. A burst of one-star reviews from accounts that never bought anything, a competitor posting under a false name, a former employee settling a score: in each case the platform is looking at your claim and at its own signals, and it sides with whichever is better documented. This guide covers the part that is the same on every platform: what counts as fake, what evidence proves it, what platforms detect without you, how to write the dispute, and what the law adds.

Platform policies and laws change, so everything below describes the pages we read for this guide, with the source named. Where a page could not be read, the text says so. Reporting steps that are specific to Google reviews are a separate procedure and are not repeated here.

The short version

A review is fake when it does not reflect a genuine experience. A review you disagree with, or one from a customer who did not pay or who rated unfairly, is not fake by that fact alone.
Prove the review fake with records you hold: no matching transaction, wrong business, reviewer connected to a competitor or to your own staff, or a pattern across accounts. Suspicion is not evidence.
Platforms run their own detection and weigh your report against it. One precise report with attachments beats five angry ones.
In the US and the UK, fake reviews are now regulated by law, but neither route is a quick removal tool. Defamation is a separate question for your own lawyer.
Never answer a fake review with a fake review, an incentive or a threat. Each one is a violation in its own right.

October 21, 2024the date the US FTC rule on fake reviews and testimonials took effect

4.4%of 30.2 million reviews posted to Tripadvisor in 2022 that it found fake or fraudulent

72%of those fraudulent submissions that Tripadvisor says it stopped before publication

Sources: FTC staff Q&A on the consumer reviews rule (effective date); Tripadvisor's 2023 Review Transparency Report release, covering reviews posted in 2022.

What counts as a fake review, and what does not

A fake review is a review that purports to be based on a real experience of your business and is not. The Competition and Markets Authority, the UK regulator, uses that wording in its fake reviews guidance (CMA208), which also says a fake review can be positive or negative and that a review is not fake simply because the trader disagrees with it.

The same guidance says a review is not fake because the reviewer did not pay for the experience. Someone who tried a free consultation or a quote visit has a real experience, and their review is genuine even if unfavorable.

Trustpilot's guidelines for reviewers give a similar definition: a fake review is one that does not reflect your genuine experience with a business's service or product. Trustpilot adds a second category that business owners often overlook, the review from someone with a special relationship to the business. Its guidelines say owners, their immediate family, anyone working with or for the business, shareholders and competitors are not eligible to review, to prevent biased reviews "both positive and negative."

What platforms usually do not treat as fake

A real customer with an inflated complaint, one who rated the courier rather than you, or one whose name you do not recognize because the order was placed under a partner's or a work email all had a real experience. Check every route before concluding there is no match. For former employees, Google lists current or former employment as a conflict of interest example, while Trustpilot's guidelines, as we read them, do not mention former employees, so read the rule of the platform you are reporting to.

The five kinds of fake review and the proof each one needs

Each kind of fake review leaves a different trail and is proved with a different record. Pick the category before you write, so the report names the right rule.

Kind of fake	What you can usually show	Rule it breaks	Strength of proof
Never a customer	No order, booking or visit under the name, email or date described; details that do not match how you operate	Not based on a genuine experience	Strong if your records are complete
Wrong business	A product you do not sell, a location you do not have, a staff name you never employed	Review is about a different business	Strong, and quick to check
Competitor	Reviewer profile tied to a rival, reviews of your competitors that are all glowing, a pattern of attacking one market	Conflict of interest	Medium: links are often circumstantial
Insider or paid	Staff, relatives or a review broker; identical phrasing across profiles	Fake or incentivized review	Medium to strong, depends on how you found it
Coordinated burst	Many reviews in a short window, new accounts with one review each, copied text, an outside trigger such as a forum thread	Fake engagement or multiple accounts	Medium: platforms weigh it with their own signals

A review can sit in more than one row. Cite the strongest rule first and mention the others in a sentence.

How to build an evidence file before you report

The evidence file is a short folder of records that lets a reviewer at the platform verify your point in about a minute. Build it before you touch the report button, because many platforms give you one report and a limited appeal.

Collect these, in this order.

The review itself. A screenshot with the date and the review's link, and the reviewer's public profile page, captured the same day. Reviews and profiles can change or vanish.
Your transaction records for the window. An export from your order system, booking calendar or point of sale for the dates the review describes. Search by the reviewer's display name, any email you can match, the amounts, the staff and the location.
A statement of the mismatch. Two sentences: what the review claims, and what your records show. Write it as a fact a stranger can check.
Pattern evidence, if relevant. A table of the suspect reviews with date, rating, account age as shown on the profile, and any repeated phrasing. Keep to what is visible on the platform.
Anything showing the connection. A competitor link, an insider relationship or a broker's advertisement. Only include it if you can document it without guessing.

Rules for gathering proof

Use public information and your own records only: no guessing passwords, no posing as a customer, no publishing the reviewer's personal data (Trustpilot lists personal information as grounds for flagging a review), and no contact meant to pressure the reviewer. The FTC rule also bars suppressing reviews through, among other things, knowingly false accusations against reviewers, so keep accusations out of public replies.

What each platform detects on its own and how to work with it

Large platforms run automated and human checks before and after a review appears, and your report is judged against what those checks already found.

The CMA's guidance lists signals that a platform might use to detect banned reviews: shared IP addresses or emails, networks of reviewers, and spikes in extreme reviews. It also says that publishers must take reasonable and proportionate steps to prevent and remove banned reviews, and must have a process to investigate suspicious ones. Those signals are only visible to the platform. Yours are your transaction records and your knowledge of your customers.

Tripadvisor's 2023 transparency report release gives a sense of scale. It says 4.4% of the 30.2 million reviews posted in 2022 were fake or fraudulent, that 72% of those were stopped before publication, and that about 2.3 million reviews were manually investigated by a moderator. It also reports ranking penalties on over 33,000 businesses for fraud. Platforms act against businesses that buy or write reviews, so your own profile has to be in order before you accuse someone else.

Platform	Grounds you can cite	What happens while it is reviewed	If the answer is no
Trustpilot	Harmful or illegal, personal information, promotional, not a genuine experience, about a different business	Review stays visible, except harmful or illegal ones, which are hidden or blurred during the check	Reply to the decision email with reasons, then use the dispute link in the footer of the final email
Tripadvisor	Its fraud rules; the guidelines page could not be read for this guide	Moderators investigate; the company reports penalties on businesses that commit fraud	Owner support pages could not be read for this guide, check them before filing
Yelp	Its content guidelines; its recommendation software decides which reviews count	Yelp's own pages blocked our reader, so the reporting steps could not be confirmed here	Read Yelp's business help pages before filing and keep expectations modest
Google Maps	Fake engagement, conflict of interest, multiple accounts, incentives	Not covered in this guide, the Google procedure is separate	See the Google-specific procedure

The Trustpilot column comes from Trustpilot's guidelines for businesses, which also say that "misuse of the flagging tool is not tolerated" and that this includes flagging reviews too quickly or for the wrong reasons. The same page says Trustpilot may reach out to the reviewer to ask them to consider updating their review or provide documentation. A clear, checkable report gives a real customer a fair chance to prove they are real.

How to write the dispute so a reviewer can act on it

A good dispute is short, names the rule, states the mismatch and points to attachments. It is written for a stranger who has about a minute and no reason to take your side.

Use this order, whatever the platform's form looks like.

Rule. Name the ground in the platform's own words, such as "not based on a genuine experience" or "about a different business".
Fact. State the mismatch in one or two sentences. "The review describes a delivery in March. We have no order under this name, email or address in March or any other month."
Proof. List attachments: the order export, the screenshot, the pattern table.
Ask. One line: "We ask that this review be removed under the policy above."

Leave out adjectives, motive and anything about the reviewer's character. If the form has a character limit, put the rule and the fact first.

An invented scenario: a burst of one-star reviews

The following is an invented scenario with generic parties, to show the order of work. Assumptions: a regional home-services company with about 400 reviews, a long-standing 4.6 average, and a customer system that logs every job with a name, phone number and postcode.

Eleven one-star reviews appear on the company's profile in 36 hours. The owner exports the jobs for the previous 12 months and searches for each reviewer's display name, any contact details the reviews mention and the services described. Seven reviewers match no job. Three describe a service the company does not offer. One matches a real job with a real complaint.

The owner sorts them. The three who describe a service the company does not offer go in a "different business" report, which is the quickest to verify. The seven with no matching job go in a "not a genuine experience" report, with the export and a pattern table showing all seven profiles are new and have one review each. The real complaint is left alone and answered in public. The owner does not report it, because it does not break a rule.

What to do when the platform refuses

A refusal means the platform could not confirm your claim from what it saw, not that you are wrong. Treat it as a request for a better file, and take the steps in order.

Use the appeal the platform offers. On Trustpilot that means replying to the decision email and then the formal dispute link. Add something new: a record you did not include, or a clearer match to the rule. An appeal with the same content gets the same answer.
Reply to the review in public. A short, factual reply tells the next reader that you have no record of this reviewer and invites them to get in touch. It does not accuse. Readers are judging you, not the reviewer.
Report a pattern to the platform's integrity or trust team where it offers a route. Trustpilot names its Content Integrity team in its guidelines for businesses. A pattern of related accounts is a different problem from one review.
Escalate outside the platform if the facts justify it. That means a consumer protection regulator for fake reviews as a trading practice, or a lawyer for a false statement of fact. These routes are slow and are not removal buttons.

The legal routes: what the FTC rule, UK law and defamation can and cannot do

Law adds three routes beside the platform process: a US regulator, a UK regulator and a defamation claim in the courts. None is a faster way to delete a review. This is information, not advice; your own lawyer decides which applies.

The US: the FTC rule on reviews

The Federal Trade Commission's rule on consumer reviews and testimonials, 16 CFR Part 465, took effect on October 21, 2024. According to the FTC's staff Q&A on the rule, it prohibits writing, creating or selling fake or false reviews and testimonials, and a business can be liable if it knew or should have known that reviews it bought were fake. It bars paying for reviews that express a particular sentiment, requires insiders to disclose their relationship clearly, bars company-controlled sites that pose as independent, bars certain suppression tactics and bars buying or selling fake social media indicators.

Three limits apply to you as the business being attacked. The Q&A says the rule has no private right of action, so you cannot sue under it; the FTC enforces it. Ordinary consumers are not liable for what they say in reviews. And the Q&A describes itself as staff guidance that is not definitive or comprehensive. The rule matters because it treats fake reviews as a trading problem and reaches the people who sell them: the Q&A names review brokers, advertising agencies, PR firms and reputation management companies.

Section 230 and suing a platform

Section 230 of the Communications Decency Act is why suing a platform over a user's review is hard in the US. Its core provision says a provider of an interactive computer service is not treated as the publisher or speaker of information provided by another. The section on effect on other laws keeps federal criminal law and intellectual property law outside that protection, among other exceptions, in 47 U.S.C. 230 on Cornell's Legal Information Institute site. In practice the platform's own policy process is the main remedy against it, and a claim generally lies against the person who wrote the review.

The UK: DMCC Act and CMA guidance

In the UK the Digital Markets, Competition and Consumers Act 2024 lists fake consumer reviews as banned in Schedule 20, paragraph 13, which also requires publishers of reviews to take reasonable and proportionate steps to prevent and remove them. According to the commencement regulations on legislation.gov.uk (SI 2025/272), Chapter 1 of Part 4 of the Act, except sections 232, 234 and 235, came into force on 6 April 2025. The CMA guidance, published on 4 April 2025, explains what is prohibited for businesses that publish reviews and says the same about reasonable and proportionate steps. You can cite that in a report to a UK-facing platform: the guidance expects it to investigate suspicious reviews and to correct ratings and counts when it removes banned ones. It does not turn your report into a legal order. The guidance we read does not set out fines, and we did not read the Act's enforcement provisions for this guide, so check them before relying on any figure.

Defamation, briefly

A false statement of fact that damages your business may be defamatory, but opinion, ratings and a harsh account of a real experience usually are not, and tests and time limits differ by country. Proving who wrote a fake review usually means a court process to identify an anonymous author: a lawyer's route, costly, and it can make the review more visible. Try the platform route first.

What fails and what backfires

Several tactics feel proportionate and cost more than the review does.

Writing fake reviews to push the rating back up. It breaks the rules of every platform named here, and the FTC rule bars writing or buying fake reviews. It also gives any platform reviewing your profile a reason to doubt your report.
Offering a refund, discount or gift to change or remove a review. Google's policy bans incentives for a revision or for removal of a negative review, and Trustpilot's guidelines ban incentivized reviews. It also tells the reviewer that pressure works.
Threatening the reviewer with legal action you do not intend to bring. The FTC rule bars suppressing reviews through unfounded legal threats and intimidation, and the platform may read the threat as the stronger violation.
Flagging every critical review. Trustpilot calls flagging for the wrong reasons misuse, and it weakens the one case that matters.
Paying a service that promises removal. We found no "special access" route in the platform pages read for this guide, and a service that posts or buys reviews for you creates risk for you under the FTC rule and UK law.
Shaming the reviewer in public. It draws attention to the review and can produce the second article you did not want.

A realistic timeline for a cross-platform dispute

Plan in weeks: a week of preparation, a wait for the first decision, one appeal, then an honest choice about the legal route.

A typical sequence for disputing suspected fake reviews

Day 1: captureScreenshot every suspect review and profile with dates and links. Do not reply yet, and do not contact anyone.
Days 1 to 3: matchSearch your records for each reviewer. Sort reviews into never a customer, wrong business, conflict of interest, pattern or genuine complaint.
Days 3 to 5: fileSend one report per rule category on each platform, with the evidence file. Reply in public to any genuine complaint.
Weeks 1 to 2: waitIn our experience decisions usually arrive in days to a couple of weeks. Reviews flagged for other reasons may remain visible during the check.
Weeks 2 to 4: appealIf refused, appeal once with something new. Report the pattern to the platform's integrity team where it has one.
Month 2 onward: decide on a legal routeWeigh a regulator complaint or a lawyer's letter against cost and publicity. Keep collecting genuine reviews in the meantime.

These are ranges from experience, not platform commitments, and a large burst or an unclear policy match can stretch each stage.

Keeping your own profile in good standing before a dispute

A platform deciding between your word and a reviewer's also looks at your profile, and a business already warned for its own review practices has a harder case.

Ask every customer for a review in the same neutral way and do not filter by mood. Do not reward reviews. Keep a log of review requests, which helps when you must show how a genuine review came about. Monitor new reviews so a burst is caught in hours, not weeks. Reputation Partners watches reviews, search results and AI answers together through monitoring, and a weekly look at your own profiles gets you part of the way if you work alone.

If this is happening to you now

Capture every review and profile today, match each one against your records, and do not reply, offer anything or contact the reviewer until you know which rule each review breaks. If the pattern is large or you are unsure, we can read it with you in a private audit.

Start with a private audit.

Questions people ask

Can I remove a fake review without proof?

Rarely. A platform has its own signals and may act on them, but a report that offers only an opinion gives it nothing to check. If you have no proof, report the facts you do have, such as no matching record, and expect the platform to decide on its own signals. A public reply is the fallback.

Will a platform tell me who wrote the review?

Usually not. Platforms generally keep reviewer identity private and share it only through a legal process. If you need to identify an anonymous author for a defamation claim, that is a matter for your lawyer and a court, not a request to the platform's support team.

Can I report a fake positive review on my competitor's page?

You can report any review you believe is fake, and a platform may investigate it. Report only what you can show, and never post or buy reviews against a rival, because that breaks the same rules and can end with penalties on your own listing instead.

Should I reply publicly to a review I think is fake?

Yes, briefly, unless the platform's process asks you to wait. Say you have no record of the person or the service, invite them to get in touch and say nothing about motive. A reply helps the next reader decide for themselves and shows the platform you tried to resolve it fairly.