I don’t normally do much off-the-cuff editorializing on my blog, but I’m intrigued by Amazon’s recent announcement that they are going to attempt to get rid of bogus reviews.
I think it is a great news. This means that serious novel writers will have less BS competition out there.
I don’t think their methodology will be perfect at first, but my guess is they’ll put together an algorithm that considers such things as length, spelling errors, how many reviews are coming from a given reviewer, how many are coming from a given IP address and/or MAC address, are devoid any of the proper nouns found in the book, it is not a verified purchase and they don’t see that it was ever read on a Kindle, etc.
All these could be spoofed and no one or even two would prove much. On the other hand, if they see 10 book reviews a day coming out of the same MAC address (a computer’s unique identifier), a reviewer that consistently scores book a “5”, they are short, have spelling errors and no mention of any of the proper nouns in the book, get the author’s gender wrong, end with “I can’t wait to buy the next book in the series!”, and such “reviews” have no (or very few) “helpful” votes, it is a reasonable guess that such a review might have been purchased on fiverr.
From there, if they look at all the reviews for the book in question and 10 of the 12 reviews (for example) have similar profiles, then it is probably a safe bet that the author is buying reviews.
It is not going to be a perfect science, but given all the work Amazon (with Google leading the way) on indexing content, my guess is that they will be right far more often than wrong.
And could a legitimate review have all these suspect characteristics? Yes, it is possible, but highly unlikely. And even if you got such a review, even if were 5 stars, would you care if it got deleted? As a rule, I won’t read a book that looks like it has BS reviews based on my own BS sensor.
I am curious about the report that they will downgrade reviews from people who were not verified purchasers. I have in fact offered free copies to people have done early reviews for me. Some have reported that in their review. Some have not. And doing so is a common practice in the traditional world of publishing. It occurs to me that this could be solved in several ways. For instance, such a reviewer could be forced to check a box stating they got a free review copy, and such would be limited to 25 (just to make up a number) by Amazon.
In short – with the one caveat ref free review copies – I’m comfortable with where Amazon is going with this, though it might take them a few iterations really get it right. I can’t imagine how a legitimate author who earns reviews the old fashion way – quality writing, well-prepared product – is going to fear this.
Other thoughts?
