Blog · Spotting fake reviews

How to Spot Fake Amazon Reviews in 2026

The five-star average lies more often than you think — here is how to read past it before you buy.

PageVigil · 8 July 2026

A product with 4.7 stars and thousands of reviews feels safe. But rating inflation is now an industry, and a high average tells you almost nothing on its own. The good news: fake reviews leave fingerprints, and once you know what to look for you can spot most of them in about a minute.

Read the middling reviews first

Skip the five-star gush and the one-star rants. Head straight to the three- and four-star reviews, because these are the hardest to fake and the most honest. Real buyers who mostly liked something will still mention the annoying bits — the strap that frayed, the app that needs an account, the sizing that runs small. If the three-star reviews describe a fundamentally different product than the glowing ones, trust the middle.

Watch for template phrasing

Paid and generated reviews tend to recycle the same scaffolding. Look for repeated openings like I was skeptical at first but, vague praise that never names a specific feature, and oddly formal product-name repetition ("This amazing kitchen scale is an amazing addition to my kitchen"). When several "different" reviewers phrase their enthusiasm almost identically, you are probably reading one writer, or one prompt.

Check the timing

Sort reviews by most recent and scan the dates. Genuine reviews trickle in steadily. A cluster of dozens of five-star reviews all posted within the same few days — especially right after launch — is a classic seeding pattern. The same goes for a wave of glowing reviews that suddenly appears after a run of critical ones, which often signals a reputation cleanup.

Mind the rating-text mismatch

One of the loudest signals is a five-star rating stapled to lukewarm text. "It's okay, does the job, a bit flimsy — 5 stars" is not how satisfied people behave. This mismatch often means the review was incentivized: the seller offered a refund, a gift card, or a free replacement in exchange for the top rating. Phrases like "received this in exchange for my honest review" or "the seller kindly refunded me" are your cue to discount the star count entirely.

Look at the reviewer, not just the review

Click through on a few enthusiastic reviewers. A profile that has posted twenty five-star reviews in one week across unrelated products — a phone case, a probiotic, a garden hose — is almost certainly part of a review farm. Verified Purchase badges help, but note that sellers can game these too by reimbursing buyers, so treat the badge as a weak signal rather than proof.

Let a tool do the counting

Doing all of this by hand is thorough but slow, and it is easy to miss patterns that only show up across hundreds of reviews. This is the gap ReviewVigil fills. It is a free Chrome extension: one click on any product page and it analyzes the visible reviews for exactly the patterns above — template phrasing, posting bursts, duplicates, rating-text mismatches, incentivized language. You get a letter grade from A to F, an estimated true rating once suspicious reviews are discounted, the percentage of reviews that look manipulated, and the specific signals it flagged so you can judge for yourself.

It works on any store that shows reviews, not just Amazon, and the free tier covers five checks a day. If you also want to know when a product page or price quietly changes after you buy, PageVigil's page monitoring handles that side. Either way, the habit is the same: never trust the average, always read the pattern.

← More from the PageVigil blog