The short version: most fake reviews are not badly written, they are boringly perfect, five stars, posted in clusters, full of the product name repeated for no reason, and written by people with no other reviewing history. Check the pattern, not the words, and check the reviewer as hard as you check the review.
A client, 312 reviews, and nine days
A few years back I worked with a small home goods seller on Amazon, a woman I will call Sarah because her real name does not matter here. She sold weighted blankets, decent margin category, competitive as anything on that platform. Her listing had 4.9 stars from just over 300 reviews and I remember thinking, before I looked at the dates, that it was a nice healthy number.
Then I pulled the dates. 312 of her reviews had landed in a nine day window, eighteen months earlier. Before that window, the product had eleven reviews. After it, almost nothing for months, then a slow trickle of real ones that averaged 3.6 stars, not 4.9.
Sarah had not bought those reviews herself, she had inherited the listing from a previous supplier relationship and had no idea the history existed. But Amazon did not care about the backstory. Her account got flagged during a review-integrity sweep, the listing was suppressed for six weeks, and she lost roughly $14,000 in sales during her busiest quarter of the year while she appealed. That is the bit nobody tells you when they say “just check the reviews before you buy.” Sometimes the seller is as much a victim of the fake reviews as you are, and sometimes the fake reviews are the whole reason the business collapses.
The bit nobody likes saying out loud
Here is the uncomfortable part. It is not only dodgy dropshippers buying fake five-star reviews. Well-known brands do it too, through review-swap networks and paid “gifting” programs where the product is free in exchange for a review that is never disclosed as such. And the flip side gets ignored almost entirely: plenty of one-star reviews are fake as well, planted by competitors, or written by people trying to extort a small seller into a refund or a cash payment in exchange for taking the review down. I have seen this happen to a coaching client whose testimonials page got hit with three separate one-star “reviews” from accounts created the same day, all mentioning a service she does not even offer. Fake reviews are not always dishonestly positive. Sometimes they are dishonestly negative, and that is the version most articles on this topic skip completely.
The other awkward truth is that the “verified purchase” badge means far less than people assume. A seller can refund a buyer privately after they leave a glowing review, and the badge stays, because the platform only checked that a purchase happened, not that it stuck. Verified purchase tells you a transaction occurred. It does not tell you the person kept the product, used it, or was paid to say something nice about it.
Step by step: how I check a review before I trust it
This is the exact process I use now, whether I am buying software for my own business or checking testimonials before recommending someone to a client.
- Sort by “most recent” first, not “most helpful.” A helpful sort buries anything written in the last month. You want to see what is happening right now, not what happened when the listing launched.
- Look for date clustering. If forty reviews landed in the same week and the product has been live for two years, that is a red flag on its own. Real reviews trickle in unevenly because real people buy things at different times.
- Click the reviewer’s profile. A genuine reviewer usually has a history, other products, other categories, opinions that vary. A reviewer with five reviews, all five stars, all posted this month, for unrelated products, is almost certainly part of a review ring.
- Read the middle, not the extremes. Three and four star reviews are harder to fake convincingly because they require specific complaints. If a listing has almost nothing between one star and five star, be suspicious of both ends.
- Search for the exact phrase. Copy a distinctive sentence from a review and paste it into a search engine in quotation marks. If it shows up word for word on three other unrelated listings, you have found a template, not an opinion.
- Check the star distribution graph, not just the average. A genuine product usually has a rough bell curve or a long tail of complaints. A product with 90% five star and almost nothing in between is a pattern, not a coincidence.
- Look for the phrase “I received this product for free” or similar disclosure buried at the bottom. Under the FTC’s rule that finally took effect after years of delay, sellers and reviewers in the US are legally required to disclose paid or incentivized reviews. Most still do not bother, which tells you something about how seriously the rule is being enforced so far.
None of these steps take more than two minutes once you have done it a few times. The habit is the hard part, not the method.
What fake reviews sound like
Forget the old advice about “too many exclamation marks” or “sounds like a robot.” AI-written reviews now sound completely normal, sometimes more polished than a real customer would bother to be. What still gives them away is structure, not tone.
Real reviews are lopsided. Someone loves the colour but hates the zip. Someone says it arrived a day late but the customer service person was lovely about it. Fake reviews tend to be evenly positive across every attribute, mention the product name unnaturally often (because the reviewer was given a script or a template with the product name inserted), and rarely mention anything that would put a buyer off even slightly. If a review reads like ad copy, that is because in a lot of cases it started life as ad copy.
I wrote a full breakdown of this exact pattern, with more screenshots and examples, in my detailed guide on spotting fake reviews before you buy, if you want the longer version with visual examples.
Where AI has made this both easier and worse
AI cuts both ways here and I will not pretend otherwise. On one side, legitimate ecommerce brands are now using AI tools to collect, sort, and surface real customer feedback faster, spotting genuine patterns in complaints before they become a wave of one-star reviews. I have written before about how ecommerce brands can use AI for customer reviews in a way that improves the product rather than just polishing the appearance of it, and coaches and consultants can do something similar with client testimonials, which I covered separately in a piece on using AI for customer reviews as a coach or consultant.
On the other side, the same generative tools make it trivially easy to churn out hundreds of unique-sounding fake reviews in an afternoon, no template needed, no repeated phrases to search for. This is why the clustering and reviewer-history checks matter more now than the old trick of hunting for copy-paste sentences. The words got harder to catch. The pattern of who posted them and when did not.
The same red flags apply to reviews of services, not just products
This matters just as much when you are hiring someone, not buying something off a shelf. I get messages every month from people asking how to tell whether an AI consultant’s glowing testimonials are real, and the honest answer is that the checklist barely changes. Look for reviewers with a real presence somewhere else, a LinkedIn profile with actual history, a named business you can check exists. I wrote a full guide on how to tell a good AI consultant from a fake one, because the testimonial page is usually the first thing that gets faked when someone has more confidence than track record.
When I reviewed SERP Empire for my own audience, I bought it myself, used it for six weeks, and reported what happened to rankings, good and bad, rather than repeating the affiliate copy everyone else was recycling. You can see that approach in my SERP Empire review. That is the test I apply to any review I read elsewhere too: did this person clearly use the thing, or are they describing features from a sales page in slightly different words?
Three quick checks you can do in under a minute
If you do not have time for the full seven-step process, here is the shortcut version I use when I am buying something small and low stakes:
- Sort by lowest rating first and read ten of them. If the complaints are specific and varied, the listing is probably real. If there are barely any complaints at all, be suspicious.
- Check whether the reviews mention the product by its exact full name repeatedly. Real customers say “it” or “the blanket” or “this thing.” Paid reviewers are often instructed to include the product name for SEO reasons and it reads unnaturally.
- Look at the total number of reviews against how long the product has existed. A product launched four months ago with 900 reviews and a brand new seller account is not organic growth, that is a purchased base.
What to do if you think you have found a fake review
Report it on the platform, most have a specific “report” flow for reviews rather than just flagging the product. Do not rely on your own judgement alone if the purchase is expensive, cross check with an independent source, a forum thread, a YouTube teardown, or a friend who has owned the thing. And if you run a business yourself, audit your own listings the same way you would audit a competitor’s. Sarah’s story above happened because nobody checked the dates on her own product until it was almost too late.
Frequently asked questions
How can you tell if an Amazon review is fake?
Check the date it was posted against the product’s listing history, look at whether the reviewer has other reviews across unrelated categories, and check the star distribution graph for an unnatural spike at five stars with almost nothing in the middle. Reviews clustered within days of each other, from accounts with no other history, are the strongest single signal.
Do verified purchase badges mean a review is real?
Not reliably. A verified purchase badge only confirms a transaction happened on that platform, not that the reviewer kept the product, used it, or was unpaid for the review. Sellers can and do refund buyers privately after a five-star review is posted and the badge stays in place.
Are negative reviews ever fake too?
Yes, and it is rarely mentioned. Competitors sometimes post fake one-star reviews to damage a rival listing, and some sellers face extortion attempts where someone threatens a bad review unless they are paid to remove it. If a batch of one-star reviews appears suddenly, from new accounts, mentioning issues the product does not have, treat it with the same suspicion as suspiciously perfect five-star clusters.
Is buying fake reviews illegal?
In the US, the FTC’s rule against fake and deceptive reviews carries civil penalties of tens of thousands of dollars per violation for businesses that buy them, write them, or fail to disclose paid ones. Enforcement so far has been inconsistent, but the legal risk for sellers is real and growing, not theoretical.
Related reading: The Best Product Review Sites to Check Before You Buy in 2026 and How to Use AI in Everyday Life.
