Book Reviews, Goodreads, and Amazon

A while back I discussed the value of book reviews with some author friends. One-self pubbed author offered some (to me) amazing information and I share it here with the big caveat (from the source) “Please take nothing here as gospel.”

First, everything offered is based on research and articles read over several years, and a lot of informed guesswork.

This stuff is a rabbit hole and I find my time is better spent reading books and writing novels than trying to see the wood for the trees with this stuff.


The above is an old graphic, the thresholds have changed and it over simplifies. It’s from an article way back, one of several that all agreed in principle on how it worked, but varied hugely on numbers, the list offered below is also an oversimplification and about the average threshold numbers from memory.

There are a shit ton of articles out there, but a fair few of them are screwed by people trying to sell services, ‘we can help you beat amazons algorithm, just send us money’, etc. As such I am always a little dubious of their ‘facts’ though the concessions of those facts adds up to what is described.

The best of these is KDP rocket, and it’s a one off fee with a whole lot of support, but it is also a rabbit hole.

If your goal is improving book sales through reviews…
Goodreads is not majorly important, its useful for getting in front of readers, but less so than it was in its hayday; Goodreads has been in steady decline for several years.

Amazon reviews do help sell books, but not because of the reviews themselves; it’s a numbers game to do with amazons site algorithms, and that’s a complex bucket of squid in seawater. The ‘rules’ change every time amazon tweak its algorithms, and amazon’s only goal in those tweaks is to sell product (ie, get books people will want to buy in front of people who will want to buy them). So the aim is to get the algorithm to work for you rather than against you.

Amazon tweaks its algorithms all the time, sometimes these tweaks are tiny, sometimes they are huge, and weirdly the impact of the tiny ones can be the most dramatic.

The algorithm works by thresholds
If a book has 15+ reviews the algorithm will be more likely to push it towards readers as a ‘you may also like this’ when they search for a book in the same sub genre (ie military scifi).

If a book has 30+ reviews the algorithm is more likely to push it towards readers as a ‘you may also like this’ when they search for a book in the same genre (ie scifi).

If a book has 60+ reviews the algorithm is more likely to push it towards readers as a ‘recommendation’ in emails from amazon. when they have previously bought books in the same sub genre (ie military scifi).

If a book has 90+ reviews the algorithm is more likely to push it towards readers as a ‘recommendation’ in emails from amazon. when they have previously bought books in the same genre (ie scifi).

All of those also apply to where books fall on the amazon sites when people search for something, ie does you book appear on the first page of a search or the tenth…

However reviews are only a part of the algorithm
The review numbers for those thresholds also change every time they tweak the algorithm, and every time they alter rules on reviews.

Of course the algorithm does not just go by reviews and genres, it also pushes books based on recent sales figures and rankings. So a new book with high first week sales and only a couple of reviews will get pushed ahead of a book that’s got 50+ reviews but has been out five years and has only a trickle of sales. The algorithm gives each book a score based on lifetime-sales/reviews/ranking etc before determining which books to chose to push based on the likelihood of further sales.

How much weighting the algorithm gives to numbers of reviews in that calculation also varies every time they tweak something. For example a couple of years ago they changed the weighting percentage of number of reviews vs all the other considerations from around 40% to 25% and the entire marketplace changed: some authors dropping off the face of the earth sales-wise despite having good reviews because the onus was all placed on the newest releases rather than books with lot of reviews. A month later it swung the other way.

Also the number of reviews is more important than the rating
The algorithm doesn’t care about the average number of stars, only the number of reviews. A one star review is crappy but if that one star review takes you over a number of reviews threshold it is actually a good thing that would lead to more sales. Except at times the average rating has been part of the algorithm… and its weighting also changes, so at times a lower sub 4 rating is detrimental before you even get in front of a prospective reader, at other times it isn’t…

Also of course, amazon does not publish the details of its algorithm, all information on the latest iteration is subject to what data mining people have done and best guess at how it currently works which is generally a few days before it changes again.

The upshot of all this is exactly how important amazon reviews are varies, quantity usually trumps quality but not always. And they only matter beyond vanity in terms of the numbers game (a one line review is just as valuable as three paragraphs of gushing praise).

Amazon’s only goal is to sell more product
Its algorithm is designed to do that, not put great literature in front of readers. The algorithm is the same one it uses for everything on amazon sites so it is treating your book in exactly the same way as it is dealing with a box of pencils.

All of this is also subverted by sponsored contact, ie paid advertising on amazon sites, which trumps the algorithm.

All of this is subject to change, and trying to guess the current import of reviews if like throwing darts at a dartboard blindfolded. You’ll hit something, you’re not guaranteed it will even be the board.

Does the book cover appeal to the eye?
All of this is also just about getting a book in front of a prospective reader on amazon who has not searched for it specifically. That’s before they will ever actually read the reviews, which most people don’t. Frankly at that point having got the book on the screen in front of someone the most important thing becomes not the number of reviews or the quality of them, but does the book cover appeal to the eye.

Is it worth your time?
Anyway, all the above is probably out of date. But ultimately any promotion of your books is good, and more reviews (preferably good ones) is also good. The question becomes simply one of time, ie is it worth your time based on the likely outcome, as opposed to other forms of promotion.

I would argue that $175 dollars is better spent on amazon advertising, that’s around the last two months worth I spent. It isn’t guaranteed but it does get you sales and amazon ranking both of which have an influence on the algorithm as well.

Bottom Line
There are no guarantees with amazon advertising, but you do only pay by the click, so it costs nothing if no one ever looks at the book, and it has always generated sales for me, without ever costing me a fortune in the process (on average I probably come out around even, but its about putting books in the hands of readers not profit for me).