Google Business Reviews disappear in both directions. A fake five-star rating a competitor bought vanishes, which is the system working. A real one-star account from a genuine customer vanishes too, which is the system misfiring, and the business owner is rarely told either happened. Google now removes reviews at a scale that touches almost every active profile. Most owners notice only when a number drops. Understanding why it happens, and what actually moves a review on or off a profile, is the difference between managing reviews and reacting to them.
Three forces take a review down
A Google review comes off a profile through one of three channels, and they do not work the same way. The first is automated moderation. Google's systems evaluate reviews as they post and, increasingly, long after, scoring language, reviewer behavior, and account signals for patterns that appear to be spam or manipulation. The second is manual enforcement, where a flagged review is routed to a human who checks it against policy. The third is legal removal, where a court order or a valid legal complaint forces a takedown, a channel that matters far more in some countries than others.
Most removals now run through the first channel. That is the one no one can see working, and the one that makes review counts move without warning.
Google removes reviews at a scale most owners never see
The volume is not a secret, because Google publicizes the part it wants seen. Google reported blocking or removing more than 240 million policy-violating reviews across 2024, along with 70 million fake edits and 12 million fake Business Profiles, and credited AI for the reach. The year before, a new large language model pushed fake-review takedowns up 45% to over 170 million. The running totals sit in a public Maps transparency report Google updates on its own schedule.
Read what those numbers count. Every one is a takedown of fake or policy-violating content, the kind of review a business wants gone. They are the enforcement Google is proud of. They are not the whole story, because the same automated net that catches the fakes also catches legitimate reviews, and that figure is one Google does not publish.
The rules that actually get a review removed
Google removes a review when it breaks a policy, not when a business dislikes it. The distinction is the entire game. The grounds that matter most for reputation work are the ones tied to authenticity. Fake engagement refers to any review not based on a real experience with the business, which is the single most reliable grounds for removal. Rating manipulation covers incentivized or paid reviews. Impersonation, misinformation, and misrepresentation cover accounts that are not what they claim to be. Off-topic covers political rants and commentary that has nothing to do with the visit. Restricted content, harassment, hate speech, personal information, and explicit content each carry their own removal path.
Now the catch. The complaint most owners want to file is not on that list. A review that is harsh, unfair, or simply wrong about the facts is not removable on those grounds alone. Google states it plainly: do not report a review just because you disagree with it. A one-star account of a really bad day stays up. A five-star review from someone who was paid comes down. The policy rewards the authenticity of the experience, not the sentiment of the rating.
How the removal process really works
There are two official routes, and both demand a policy reason. The first is flagging the review directly from the Business Profile, using the report icon and selecting a violation. The second is the Reviews Management Tool, where a business confirms the account, selects the profile, and chooses to report a review for removal. Evaluation typically takes several days, and the tool shows the status as pending, reviewed with no violation found, or escalated. If a request is denied, Google allows one appeal per review.
There is a third way a review gets flagged, and it runs blind. Anyone with a Google account can report a review from its menu, not just the business, which means customers, competitors, and a reviewer's own detractors can all push a review into the moderation queue. That path has no feedback loop. The person who flags it sees no status, gets no reason, and is never told whether the review came down because of their report, someone else's, an automated sweep, or not at all. Even the business, watching the same review, cannot tell which flag moved it. The Reviews Management Tool at least shows a business the status of its own request. Every other flag disappears into the same silent pipeline, which is why crediting any single removal to any single action is mostly guesswork.
The failure modes are as consistent as the process. Flagging a review on the wrong ground gets it reviewed and left up, the shutout the Stryde agency learned eight times on its own profile. Resubmitting the same request does not restart the clock or improve the odds; it signals manipulation and can draw scrutiny to the account. The move that works is narrow: identify the actual policy the review breaks, usually the absence of a genuine experience, file once on that ground, document it, and use the single appeal if the first pass fails.
The deletions you did not ask for
The harder problem is the reviews that leave without anyone flagging them. An analysis of 60,000 Google Business Profiles by GMBapi.com, published on Search Engine Land, found deletions climbing sharply across the first half of 2025, with a growing share of monitored locations losing at least one review in a given week. The patterns underneath are the useful part. In English-speaking markets, deleted reviews skew heavily toward five-star ratings, consistent with automated systems hunting incentivized and fake positive feedback. Medical and home services see reviews removed early, within months of posting, while restaurants and retail see reviews pulled years later as detection improves and old profiles get re-scanned. Germany operates the opposite way, with more low-rated reviews removed and faster takedowns, reflecting its strict defamation law rather than its spam filters.
Two harms arise from this for a business. Trust erodes when legitimate reviews, good or bad, disappear without explanation. And the data distorts, because deleted reviews move the star average and the signals that feed local ranking. A profile can lose standing without a single new negative being posted. This is a vendor dataset rather than a Google disclosure, so treat the specific shares as indicative, not official. The direction is not in dispute.
What Google will not tell you
Here is the asymmetry that governs it all. Google publishes the takedowns and withholds the collateral. It reports how many fake reviews it caught, but never how many real ones it swept up with them. It does not publish how long a valid removal request takes to clear, or how those success rates shift during a broader enforcement push. The one number that would tell an owner what is happening to their own reviews this month is the one number Google keeps to itself. That is why review monitoring has become its own discipline, because the record a business keeps is the only record it controls.
What to do about it
The defensive posture is steady and unglamorous. Generate reviews the compliant way, at a natural pace, on the customer's own device, with no incentive attached. When done consistently, authentic reviews outnumber and outlast anything the filters question. Monitor the profile for removals as closely as for new reviews. Keep a dated record of what was posted and what vanished, because Google will not hand that history back. When a review genuinely violates policy, flag it once on the correct ground and appeal once if needed, not the same complaint on repeat. When an enforcement wave is running, and removals stall, wait it out. Pushing into a backed-up queue only flags the account. On Google, the business that keeps its own records and files clean beats the one that pushes hardest.
The lesson
Google removes reviews to protect the integrity of its ratings, and the same machinery that cleans up the fakes occasionally takes the real ones with it. The enforcement is largely automated, largely invisible, and reported only on the side that flatters it. A business cannot change how Google moderates. It can control the quality and pace of what it generates, the accuracy of what it flags, and the completeness of the record it keeps. Do those three well, and the removals that matter stop being a mystery and start being a managed risk.
Part of the Review Management pillar. Related: The FTC's Fake Review Rule and Review Velocity.