Before I went to bed last night, I saw that the very first review of my new book, Mercury Rests, had been posted on Amazon.com. I was relieved and gratified that it was a five star review. Then I went to bed and had a dream that a good friend of mine had left a one star review, excoriating the book and bringing its average review score down to three stars. This is what it’s like being an up-and-coming author these days: bad reviews are literally the stuff of nightmares.
To someone who has never tried to sell a book on Amazon.com, obsessing over reviews may seem silly or even childish. Who cares what some faceless troll on the Internet thinks about your book? Dust yourself off and keep writing!
The effect of a bad review goes far beyond the impact that it has on the author’s ego, however. The prominence of a book on Amazon.com is determined primarily by two factors: how well the book has sold and how positive its reviews are. More highly rated books are displayed more prominently, which leads to more sales. Increased sales leads to even more prominent display, which leads to still more sales. Through the miracle of the positive review snowball effect, a few hundred rave reviews can transform an otherwise unremarkable book into a worldwide bestseller. Witness the success of author John Locke, who was propelled to international prominence and wealth on the strength of positive reviews of his crime thrillers. Locke has even written a book called How I Sold 1 Million eBooks in 5 Months. The only problem with this book, as veteran author Lee Goldberg points out, is that Locke left out a key part of his strategy: pay people to write fake reviews.
Locke has admitted to the New York Times that he paid 300 people to write fake reviews for his books. Many of these fraudulent reviews have evidently been removed, but the damage is done: hundreds of thousands of readers have been bilked into buying Locke’s drivel and he will undoubtedly continue to sell vast numbers of books simply on the value of name recognition.
You may be tempted to say, “So what? The people who bought those books based on fake reviews have only themselves to blame. They should have recognized that those reviews were fake. Caveat emptor.”
Putting aside the issue of whether readers should be savvy enough to spot fake reviews (which are not always as obvious as OMG BY THIS BOOK ITS SO GRATE), this criticism misses several important points:
First, Amazon is presenting these reviews as genuine. There is no disclaimer at the top of the list of reviews disavowing responsibility for fraudulent reviews. Yes, Amazon has strict Terms of Service that prohibit the posting of fraudulent reviews, but that only makes the situation worse by giving the impression that Amazon proactively polices the reviews section (as far as I can tell, while they respond to complaints, they do not have any procedures in place to proactively identify and remove fake reviews). No one goes to Craigslist or eBay expecting completely reliable information, but Amazon has a cachet of respectability that lends credence to reviews that appear on their site.
Second, reviews on Amazon.com are often presented in aggregate form, for example on a search results page where a thumbnail of a book is displayed with five stars and the number 300 in parentheses, indicating how many reviews the book has received. There’s no “Oh, by the way, 200 of those 300 are fake” notice on these pages. The only way to determine that any of the reviews are fake is to page through the reviews one at a time, and unless you want to spend more time reading reviews than reading books, that isn’t a viable option. (And again, this assumes that you can spot fake reviews, which is harder than it sounds.)
Third, as I’ve mentioned, the prominence of a book on Amazon is determined in part by reviews. It’s all well and good to say, “I’m a savvy consumer; I can make up my own mind about which books to buy,” but the fact is that when you shop on Amazon.com, you’re only going to see an infinitesimally tiny percentage of all the books that are available for sale, and which books are displayed is going to be determined in large part by what other people have purchased and which books have gotten lots of good reviews. Essentially what people like John Locke have done is to crowd out more deserving books by gaming the system. No matter how savvy a shopper you are, you can’t choose to buy a book that you’ve never heard of – and you might even buy a John Locke book just to see what all the fuss is about.
Finally, as Gresham’s Law states: Bad money drives out good. Or in this case: bad reviews drive out good reviews. As authors realize that it’s much easier to pay for good reviews than earn them, the business of creating fraudulent reviews will boom. And as consumers become more aware of the problem, genuine good reviews will become suspect. The whole system will collapse from the weight of fake reviews.
As someone who has spent literally hundreds of hours trying to locate and contact Amazon reviewers who might be willing to review my books – and over a thousand dollars of my own money sending paperback copies to reviewers with no guarantee that I will get a review (let alone a five-star review!), it’s hard to overstate how angry I am with shysters like John Locke for taking the easy way out. Not only is paying for fake reviews an insult to authors who have worked their asses off to write good books and get good reviews; it’s also an insult to conscientious reviewers who spend countless hours writing genuine reviews for which they are not paid a dime. And finally, it’s an insult to readers, who are being sold a bill of goods.
Locke couldn’t have pulled off this fraud without some complicity from Amazon.com, however. Amazon.com has access to loads of information that is not available to the consumer (such as the IP addresses of reviewers and their purchase histories) that could help them identify fraudulent reviews and ban reviewers with a history of fraud. Sudden bursts of reviews (as happened with Locke) should also have sent up some red flags in Seattle. There’s simply no excuse for letting Locke get away with these sorts of shenanigans.
Imagine if you could set up 300 dummy websites with links pointing to your book’s website and fool Google into making your book come up first in search results for “Best book ever.” You can’t, because Google is too smart to fall for such an obvious ploy. Google looks at things like the reliability of those sites, whether they are all using the same IP address, and whether they all came online last week. There’s a whole industry devoted to gaming Google (called Search Engine Optimization or SEO), and it gets harder and harder to do every year, because Google is always tweaking its algorithms to shut out scammers.
But that same strategy works wonders on Amazon.com because Amazon’s rating system is completely transparent and democratic: if a book has one review with a five star rating and a second reviewer adds a review with a one star rating, the book now has an average of three stars. Simple. Unfortunately, a completely transparent and democratic system is easily exploited, and there’s no reason to think the fraud in such an easily exploited system is limited to a few outliers like John Locke – he’s just the most shameless perpetrator.
Allowing its shoppers to be lied to and manipulated isn’t good business and Amazon knows it. They are going to have to do something about the problem of fake reviews – over and above simply reacting to specific allegations of fraud. At the very least, they are going to have to be more proactive about removing reviews that appear fraudulent and shutting down unscrupulous reviewers. Eventually, however, I think they are going to have to go to a more Google-like system, where they use a proprietary algorithm to give different weights to different reviews, depending on the reliability of the reviewer.
If Amazon doesn’t do something about this problem – and soon – look for someone else to fill the need for reliable product reviews. I’m thinking of something like Klout, which ranks a person’s influence across various social media platforms. By linking a person’s Klout score to their reviews and weighting the reviews accordingly, you could get a much more reliable aggregate product score. Or maybe Google will step in. They’re already the experts at this sort of thing; expanding into product reviews would be a natural fit.
And if the idea of Google stepping onto their turf doesn’t scare Amazon – well, then they aren’t as smart as I think they are.