Amazon has more than 1.5 billion product reviews. Studies estimate that anywhere from 30% to 42% of them are fake or incentivised. That means every time you rely on a star rating, you're rolling the dice. The patterns below won't catch every bad actor — but they'll make you a significantly harder target. And once you can spot these patterns manually, you'll understand exactly what automated tools like BuyWise are doing on your behalf.
A flood of 5-star reviews in a short window
Genuine products accumulate reviews over months and years. If a listing shows 200 five-star reviews posted within the same week, that's almost never organic. Sellers using review farms push batches at once to establish an initial rating and break into Amazon's search algorithm. Look at the review date histogram — a natural product has a slow, rising curve. A boosted one has sharp spikes followed by dead periods. The contrast is often dramatic once you know to look for it.
Reviewers with no history or identical profiles
Click through to a few reviewer profiles. Red flags: joined recently, reviewed only this product, reviewed many unrelated products on the same day, or has a generic name and no photo. A cluster of reviewers all posting on the same date from accounts created the same week is a near-certain signal of coordinated fake review activity. Some sellers operate whole networks of these accounts, often managed through encrypted messaging apps like Telegram.
Generic, vague praise that doesn't mention the product
Real reviews say things like "the strap broke after two weeks" or "fits perfectly with my 2022 MacBook Pro." Fake reviews say "Great product! Very happy with my purchase. Would recommend to everyone." The vaguer the review, the less useful — and the more suspicious. Paid review writers are typically given minimal product information and a word count to hit. Specificity is effort. Fake reviewers avoid effort.
Excessive exclamation points and ALL CAPS enthusiasm
Paid reviewers are often incentivised to sound enthusiastic. Real buyers are more matter-of-fact. Natural language models trained on verified-purchase reviews consistently show that authentic reviews have lower sentiment intensity than incentivised ones. When every sentence ends with an exclamation point and the reviewer describes buying a power strip as a life-changing experience, the signal is clear.
Unverified purchases dominating the review count
Amazon labels reviews from confirmed buyers as "Verified Purchase." A product where most reviews lack this label wasn't necessarily bought through Amazon — which makes those reviews much easier to fake. Filter to Verified Purchase only and see if the rating holds up. In many cases, a product with a 4.6-star overall rating drops to 3.2 stars once you apply the verified-purchase filter. That delta is your answer.
Star rating doesn't match the written reviews
Scroll down and read the 3-star reviews — they're often the most honest because they weren't specifically incentivised. If the written reviews describe real problems but the overall score is suspiciously high, padding is almost certainly happening. A 4.7 average alongside dozens of 3-star mentions of the same defect is a giveaway. The disconnect between the number and the narrative is usually a reliable signal.
Suspiciously new listing with thousands of reviews
A product listed three months ago with 4,000 reviews is acquiring them at roughly 44 per day. Amazon's genuinely best-selling products rarely do that organically. Cross-check the listing date (visible in the product details section) with the review count. If the math doesn't add up, the reviews don't either. New sellers in competitive categories are often the most aggressive at review manipulation because they have no organic reputation to build on.
The same product relisted under a new ASIN
When a product accumulates too many negative reviews, sellers sometimes create a fresh listing under a new ASIN to start with a clean slate — sometimes importing fake positives from the start. If the images look identical to another listing you've seen, check whether the ASINs differ. Tools like BuyWise flag when a product's review velocity is inconsistent with its listing age, catching this pattern automatically.
What to do when you spot fake reviews
If you identify a product with manipulated reviews, you have a few options. First, simply don't buy it — there's almost always a competing product with more legitimate social proof. Second, report it to Amazon using the "Report abuse" link on any individual review. Amazon takes these reports seriously because they undermine the platform's core trust proposition.
Third, look for the product on a different retailer. The same product sometimes sells on Walmart or Best Buy with a completely different (and often more accurate) review set, because it's harder to import fake Amazon reviews to other platforms. A product with a 4.7 on Amazon and a 3.1 on Walmart for the same SKU is telling you something important.
Why manual checking only goes so far
The eight patterns above are learnable, but applying them manually to every product takes five to ten minutes per listing — and sophisticated fake review operations have adapted to many of them. Modern fake review campaigns use AI to generate varied, specific-sounding text, stagger posting dates across weeks, and create aged accounts with plausible review histories. The arms race has shifted.
This is where automated tools become useful. BuyWise analyses over 50 signals simultaneously — including ones that are impractical to check manually, like cross-referencing reviewer behaviour patterns across thousands of products, or detecting subtle language anomalies at scale. The goal isn't to replace your judgement; it's to handle the groundwork so your judgement can focus on the signals that matter.
Don't do this manually for every product
BuyWise checks all 8 of these signals automatically — plus 40+ more — in seconds. Free to use, no account required.
Add to Chrome — It's Free