It's almost never a bug. It's a filter doing its job.
The first thing to let go of is the idea that something broke. Reviews disappear because a system decided they looked risky, and that system is working exactly as designed, even when it's wrong about your review.
Google's spam filter scans every review against patterns it associates with fakes: where it came from, who wrote it, how it showed up, and what it says. When something pattern-matches, the review gets pulled. The filter is tuned to be aggressive, because Google has decided that letting a fake review stand is worse than catching a real one in the net.
So a genuine review can vanish simply because it shared a few traits with the fake ones. Annoying, but fixable once you know what the filter is twitchy about.
What actually trips the filter.
Most disappearances trace back to a short list. None of these are guaranteed to pull a review, but each one nudges the odds, and they stack.
- The review came from the same device, network, or IP as your business. The single most common own-goal.
- The reviewer has a thin profile: no photo, no review history, an account made yesterday.
- A burst of reviews landed in a tight window. Ten on a Tuesday after months of silence reads as a campaign.
- The text contains a link, a phone number, a price, or a URL. The filter treats those as promotional.
- The review names a competitor, or reads as off-topic ranting rather than a customer experience.
- The reviewer deleted their Google account, which quietly takes their review with it.
The tablet-by-the-counter trap.
Here's the mistake that bites good, honest businesses the hardest. You keep a tablet at the counter, or you hand customers your phone, and you ask people to leave a review right there before they go. It feels efficient. It's filter bait.
Every one of those reviews comes from the same device on the same network. To Google, that's a pile of reviews from one location with one fingerprint, which is exactly what a fake-review operation looks like. The filter can wipe the whole batch, and it doesn't care that every customer was real.
Google's 2026 policy tightened this further, discouraging on-premises kiosks and pressured, ask-on-the-spot setups. The fix is boring and it works: let people review from their own phone, on their own time, in their own words.
- What the 2026 Google review policies changed
On-premises asking, incentives, and AI text, in plain language.
If you bought reviews, this is the bill.
Sometimes the disappearing reviews are the ones you paid for, and the sweep that takes them can take your real ones too. Bought and incentivized reviews are precisely what the filter exists to catch, and a profile that gets actioned for them often loses honest reviews in the same pass.
If a batch of suspiciously perfect reviews vanished at once, that's not Google malfunctioning. That's Google catching up.
- What buying Google reviews actually costs
The FTC rule, the Google penalty, and the cheaper fix.
Can you get a filtered review back? Usually not.
This is the part people don't want to hear. There's no real undo button. A filtered review isn't deleted from existence, but you can't manually restore it, and Google rarely reverses the call.
You have two long shots. You can report the missing review through Google Business Profile support and explain that it's legitimate, which occasionally works and usually doesn't. Or the customer can try posting it again from a stronger account, ideally one with some history and a profile photo. Neither is reliable.
So the honest strategy isn't recovery. It's prevention. Stop losing the next hundred reviews instead of fighting for the one that's gone.
The most filter-proof review looks unmistakably human.
Flip the filter's logic around. It's hunting for reviews that look manufactured, so the safest review is the one that looks like a real person had a real experience and said so in their own voice.
That means a few specific things. The review comes from the customer's own phone, not your counter tablet. It arrives spaced out across the week, not in a Tuesday avalanche. And it has texture: the tech's name, the actual job, the thing that surprised them. Generic praise in a burst looks like spam. A specific story from a real person does not.
This is where guided reviews quietly help. When a customer writes from their own device, in their own words, about the specific thing that stood out, you're collecting exactly the kind of review the filter is least likely to touch. You're not gaming the system. You're producing the real thing it's trying to protect.
Next step
Collect reviews that look like what they are.
Real customers, their own phones, their own words, spaced out and specific. That's the profile the filter leaves alone.