How Changes to Amazon’s KDP Select and Review Rankings Impact Authors

Here is how some authors have been gaming self-publishing on Amazon:

Write a bunch of short books and get a bunch of friends (or paid reviewers) to give you 5 stars. Add your books to KDP Select—the program that pays authors every time a book is borrowed—and make some money.

This will be a strategy of the past if the latest changes Amazon is implementing have their intended impact. While these two changes are not tied together, they do form a solid foundation for gaming two hugely popular Amazon programs.

Let’s look at each one and consider the ramifications.

KDP Select

KDP Select, the program that pays a Kindle eBook author a couple dollars every time their book is borrowed (only if the eBook is available exclusively from Amazon), has always been tilted in favor of authors of short books. Authors of books that are 60 pages (for example) get paid the same amount of money each time a book is borrowed as an author’s whose book is 300 pages. Unfair, considering it is far more time consuming to write those longer books. I’m sure it was also much easier for Amazon to track and calculate “borrows” as a percentage of all borrows, which is how they computed the payout.

This leads to an incentive to crank out books without regard to quality and depth, which unfairly competes against the fiction writer who might spend months or years on their novel. I think it’s also safe to say that a reader gets more hours of enjoyment from the longer book. Since KOLL (Kindle Owners Lending Library) allows a single free borrow per month, a short book isn’t going to be as satisfying. It won’t have the same value, which is a reflection on the value of Amazon Prime, of which KOLL is a benefit.

The upshot is that it makes KOLL less attractive to fiction writers and it is fiction readers that are considered the most voracious readers and therefore more valuable to subscription programs* like KOLL and Kindle Unlimited (KU).

I think this change addresses the question of how to get more quality novels to be part of KDP Select.

So what exactly is the change?

Each month Amazon transfers a few million dollars into their KDP Select fund which is used to pay authors who participate in KDP Select. Up until June 30, 2015, Amazon would look at the number of books in the program and figure out the number of times each book was borrowed. If your book was borrowed, you get a percentage of the pie based on the number of times it was borrowed relative to all other borrows that month.

The new method of calculating a book’s share of the pie is now based on the number of pages a reader reads. The author of a long book is potentially going to make more money than the author of a short book.

Of course this change also has a way of rewarding quality. A long book that does not capture the interest of the reader simply won’t be read “cover-to-cover.” Authors are incentivized to improve the quality of their work, or at least be honest in how they present it.

Simply getting a subscriber (KOLL or KU) to download the book will not generate KDP Select revenue in and of itself.

Reviews and Review Rankings

Reviews, and the star ratings that score them, are the life-blood of a book. They are used by readers and potential business partners alike to judge a book. Unfair as that may be it is one of the few signals someone can use to make a snap decision about whether to buy or borrow a book.

It is also often used by advertising firms, such as those that promote KDP Select books, to help determine whether a book is worthy of promotion to their readers. Just about all of them are the same or are similar to this one from the book promo service BookGorilla:

There are no doubt thousands of quality books that do not meet this criteria (like Henry Miller’s classic Tropic of Cancer!). It is a system that favors the socially connected, or those willing to pay one of the many review services who have sprung up to write positive reviews for payment.

(By the way, books are not the only product that suffers from inaccurate or unfair reviews. A manufacturer who releases a product and then improves it based on customer feedback [isn’t that what we want?] will be stuck with unfavorable ratings and rants forever. Prospective buyers have to wade through early, sometimes irrelevant, reviews to see if a particular issue was reported by another reviewer as resolved).

Amazon’s new artificial intelligence system aims to change this

• Gone is the simple averaging of ratings. Reviews will now be weighted based on a number of factors.
• Gone is the rating that changes only when new reviews are added. The overall rating may now change over time.

Amazon spokesperson Julie Law wrote in an email to InformationWeek:

The enhanced system will use a machine-learned model to give more weight to newer, more helpful reviews from Amazon customers. The system will continue to learn which reviews are most helpful to customers, and improve the experience over time.

Authors should keep these key factors in mind

1. More weight will go to reviews based on verified purchases (books bought from Amazon rather than other retailers).
2. Votes on the helpfulness of reviews count in overall weighting.
3. Averages are subject to change over time.
4. Newer reviews receive greater weighting.
5. The words used by reviewers, and the length of the reviews, matter.

How does Amazon weight all these? It's anyone’s guess. It's comparable to trying to figure out how Google determines search result rankings. Both companies consider this information proprietary because they don't want the public gaming their results.

In short, it’s going to become a lot harder to game reviews.

Where does this leave us?

It looks like the goal is to swing the pendulum back to quality. Amazon wants reviews to be reliable and trustworthy. Their tweaks to KDP Select aim to make it more fair and attractive to authors.

We’ve had tremendous disruption in the publishing world the past seven years which has unfairly favored those willing or able to manipulate the system. Amazon itself has taken heat for being complacent in its role as a leader. Not only is it the single largest seller of books, but perhaps more impactful, the book “buy-page” on Amazon has a way of influencing sales and author reputation beyond the confines of its website.

I consider these welcome changes that stand to benefit committed authors everywhere. Let’s hope it works.

* * * * *

Postscript

*It is also worth pointing out the business benefit for Amazon if they succeed in attracting better quality books, especially to their subscription programs. While many analysts don’t consider KOLL a formal book subscription program, KU is and it was late to a market pioneered by 24Symbols, Scribd and Oyster (the so-called “Netflix of books” model).

It is widely reported that these other services pay publishers by the borrow, a costly proposition for the services and a bargain for avid readers. Since Amazon sells the most eBooks, and they have a proprietary eReading app, they alone are the ones who can switch to this potentially more accurate way of measuring and calculating consumption of books (counting pages). Better information can inform all sorts of negotiations and decisions like which books to add to the subscription program, which ones to avoid, how to incentivize authors, etc.

In light of the KDP Select and ratings changes it remains to be seen how this will play out for the subscription business model. But we may have gotten a clue on June 30 when Smashwords founder Mark Coker announced Scribd was making dramatic cuts to their catalog of romance and erotica titles.