Every Amazon rating is a summary of history, and summaries hide things. Two products can both show 4.6 stars and 800 reviews while having nothing else in common: one earned them steadily across three years, the other collected most of them in a single enthusiastic month. The average cannot tell you which one you are looking at. The dates can.
The thirty-second check
Open the reviews and sort by most recent. Read the first page. Then glance at the spread: are the reviews distributed across months and years, or do they arrive in bursts? A burst of five-star reviews packed into a two-week window has a small set of explanations, and most of them involve a seller campaign rather than 400 strangers spontaneously agreeing.
Why recent reviews outrank old ones
Products change under the same listing. Sellers reformulate, switch factories, or swap components while the listing keeps its accumulated stars. The reviews from three years ago describe a product that may no longer exist; the reviews from last month describe the one that will arrive at your door. When old reviews glow and recent ones complain about quality, believe the recent ones.
Other tells worth ten seconds each
Check whether reviews are marked as verified purchases. Read the one-star and three-star reviews first, because complaints are specific where praise is generic, and patterns in complaints (arrived broken, not the size pictured, stopped working in a month) are the most honest data on the page. And be suspicious of reviews that describe a different product entirely, which happens when a seller reuses an old listing's rating history for something new.
The vacuum nobody filled
A few years ago, free tools existed that analyzed review authenticity at scale. Millions of people relied on them, and they are gone, shut down or abandoned, while manufactured reviews grew more sophisticated. We think that vacuum is unacceptable. We are True Finds, and reading the reviews so you do not have to guess is on our roadmap, right behind price truth, where we started.