SellerSort
Try it free

Who actually writes product reviews, and why it skews your average

Mostly customers at the extremes. Research on online reviews finds ratings pile up at five stars, with a smaller heap at one star and few in between, because the people who buy are already inclined to like the product and the people who bother to write are mostly the delighted and the disappointed. Your average describes the customers who chose to write, not all of your customers.

What this comes down to

  • Reviews come mostly from customers at the extremes, which is why ratings pile up at five stars with a smaller heap at one.
  • Researchers call this a J-shaped distribution and trace it to two self-selection biases: who buys, and who bothers to write.
  • A simple average hides the shape; the spread and the text of the middle ratings tell you more.
  • Asking every customer for a review, not only the happy ones, is the honest way to hear from the quiet middle.

SellerSort Team · Updated

The shape of almost every review page

Plot the ratings of a typical product and you do not get a bell curve. You get a tall bar at five stars, a shorter one at one star, and very little at two, three and four. In 2009, Nan Hu, Jie Zhang and Paul Pavlou named it in Communications of the ACM: the J-shaped distribution of product reviews.

Their point was not that customers are unusually happy. It was that the people who write reviews are not a random sample of the people who bought.

Your average rating describes the customers who chose to write, not the customers you have.

Two filters between your customers and your reviews

  1. Purchasing bias. People who expected to like a product are the ones who bought it. Someone who suspected the fit would be wrong never became a customer, so they never became a reviewer either.
  2. Under-reporting bias. Among buyers, those with a middling experience mostly stay silent. Writing a review takes effort, and “it was fine” rarely feels worth it. Delight and disappointment do.

Put the two together and the page fills with the extremes. The customer who found the product a bit flimsy, returned it, and moved on is the one you almost never hear from.

The other writers, and what the rules say about them

Customers are not the only people who write reviews. Google’s product ratings policy and the FTC’s reviews rule each name the other kinds, and say what happens to them:

Reviewers other than ordinary customers, and the rules on each
Who wrote itWhat the rules say
A customer given something for reviewingAllowed by Google only if the incentive does not depend on the review’s sentiment, and marked with is_incentivized_review. The FTC rule bans incentives tied to a particular sentiment.
An employee or insiderGoogle removes reviews written with a conflict of interest, such as by employees.
Someone who never used the productA fake review, banned by the FTC rule in force since 21 October 2024.
Another site’s reviewersGoogle product ratings will not take reviews a merchant syndicated from other sources; you must collect and own what you send.

So the reviews that survive those rules come from real buyers, and the J-shape is about which real buyers chose to write.

What that does to the numbers you read

The researchers’ advice was to stop relying on the simple average. Two products at 4.3 stars can hide very different shapes: one with steady fours, one split between fives and ones. The second has a problem that a slice of buyers hit hard, and the average gives no hint of it.

It also means the rarest reviews are often the most useful. A two- or three-star review usually comes from someone who wanted the product to work and is telling you precisely where it fell short, without the heat of a one-star rant.

Who is missing, and how to hear from them

The quiet middle does talk, just not on your review page. Their complaints turn up in return reasons, support tickets and pre-sale questions, covered in where to find customer feedback beyond reviews.

On the review page itself, the honest fix is to ask everyone. Google’s policy for Maps and Business Profile reviews says businesses must not selectively solicit positive reviews from customers, and the FTC’s guidance on its review rule warns that asking only customers you think are happy could break the FTC Act even though the rule has no specific ban on it. Asking every customer is both the compliant choice and the one that fills in the middle of the J.

Reading your own reviews with the bias in mind

  1. Look at the spread of ratings, not only the average.
  2. Read the two- to four-star reviews first. They carry the most specific complaints.
  3. Count topics, not stars. A complaint that appears in a steady share of reviews matters more than one furious review, which is the method in what customer review analysis is.
  4. Paste a batch into the free Review Analyzer to see the topics and severity without reading every one.

Questions

Why are most product reviews five stars?

Because reviewers are self-selected. Buyers already expected to like the product, and among them the delighted are far more likely to write than those who found it merely fine. Researchers call the result a J-shaped distribution.

Are one-star reviews less trustworthy?

Not less trustworthy, but more extreme. A one-star review is often accurate about what went wrong and exaggerated about how often. Check how many other reviews raise the same topic before treating it as common.

Should I only ask happy customers for reviews?

No. Google’s review policy for Maps and Business Profile forbids selectively soliciting positive reviews, and the FTC warns that doing so could violate the FTC Act. Ask every customer the same way.

Read next