Published on 2026-07-25
Are some negative reviews actually sabotage?
A one-star review complaining about a dish the restaurant has never served isn't a bad customer experience. It's a sign the review might not be real at all.
A one-star review shows up on a small bakery's page. The reviewer complains about slow table service and a rude waiter. The bakery has three seats and a counter. Nobody who ever ordered a loaf of bread there waited on a waiter, because there isn't one. The owner reads it twice, confused, then a little sick to her stomach, because the review is public, permanent, and dragging her rating down for a meal that never happened.
We usually assume a bad review means a bad experience. Most of the time, that's exactly what it means. But not always, and it's worth knowing the other version exists. The US Federal Trade Commission says it plainly in its own consumer guidance on reading reviews: 'fake reviews are not always positive. Sometimes, a company might post fake negative reviews to harm a competitor.' The same playbook that inflates one business can be pointed the other way, at whoever it's competing against.
An arms race, not a one-off grudge
Harvard Business School researchers Michael Luca and Georgios Zervas studied this directly using Yelp data, and one of them described what they found as an 'arms race': some businesses hire people to post glowing reviews of themselves, while others lean on family and friends to do the same, and to leave unfavorable reviews aimed at a rival down the street. The pattern wasn't random. Small, independent restaurants with weak existing ratings were the most likely targets and the most likely offenders, because for a business that size, a single star genuinely moves revenue by an estimated five to nine percent. That's not a rounding error. For a struggling shop, that's the difference between making rent and not.
It's also, structurally, a hard thing for any platform to police. A review posted by an actual account, using ordinary words, aimed at a real business, looks like every other review in the queue. There's no watermark on malice. Filtering systems can flag patterns (bursts, duplicate phrasing, brand new accounts) but a patient, one-off sabotage campaign, spread out and written carefully, can slip past exactly the same checks that catch clumsier fakes.
The details that don't quite fit
A fake negative review often gives itself away the same way a fake positive one does: through details that don't match the actual place. A complaint about a menu item the kitchen has never made. A one-star rating for 'terrible parking' at a business with no parking lot to speak of. A review of a haircut left on a page for a hardware store. Whoever wrote it may never have set foot inside, and the mismatch is the tell, more reliable than the tone or the star count alone.
- The complaint describes a product, service, or feature the business doesn't have
- The account has posted only this one review, ever
- The review appeared right around a competitor's launch, price drop, or a bad news cycle for the business
- The star rating is extreme but the written detail is vague or generic
- Several harshly negative reviews cluster within days of each other
None of this is a reason to wave away every bad review a business gets. Most negative reviews are exactly what they look like: someone genuinely unhappy, telling other people why. Dismissing every complaint as an attack from a rival would be its own kind of blindness, and just as unfair to the customers who had a real, disappointing experience worth knowing about. The point is narrower: when a single review contradicts basic, checkable facts about the place it claims to describe, that contradiction is worth more weight than the number of stars attached to it.
If you're the one reading reviews before booking a table or hiring a plumber, the habit that protects you is simple: read a handful, not just the loudest one. A business with hundreds of reviews and a single wildly out-of-place complaint is telling you something different than a business where the same specific problem shows up again and again. One is noise. The other is a pattern, and only a pattern is actually evidence.
A star rating tells you how someone felt. It never tells you whether they were ever actually there, or whether they had a stake in what you'd decide next.