Amazon and Gaming the Objective Review

First, let me congratulate Amazon on a recent update. I presume authors reading this have noticed that when you do a search on a title, Amazon has changed how the customer reviews are presented: 4.3 (80% 4 or above). Why is this a good thing … and why does more need to be done …? Read on.

The Good Thing

Amazon’s new reporting of customer ratings is a step toward overcoming the pull of the extremes. A simple example is this: The first two readers gave the book a 4 and a 5. Along comes one curmudgeon who gives it a one (they thought it was a ghost story when it wasn’t). The book now has a customer score of 3.3. Though the book will never achieve a 5-star rating (ever), readers looking for a solid mystery will see that 66% of the raters gave it a 4 or above. In other words, most readers scored the book high, indicating the low scores are random and incidental.

Disturbing Trends

There are two disturbing trends, the first promoted in how-to-books on independent publishing and by marketing/sales gurus. That is to game the customer review system by having one’s followers or sites offering customer reviews flood a book’s page (particularly a new book) with 5-star reviews. Not a bad plan, except it makes a score of 5 meaningless and it hurts all buyers. If the scale means nothing on what does a mystery reader base their purchase? Further, it hurts new or other authors of equal quality with fewer followers.

The other, far more disturbing trend is this (recently noted in a New York Times article). A book consistently scoring in the 4-5 range is hit by a host of 1s. Within days of each other. To be clear, to be a 1 a book should be abysmal. So, a book that is truly a 4 or 5 cannot turn into a 1 overnight. It cannot. How does this happen?

Two ways. One, a reader dislikes something about the book that is a trigger point for them. That reader gives the book a 1 based on their trigger, then convinces friends or their reading group to do the same. It is a protest of sorts having nothing to do with the quality of the book, but a way to dis- and ban books that do not agree with one’s belief whether those beliefs are mainstream or not. And second, it may be used by authors or marketers to increase a book’s sales by effectively taking out competition.

What makes this possible is that a rater doesn’t have to read a book to rate it, which tells us that all customer scores are suspect. Amazon’s own policy reads: You can leave a rating or review for a book that you didn’t purchase on the site*. And when an Amazon review indicates a certified purchaser, it means: You must have spent $50 on Amazon.com, using a credit or debit card, in the past 12 months, to: Create reviews (including star ratings) Answer customer questions. Submit helpful votes. * These policies make it possible to line folks up to either praise or torpedo a book.

What Can Be Done

I don’t know how to stop this, fight it, or change it. Any change is unlikely with the publishing environment so competitive, readership down, and gaming the system (though inherently immoral) considered a legitimate marketing tool.

The only sensible solution is that all scores (but especially 5s and 1s) require a written review to count. The review cannot be one word “garbage” or “wonderful” but an explanation of fifty words or more about why the score given is valid and that gives some assurance the book has been read. This way the poor person trying to discover a new mystery can decide if the score of 5 or even 1 has any virtue.

In the meantime, Amazon has taken a step in the right direction.