It is a fair question, and the honest answer sits between two extremes that both circulate online: "reviews do not matter at all" and "reviews are the single biggest lever for sales". Neither is accurate. Reviews are one input among several that shape whether a browsing reader decides to buy, and their weight varies depending on the reader, the genre and how uncertain the buyer feels before reviews are taken into account.
This guide separates the different things people mean by "reviews" — count, rating, content and recency — and explains what each can plausibly contribute, without asserting causal weight that Amazon has not published. It builds on our earlier guide to [why a book has no reviews](/resources/book-sales-growth/why-does-my-book-have-no-reviews) and the [cornerstone review-acquisition guide](/resources/book-sales-growth/how-do-i-get-more-reviews-for-my-book).
What reviews can plausibly do
Reviews function mainly as social proof at the point of decision: they tell a hesitant buyer that other people bought the book and were willing to say something about it. For a debut author with no other track record, this can matter more than for an established author whose name already carries some trust. Reviews can also surface useful information — who the book is genuinely for, what to expect — that a description alone might not convey as convincingly.
What Amazon does not publish
Four distinct signals, not one
| Signal | What it may indicate | What it does not tell you |
|---|---|---|
| Review count | How many readers were engaged enough to write something | Total sales, since most readers never leave a review |
| Average rating | Broad reader satisfaction, at a glance | Whether the book suits a particular reader’s taste |
| Review content | Specific strengths, weaknesses or audience fit, in the reviewer’s own words | Whether those points apply to every reader |
| Recency | Whether the book is still being actively read and discussed | Current sales volume on its own |
A reader relevance factor also matters in practice: a review from someone who is clearly the target reader for a book tends to carry more weight with other similar readers than a generic one-line review, even though Amazon does not formally score reviewer relevance in any published way.
Reviews versus actual demand
A common confusion is treating reviews as a stand-in for demand. They are not the same thing. A niche non-fiction title can have excellent, thoughtful reviews from a small, satisfied readership and still sell modestly, simply because the addressable market is small. Conversely, a book can sell reasonably well in its first weeks, on the strength of visibility and price alone, before it has accumulated many reviews at all.
Example
Imagine two comparable business books. One has thirty reviews averaging 4.8 stars but reaches very few potential buyers because its categories and keywords are poorly matched to how readers search. The other has six reviews averaging 4.2 stars but strong metadata and a well-targeted description. The second book may well outsell the first, because reach and relevance are doing more work than review count in this case.
When reviews are worth prioritising
Reviews are most worth prioritising once the product page is already reasonably strong — clear metadata, an accurate description, a genre-appropriate cover — and traffic exists but conversion seems to lag. In that specific situation, a lack of reviews may be adding to buyer hesitation. Prioritising reviews before addressing weak metadata or low traffic tends to produce a smaller return, because there is little traffic for the reviews to influence yet.
Checking your own book
Where do reviews sit in your diagnosis?
- Is the book already getting reasonable views or clicks on its product page?
- Is the description, cover and pricing already doing a fair job, as far as you can tell?
- Do you have very few reviews relative to a reasonable expectation, given sales so far?
- Have you checked traffic and metadata before assuming reviews are the limiting factor?
Common mistakes
- Treating review count as a proxy for sales. Most readers who buy a book never leave a review, so review count understates actual sales and should not be read as a sales figure.
- Assuming a formula links reviews to ranking. Amazon does not publish how, or whether, reviews factor into ranking directly; treating an unofficial formula as fact leads to wasted effort.
- Chasing reviews before fixing a weak product page. Reviews cannot compensate for poor metadata or low traffic; fixing those first tends to matter more.
- Ignoring review content in favour of the star average. The average rating alone misses useful, specific information in what reviewers actually wrote about who the book suits.
- Assuming a niche book with few reviews has failed. A small addressable market can produce modest review counts and modest sales together, without either indicating a problem.
Frequently asked questions
Does Amazon use star ratings directly in its ranking algorithm?
Amazon does not publish the details of its ranking system, including whether or how star ratings are used. Any specific claim about this is not confirmed by Amazon.
Is a 4.0 rating too low to sell well?
There is no published minimum rating required for a book to sell. Amazon does not publish a threshold, and many books with a range of ratings sell at a range of levels for other reasons.
Do recent reviews matter more than older ones?
Recent reviews can suggest a book is still being actively read, which may reassure some buyers, but Amazon does not publish a specific weighting given to recency.
Can a book sell well with almost no reviews?
Yes. Reviews are one input among several; strong metadata, a well-matched cover and description, and genuine reader demand can all support sales without a large review count.
Should I focus on getting reviews or improving my product page first?
Generally the product page — description, categories, keywords and cover — first, since reviews mainly help convert traffic that is already reaching the page.