Every Sunday when I was a kid in the tiny New York town of Hillsdale, my parents would drive five miles down a dirt road to the local grocery store to pick up the New York Times. My parents were transplants from the greater metropolitan area of New York City; I imagine that great hunk of newsprint helped them stay connected to the life they had—albeit willingly—left behind.
My mother immersed herself in the notoriously challenging crossword puzzle. My father delved into the sports pages to follow his beloved Yankees and into the financial pages to follow his beloved stock portfolio.
I went for the Book Review.
The New York Times Book Review was to me, and indeed to much of the literary establishment of the 1970s, the first and last word on books. I devoured the reviews that praised or damned, that elevated a writer to the national limelight or smashed all hope. And of course I dreamed of the day that one of my books would appear on its pages.
Forty-odd years later, I have resigned myself to the fact that no book of mine is ever going to be reviewed by the Times.
Goodbye to the gatekeepers
The dark underbelly of this system of appraisal was that only a tiny number of authors would ever see their work reviewed in the likes of the New York Times, the L.A. Times, or the New Yorker. Everyone else was, as we used to say, S.O.L.
Amazon, that 10,000-pound gorilla everyone loves to hate, has become the New York Times Book Review of our day. Writers, not to mention purveyors of garden gnomes and potpurri, feel only as good as their latest Amazon reviews. These reviews are fundamentally different from reviews by literary critics. Readers, not critics, are dispensing judgment. One can argue that readers’ opinions count more. However, it has always been possible for a work of art to be a critical success and a popular failure, and vice-versa. This raises the age-old question for artists: would you rather be critically acclaimed or sell a million books? The correct answer is both.
The changing calculus of book reviews and ratings has led to its own challenges. Sellers of books have figured out how to game the review system—something that was nearly impossible in the days of literary institutions and gatekeepers, and something that puts readers and buyers in a bind. How can we take a chance on a product we’re not familiar with if we think its reviews are fake?
My reviews got a failing grade
The explanations for the review ratings were often somewhat less than transparent. Here’s what Fakespot said about the reviews for What Remains Unsaid, with my comments in red:
- Our engine has analyzed and discovered that 26.3% of the reviews are reliable. (Based on…?)
- Interesting tidbit: the most used word by reviewers is book. (Well, yes, it is a book.)
- How are reviewers describing this item? unsaid. (Huh?)
- Our engine detects that in general the reviewers have a suspiciously positive sentiment. (I’d argue that 19 reviews is not a high enough sample size to assign any meaning to either a positive or negative trend.)
- Our engine has profiled the reviewer patterns and has determined that there is high deception involved. (What does “high deception” mean, exactly? I hope it’s not anything akin to “high crimes and misdemeanors.”)
Three tips for indie authors
So what can you do to avoid your reviews being downgraded? Here’s what I gleaned from running my reviews through the two fake review detectors.
Spread out your review requests. Some of the downgrading happened because too high a percentage of the reviews came on the same day. I can see why that would raise red flags. But the reality of independent publishing is that you have to take advantage of the flurry of activity that accompanies the launch of your book. And Amazon’s sales algorithms work in a way that encourages lots of sales on the first day or two of publication, so you rise through the rankings.
Be wary of sending free review copies to Amazon reviewers. Of course for traditional book reviewers, there’s nothing unusual about receiving a free copy of the book from a publishing company. But free product in return for reviews is verboten on most product review sites. And on Amazon, readers who don’t purchase a copy are not VERIFIED PURCHASERS, which discounts their input. Better to direct those people to post a review on Goodreads.
Don’t take fake detection too seriously. Maybe I’m saying that only because my reviews fared so poorly. Yes, I’m all for sussing out fake reviews. They have become a big problem, not only on Amazon but on other sites such as Yelp. But it seems to me that an algorithm is not always great at distinguishing between a legitimate author with a small number of legitimate reviews by fans and someone who is gaming the system. And yet I have little choice but to play along.
Write a fantastic book about something you are passionate about, pay attention to the details of producing it, and don’t worry too much about reviews.
The algorithms are coming
And you thought automation was endangering only factory jobs and maybe eventually service jobs. It’s coming after the arts, too. Algorithms will write the books, as in a recent experiment described in Wired. We already have news-writing bots. Can review-writing bots be far behind?
Fight back while you can. If you’re a human, leave a review of What Remains Unsaid or Dance of Souls. Just don’t submit yours on the same day as everyone else. And don’t make them it too positive. And don’t give five stars to too many other products.
My algorithm thanks your algorithm.
- The Book Stop on fake reviews
- The New York Times talks about reviews for hire
- Forbes on Amazon’s fake review problem
- David Gaughran talks about another problem: fake books