Speechwriting: Corporate, Weddings, Retirement

6/21/15

Amazon's New Review Ranking System

I've been reviewing on Amazon since 1998. Long time. It started with seeing a book without any information from even the publisher, let alone reviews. Since then, my adventures have been many. Met reviewers in New York City -- at the Algonquin Hotel. Tangled heavily with a disgruntled author. I help lead a Facebook group discussing reviews (and chastising unethical reviewers). Google me to see one or two.

Most notably, albeit quietly, I was ranked as high has 59. I was even, in some circles, kind of famous for street cred I earned (more Googling for you). Now, I rank near 1,100. Take a look. At some point in 2008, Amazon restructured their ranking system which reduced the impact of my earlier reviews and I slowly lost ground.

Meanwhile, a machine of corruption regarding reviewing has kicked in. If say, you are an unscrupulous reviewing, you might take payola from a desperate self-published author, or you might trade reviews as you publish your book. a little quid pro quo. There are more scams than I have time or inclination to describe.

This week, Amazon announced a new upgrade (hopefully) to their ranking system. Again, the freshness of the review is a factor. What else? How many votes a product gets. Those two factors already exist, but they have adjusted them further, enough so as to make an announcement.

Whatever the actual algorithm is, I don't know. But I hope it reduces fraud in all its forms. It is hard to do since this is all AI stuff, not human mind reading (impossible, but it would help).

I'm hopeful. But realistic.

Here's the letter:

Hello,

We want to thank you for your contributions to Amazon’s customer reviews. The product reviews you provide help customers make more informed purchasing decisions.

Starting today, Amazon is enhancing the customer reviews system, adding a few changes we hope will help make product feedback even more useful to customers. The enhanced system will use a machine-learned model to give more weight to newer, more helpful reviews from Amazoncustomers. The system will continue to learn which reviews are most helpful to customers and improve the experience over time. This change will present itself in two ways:

- Star rating: A product’s overall star rating will now consider factors including the age of a review, helpful votes by customers, and whether reviews are from verified purchasers.
- Review ranking: Similar machine-learned factors will help determine a review’s ranking in the list of reviews.

Our goal is to help customers make even more informed purchasing decisions.

Thank you,
Amazon Customer Reviews Team