The mechanical steps for Google, Yelp, Facebook, and Trustpilot, what each platform will and will not act on, how to escalate when the first attempt fails, and what to do when the answer is a permanent no.
Published 17 July 2026. General information about platform policies, not legal advice.
THE QUICK ANSWER
Flag the review from your own Business Profile, choose the one policy category it actually breaks, and appeal once in Google's review management tool if the first decision goes against you. If the review is genuinely from a customer, no platform will remove it, and a short public response is the real remedy.
Why most removal attempts fail
Business owners flag reviews because they are false. Platforms remove reviews because they violate a policy. Those two things overlap far less than anyone expects, and the gap between them is where nearly every failed removal lives.
A moderator cannot verify what happened in your shop. They were not there, you have every incentive to say the review is fabricated, and the reviewer has every incentive to say it is not. So the platforms built their rules around things that can be checked from the outside: account behaviour, duplicated text, conflicts of interest, personal information, off-topic content, threats. Truth is not on that list because truth is not observable from a moderation queue.
This has a practical consequence that changes your whole approach. Stop asking “is this review false?” and start asking “which written rule does this review break?” If you cannot answer the second question in one sentence, the flag will fail, and you should spend that hour on a response instead.
Checkable from outside
Account age, review patterns across cities, identical text on several listings, an employee's name and phone number in the body, a competitor posting about their own category. These get removed.
Only checkable from inside
“We never charged that.” “Nobody said that to her.” “He was never here.” All may be true. None can be verified by a moderator, so none of them, on their own, get a review removed.
Build the evidence pack first
Do this before you flag anything, and do it the same day. Reviewer profiles get deleted, review text gets edited, and a screenshot taken after the fact is worth much less than one taken while the evidence was live.
The direct URL of the review and a screenshot that includes the date and the reviewer name as displayed
The reviewer's profile page, screenshotted before it disappears, showing account age and their other reviews
Your transaction, booking, or ticket records for the period the review describes, showing no matching customer
Staffing rosters or schedules if the review names an employee who was not working
Any other listing where the same account posted near-identical text, which is the strongest single signal of coordinated fake reviews
One sentence naming the exact policy the review breaks. Not a paragraph about how unfair it is
One warning about the last item. Searching for the same text on other listings is how coordinated campaigns get caught, and it is the evidence platforms respond to most reliably. Searching for the reviewer as a person is not. Do not try to identify who wrote a pseudonymous review, do not contact people you suspect, and do not post anything about a reviewer's identity. That turns a moderation problem into a legal one, and it will be your legal problem rather than theirs.
The process, platform by platform
These four run on genuinely different rules. Read the one you need rather than applying Google's logic everywhere.
Google Business Profile
Removal bar
Violates a specific Google review policy, not simply untrue
Step by step
1
Sign in to the Google account that owns the Business Profile. Search your business name while signed in, or open the profile from business.google.com.
2
Open Reviews, find the review, and use the three-dot menu on that review. The option reads Report review or Flag as inappropriate depending on the surface you are in.
3
Pick one policy category, the one that fits best. Spam and fake engagement, off-topic, conflict of interest, personal information, and harassment are separate categories with separate bars. A wrong category is the single most common reason a valid report dies.
4
Submit once, from the owner account. Asking eight staff members to flag the same review adds no weight and can read as coordinated manipulation.
5
Wait. Automated screening happens fast, human review takes longer, and there is no notification when a decision is made. Check the review itself.
6
Open Google's review management tool to see the status of everything you reported. Reports sit in Pending, Approved, or Not approved.
7
If the decision was Not approved, use the appeal option in that same tool. The appeal is read by a different reviewer, so restate the specific policy line rather than repeating that the review is unfair.
8
If the appeal fails and the content is unlawful rather than merely against policy (defamation, a court order, disclosure of private data), that is a separate legal request path through Google's legal removal request form, not the flagging queue.
Usually qualifies for removal
Posted from an account with no plausible connection to your business, often with a pattern of reviews across unrelated cities
Content about a different business, a different location, or nothing to do with a customer experience
Left by a competitor, a current or former employee, or anyone with a stake in your rating (conflict of interest)
Names an individual employee alongside personal details: phone number, home address, immigration status, medical information
Profanity, slurs, threats, or sexual content aimed at staff
Incentivized or purchased content, in either direction
The same text posted repeatedly, or duplicated across several of your locations
Will not be removed
A real customer describing a real visit inaccurately or unfairly
A one star with no text. Google allows ratings without a review
A complaint about your prices, your policies, or your refusal to do something
A review from someone who never bought anything but did interact with you: a phone call, a quote, being turned away at the door
Anything you can only rebut with your own account of events. That is a factual dispute, and platforms do not adjudicate those
A review you believe is retaliation, unless you can point to the policy it breaks rather than the motive behind it
Google removes content that violates a written policy. It does not remove content because the business says it is false. That single distinction explains almost every failed removal attempt. Before you flag anything, find the sentence in Google's prohibited and restricted content policy that the review breaks, and write your report as if the person reading it has ten seconds and no context about your business.
Yelp
Removal bar
Violates Yelp's Content Guidelines, judged by a human moderator
Step by step
1
Log in to your Yelp for Business account and open the review from your dashboard, or open it on the public page.
2
Use the three-dot menu on the review and choose Report review.
3
Select the guideline it breaks. The options that carry weight are conflict of interest, lack of firsthand experience, privacy violation, and threats or harassment.
4
Add the supporting detail Yelp asks for in the free-text box. Be factual and short: what the reviewer claims, what your records show, and which guideline that maps to.
5
Yelp moderators read these individually. A decision typically arrives by email within a few days to a couple of weeks.
6
If it is declined, you can report again only if you have new evidence. Re-reporting the same case with the same facts will not change the outcome.
7
Regardless of outcome, use both response channels: the private message to the reviewer and the public comment. On Yelp these are separate tools and they do different jobs.
Usually qualifies for removal
The reviewer was not a customer and had no firsthand interaction of any kind
Written by a competitor, an employee, or a family member of either
Repeats something the writer heard from someone else rather than experienced
Contains someone's personal information, or a photo of a person who did not consent
Threats, harassment, hate speech, or lewd content
Promotional content or a solicitation dressed up as a review
Will not be removed
A negative but firsthand account, however harshly worded
A review of an interaction that did not end in a purchase
A review whose facts you dispute, where both sides are simply asserting
A review you believe was triggered by a competitor, absent evidence of who wrote it
Yelp is the strictest of the four about not removing negative content, and it is unusually direct about the fact that businesses cannot pay to have reviews taken down. It is worth understanding Yelp's recommendation software separately: reviews it does not recommend get moved behind a link at the bottom of your page and stop counting toward your rating. That is not removal, and you cannot request it, but it means some fake reviews quietly stop mattering without any action from you. Check whether the review is actually recommended before you spend a week fighting it.
Facebook Recommendations
Removal bar
Violates Community Standards, a narrower bar than the review platforms
Step by step
1
Open your Page and go to the Reviews or Recommendations tab.
2
Use the three-dot menu on the recommendation and choose the report option, then select the Community Standards category it breaks: spam, harassment, hate speech, or false information as defined by those standards.
3
Report the account separately if it looks fabricated. A recommendation left by a fake profile is often removed as a byproduct of the profile being actioned, which is a faster route than arguing the content itself.
4
Check Support Inbox on your Page for the decision. Facebook reports appear there rather than by email.
5
If a coordinated wave hits you in a short window, report the pattern rather than each item alone. Note the timestamps and the shared characteristics in each report.
6
As a last resort you can switch Recommendations off entirely in Page settings. Understand the trade: this hides every recommendation you have, including years of good ones, and it is visible to anyone who was reading them yesterday.
Usually qualifies for removal
Spam, including repeated identical posts across many Pages
Harassment, hate speech, or threats against you or your staff
Content from an account that violates Facebook's authenticity rules
Content that has nothing to do with your business at all
Will not be removed
An unfavourable recommendation from a real account describing a real visit
A dispute over what happened, where neither side breaks a Community Standard
A recommendation from someone who never transacted with you. Facebook has no purchase-verification concept here
Facebook applies Community Standards, which were written for the whole platform rather than for commerce, so ordinary business disputes rarely clear the bar. The practical lever is the account, not the post. If the profile is days old, has no photo, and has recommended six businesses in one afternoon, report the profile. The other practical lever is that Facebook recommendations carry less weight with searchers than Google does, so ration your energy accordingly.
Trustpilot
Removal bar
Violates Trustpilot guidelines, and the reviewer can be asked to prove the experience
Step by step
1
Log in to your Trustpilot Business account and find the review in the Reviews section.
2
Use the flag option on the review and choose the reason: harmful or illegal, advertising, not based on a genuine experience, concerns a different business, or contains personal information.
3
Write the explanation carefully. Trustpilot reads these, and the free text is where you state that your records show no transaction, no booking, and no support contact matching this reviewer.
4
Attach or reference your evidence: order records for the period, a support ticket export, or whatever shows the described interaction never occurred.
5
The review is placed under a flagged notice while Trustpilot assesses it. This is visible on your profile, so respond publicly as well rather than waiting in silence.
6
Trustpilot may ask the reviewer to provide documentation of the experience. If the reviewer does not supply it within the window Trustpilot gives them, the review comes down.
7
If the flag is rejected, you can reply publicly and that reply sits directly beneath the review permanently.
Usually qualifies for removal
No genuine buying or service experience behind the review, which Trustpilot can test by asking the reviewer for documentation
The review is about a different company with a similar name
Advertising, referral links, or promotional content
Personal information about staff or other customers
Harmful, discriminatory, or unlawful content
Will not be removed
A genuine customer's harsh account of a genuine transaction
A review about an experience you would rather was not public but which did happen
A review you dispute on the details while agreeing the person was a customer
Trustpilot has the most useful mechanism of the four, because it can put the burden of proof on the reviewer rather than on you. That only works if your flag says clearly that no transaction exists, and only if your own records can back it up. Vague flags that say the review is unfair get treated as what they are.
The Google escalation path, in order
Google is the one that matters most, because it feeds the map pack, and it is the one with the deepest escalation ladder. Work through it in this order and stop as soon as you get the outcome. Skipping to the top rarely works, because the higher rungs will ask what happened at the lower ones.
1
Flag the review, once, from the owner account, with the correct policy category.
2
Check the review management tool for status rather than waiting for an email. There will not be one.
3
Appeal the Not approved decision inside that tool. Restate the policy line, add the evidence pack, and drop the narrative about how much this review has hurt you. It does not help.
4
Contact Google Business Profile support through the help flow attached to your profile. You get a human, a case number, and a chance to explain a pattern that an automated pass would miss.
5
Post the case in the Google Business Profile Community forum with the review link and the policy you are citing. Product Experts read it and can occasionally escalate what support declined.
6
For unlawful content, use the legal removal request form. This is a different queue with a different standard, and it is where genuinely defamatory content belongs. Expect to describe the specific statement and why it is unlawful in your jurisdiction.
Two things to avoid at every rung. Do not buy removal services that promise guaranteed takedowns: the methods generally violate platform terms and the profile that gets penalised is yours. And do not offer a refund in exchange for deleting a review in writing, which creates a record that reads badly to regulators and to the next reader who sees it quoted.
When you are hit by a wave rather than one review
A coordinated burst is a different problem from a single fake, and it needs a different report. Six one stars in ninety minutes from accounts with no history is a pattern, and pattern evidence is the kind platforms act on fastest. Report each item individually, because that is the only mechanism available, but describe the pattern in every single report: the time window, the account characteristics, the shared phrasing.
Then check whether the wave is actually aimed at you. Waves often follow a viral post, a local news story, or a mistaken identity with a similarly named business in another state, and if it is mistaken identity, saying so publicly and briefly resolves it faster than any flag. Whatever the cause, do not paste the same reply under all of them. Identical replies twenty times over look automated, read as defensive, and on some platforms risk being treated as spam from your own account.
When removal is refused, the response is the remedy
Most fake reviews stay up. That is worth accepting early, because the energy spent on a fourth appeal is energy not spent on the thing that actually moves a purchase decision. Our judgement, and it is a judgement rather than a measured finding: people who go looking for the negative reviews are reading the owner's reply as evidence of how the business behaves when something goes wrong. On that reading, a composed reply under an unhinged review does more for you than the removal would have.
Respond within a day, publicly
An unanswered accusation reads as an admission. A calm, specific reply is the only part of this process you fully control, and it is written for the next reader rather than for the person who posted.
State the record, do not litigate it
Say what your records show in one sentence. Do not argue point by point, do not accuse anyone of lying, and never reveal anything that identifies a customer or a patient.
Offer a verifiable next step
Invite the person to contact a named role at your business with a booking or order reference. If no such reference exists, the absence answers the question for every reader without you saying a word about it.
Bury it with velocity
A single fake one star matters enormously on a profile with eleven reviews and barely at all on a profile with four hundred. Systematic asking is the only durable defence against content nobody will remove.
Keep the file open
Save the evidence pack. If the same account or the same pattern reappears, a documented history of coordinated activity is far more persuasive to a platform than a single report ever is.
WORKED EXAMPLE · ILLUSTRATIVE, NOT A REAL REVIEW OR A REAL BUSINESS
★☆☆☆☆ · Wendell O. (invented for this example)
“Worst experience anywhere. They took my deposit in March, never did the work, and the owner swore at me on the phone. Total scam, avoid this place.”
Flagged under spam and fake engagement, which is the category the evidence actually supports: the account is four days old and has posted near-identical text about two other shops in the same trade. Not approved. Appealed with that same policy line and the screenshots attached. Not approved again. Correct category, real pattern evidence, and Google still did not act.
The public reply that works anyway
“We take this seriously enough that we checked every deposit taken in March against our booking system, and we have no record of a job, a payment, or a call under this name. That may mean we have the wrong name, and if so we want to fix it: please email the office with your invoice or deposit reference and ask for our service manager. We answer the phone ourselves and we do not speak to customers that way.”
Note what the reply does not do. It never calls the writer a liar, never uses the word fake, never mentions the failed flags, and never gets sarcastic. It states a checked record, leaves an honest door open, and answers the swearing allegation once, flatly. Every reader draws their own conclusion, which is the conclusion you wanted.
Where the legal option genuinely fits
Occasionally a review is not just fake but defamatory: a false statement of fact, published, causing real damage. Platforms in the United States are generally not liable for what users post, so the target of any legal action is the author, not Google. Some businesses obtain a court order against the author and then submit that order through the platform's legal removal path.
Two cautions before anyone reaches for a lawyer. First, this is slow and it is expensive, and it usually costs more than the review does. Second, suing a reviewer is itself a news story in a way that the original review was not, and many jurisdictions have anti-SLAPP statutes that can leave you paying the other side's costs. Where the review is genuinely defamatory and genuinely damaging, talk to a lawyer in your jurisdiction. Where it is merely infuriating, write the reply and move on. If you run a practice where every reply carries professional-conduct risk, the law firm guidance covers responding without confirming that anyone was ever a client, and the dental version covers the same problem under HIPAA.
This guide is general information about how platform policies work. It is not legal advice and no lawyer wrote it.
The structural fix nobody wants to hear
Removal is defence, and defence does not compound. The businesses that shrug off fake reviews are the ones with enough real ones that a single fabricated one star moves the average by a rounding error. That is not a mindset, it is arithmetic: one bad review against fourteen is a visible dent, one bad review against three hundred is invisible. So the durable answer is a systematic ask after every good job, which is one tap if you hand people a direct Google review link and QR code rather than hoping they find your profile.
Draft the reply while you are still annoyed
The hardest part of answering a fake review is writing something calm on the day it lands. Paste the review into the free response generator, pick a tone, and edit what comes back. Five free per day, no signup for the first one.
Praisly watches Google, Yelp, Trustpilot and Facebook, drafts responses, scores a dispute before you submit it, and asks happy customers at the right moment. Plans are published on the pricing page starting at $59 a month, month to month, cancel in two clicks, no sales call.