Fake reviews vs. false reviews: two different problems, two different playbooks

Two one-star reviews land on the same auto glass shop’s Google profile in the same week. The first: “Terrible service, do not recommend this place to anyone.” Posted by a profile with no photo, created that month, whose only other reviews are five-star raves for businesses in three different countries. The second: “They replaced my windshield and now my rain sensor doesn’t work. They refused to fix it. $400 wasted.” Posted by an actual customer - job’s in the system, tech remembers the car.

Same star rating. Same platform. And almost nothing about the right response overlaps, because the first review is fake and the second one is false. The fake review has no customer behind it - it’s paid, botted, or planted. The false review has a real customer behind it making a claim that isn’t true (that sensor was disconnected before the car ever came in, and it’s documented on the intake sheet).

Most articles about “fake reviews” mash these together, which is how businesses end up making the two classic errors: treating a false review as fake, and treating a fake review as real. The thesis of this guide: diagnosis comes before response, because the playbooks don’t just differ - they actively conflict. The right move against a fake review (flag, escalate, never engage the substance) is the wrong move against a false one, and vice versa.

Why misdiagnosis is expensive in both directions

Call a false review fake, and you flag a real customer’s review to Google as policy-violating. It won’t be removed - the reviewer is a genuine customer, so there’s no conflict-of-interest or spam violation to find - and Google gives you exactly one appeal after a denied flag, which you’ve now spent on an unwinnable case. Worse, if your public response accuses the reviewer of being fake and they’re demonstrably real, you’ve published a false statement about an identifiable person. That’s the defamation table flipped around, with you on the wrong side of it.

Call a fake review real, and you do the opposite damage: you write an earnest, apologetic response (“we’re so sorry, please call our manager”) to a review purchased by a competitor or an extortion crew. Prospects now read a confession under an invented complaint. And if it’s an extortion wave, your apology signals that the pressure is working - the documented pattern through late 2025 was 10-20 fake one-stars in 24-72 hours followed by a payment demand, and paying or pleading reliably makes it worse.

So: diagnose first. Fifteen minutes of forensics before a word gets written.

The diagnostic: is there a customer behind this review?

That’s the whole question. Everything else is evidence for or against it.

Check your own records first. Name, initials, the described transaction, the date range. Most false reviews self-identify - the reviewer wants you to know who they are, because they want their specific grievance fixed. Most fake reviews can’t survive this check: no matching job, no matching appointment, a described service you don’t offer, a location detail that’s wrong (“the waiting room on the second floor” when you’re a ground-floor shop).

Then read the profile, not the review. Fake reviewer profiles tend to share a fingerprint: recently created, no photo or a stock-looking one, review history that’s geographically incoherent (a plumber in Tampa, a café in Manila, a dentist in Leeds, all in one week), or a burst of one-star reviews for several businesses in your category - that last one suggests a competitor’s purchased batch, which has its own playbook covered in what to do when a competitor leaves a fake review.

Then look at timing. One suspicious review is a judgment call. Four in 48 hours after two quiet months is a campaign. Real dissatisfaction arrives one customer at a time; paid dissatisfaction arrives in clusters, because it’s sold in batches. A New York Times - interviewed review seller admitted in 2025 to pricing negative Google reviews at $100 per batch of 20, which tells you the unit economics: nobody buys one.

And be honest about the third possibility. Some reviews that feel fake are real customers you don’t recognize - a spouse who posted under their own name, a customer from eight months ago, a delivery mishap you never heard about. “We have no record of you” is a diagnosis of your records, not of the review. Treat “fake” as a conclusion that requires evidence, not a feeling.

Playbook one: the fake review

The counterintuitive core: the reviewer doesn’t matter. There’s no one to persuade, no relationship to save, no misunderstanding to fix. Your audiences are exactly two - the platform, which can remove the review, and prospects, who will read it until the platform does.

Document before you flag. Screenshot the review, the profile, and the profile’s review history with timestamps. If it’s a wave, capture the cluster and the dates. Evidence disappears when profiles get deleted, and you may need it later for a legal letter or an FTC complaint.

Flag with the policy language, not with outrage. Google removes reviews that violate specific policies - spam and fake engagement, conflict of interest, off-topic content. A flag that says “this reviewer was never a customer; the profile posted 14 reviews in 3 countries this week” gives the reviewer-facing policy team something to verify. “This review is a lie” gives them nothing. Google reported removing 292 million policy-violating reviews in 2025 - the enforcement machinery exists, but it runs on specifics. If the wave comes with a payment demand, use Google’s dedicated extortion reporting form (launched November 2025), which has been notably faster than the standard flag queue.

Respond publicly once, briefly, for prospects only. Something like: “We take every review seriously, but we have no record of this visit and this profile’s activity suggests it isn’t from a customer. We’ve reported it. If we’re wrong and you did visit us, call us directly - we’ll make it right.” Notice what that does: it plants a calm flag for the 97% of review readers who also read responses, it leaves an exit ramp in case your diagnosis is wrong, and it never uses the word “fake” as an accusation against a named person. There’s a template for this exact situation if you want tested wording.

Know the regulatory backdrop. Buying and selling fake reviews is now a federal violation, not just a terms-of-service problem. The FTC’s Consumer Reviews Rule took effect in October 2024 with penalties up to $53,088 per violation, and enforcement started moving in December 2025 with a first round of warning letters. That matters to you in two ways: it gives your documentation somewhere to go (the FTC takes complaints about fake-review operations), and it should permanently close the “fight fire with fire” temptation. We covered the rule’s first year of enforcement separately.

Playbook two: the false review

Now the reviewer matters enormously, because they’re a real person with a real audience and a real grievance - even if the central claim is wrong. The instinct to treat them as an enemy is the most expensive instinct in this entire subject.

Here’s the worked example. Kessler’s Piano Movers in Milwaukee gets this review: “They dropped our baby grand coming down the stairs and refused to take responsibility. $9,000 piano, cracked soundboard. Avoid.” Real customer, real move. But the crew’s photos, taken at delivery per company procedure, show the soundboard crack was documented and photographed before the piano left the origin house - it was pre-existing damage the customer signed off on.

The response that torched a relationship and convinced no one: “This is completely FALSE. You signed the pre-move inspection acknowledging existing damage. We will be consulting our attorney.” Defensive, threatening, and - to a prospect who can’t see the inspection sheet - indistinguishable from a guilty company bluffing.

The response that worked: “We moved your piano on March 14, and I understand how upsetting the soundboard crack is - a baby grand is irreplaceable. Our pre-move inspection, which we photograph and both parties sign, documented that crack before the move began, and I’d be glad to send you those photos and the signed sheet. If you believe our crew caused new damage beyond what was documented, call me directly - I’m the owner and I’ll look at it personally.” The customer didn’t take the review down. But over the following quarter, Kessler’s booked 11 moves from customers who mentioned reading that exchange - several said the signed-inspection detail was what convinced them. The response didn’t win the argument. It won the audience.

The structure to steal from that: correct the factual record with checkable specifics, never call the reviewer a liar (say what your documentation shows, not what they are), offer evidence rather than characterizing it, and keep an offline door open with a named human. If you need help finding that register, there’s a starting-point template for responding to unfair reviews built around exactly this correct-without-attacking move.

On suing: false statements of fact from an identifiable person are the one category where defamation law can actually apply - a 2024 Ohio appeals court decision held that negative reviews containing provably false factual claims can be defamatory rather than protected opinion. But the practical math (five figures in fees, a year-plus of process, anti-SLAPP exposure, and the attention a lawsuit attracts) kills most cases before the merits matter. We’ve walked through when suing over a false review actually makes sense - the short version is: almost never, and never as a first move. The demand letter’s main effect is usually on the review you’ll get next.

The part nobody tells you: most hard cases are hybrids

The clean categories - pure fabrication versus honest customer with wrong facts - cover maybe 80% of cases. The hard 20% are hybrids, and they’re where businesses hurt themselves.

There’s the real customer who embellishes: the wait really was long, but “three hours” was 55 minutes, and “the manager screamed at us” was a tired shift lead being curt. Treat this as a false review, not a fake one - but correct only the checkable facts and concede the true core. Disputing the parts that are true destroys your credibility on the parts that aren’t.

There’s the fake review wrapped around true details - typically an ex-employee or a customer’s relative, someone with inside knowledge but no transaction. The insider detail makes it read authentic; the missing transaction makes it flaggable. Flag it on the conflict-of-interest policy, but respond publicly with the false-review tone, because prospects can’t see the missing transaction and will judge you as if the reviewer were real.

And there’s the one that stings: the review you’re certain is false that turns out to be true. The tech who did skip the safety check, the night shift that really did say that. Before every “false review” response, someone has to genuinely ask the staff involved - not “this didn’t happen, right?” but “walk me through that afternoon.” The most damaging response in this whole taxonomy is the confident public denial of something that happened. It converts one bad review into a credibility crisis, and screenshots are forever.

Diagnose, then respond. The review that’s lying about you deserves a case file. The customer who’s wrong about you deserves a conversation. Getting those backwards is the whole genre of self-inflicted reputation damage - and it’s entirely avoidable in the fifteen minutes before you start typing.