How to Read a Product Review Before You Trust It
A five-star average tells you almost nothing on its own. The distribution underneath it tells you a great deal.
Online reviews are the main way most of us decide what to buy, and they are also the part of the buying process most exposed to manipulation. You do not need to distrust all of them — you need a short routine for telling the useful ones apart.
Read the distribution, not the average
A 4.4-star average built from mostly fours and fives is a solid product. The same 4.4 built from a wall of fives and a hard cluster of ones is two different populations: people who received a working item, and people whose experience went badly wrong. Always open the histogram, then read the one- and two-star reviews first. They tell you the failure modes.
Look for specifics that only an owner would know
Genuine reviews are boringly concrete. They mention how loud it is at 2 a.m., that the power cable is shorter than expected, that the filter is fiddly to remove. Manufactured reviews stay abstract and enthusiastic: great product, fast delivery, highly recommend. Length is not the signal — specificity is.
Check the timeline
Dozens of reviews arriving within a few days, then near-silence, is a pattern worth noticing — especially on a newly listed product. So is a long run of reviews that describe a completely different item, which usually means the listing was repurposed after the fact.
Read the negatives with judgement too. A one-star review about a late courier tells you about the delivery company, not the product. Weight the complaints that describe the item itself and that repeat across multiple reviewers.
Separate reviews from review sites
A published "best of" list is a different thing from customer feedback, and it is worth asking how it was assembled. Reasonable questions: did the author physically test the products, or compile specifications? Is the ranking order explained? Is there a disclosure about commercial relationships? Would a product with a poor result ever appear on the list at all?
Commercial relationships are not automatically disqualifying — most independent publishers, including this one, disclose that they may earn a commission on links. What matters is whether the disclosure exists, whether the criteria are stated, and whether anything ever gets criticised.
Signals that a page is advertising rather than journalism
- A countdown timer, a stock counter, or "only 3 left" language
- Every single product in a category rated positively
- Comparison tables where the competitors are unnamed generic alternatives
- No author, no date, no way to contact anyone
- Testimonials with stock-photo portraits and first names only
None of these mean the product is bad. They do mean the page is a sales asset, and should be read as one.
A practical two-minute routine
- Open the rating distribution; read three negative reviews and three positive ones.
- Search the product name plus "problem" or "returned" and read what comes back.
- Find one source that is not selling the product — a forum thread, a video, a consumer body.
- Check the returns window and who is responsible for it before paying.
Two minutes will not make you an expert. It will reliably catch the purchases you would have regretted.