For those who grit their teeth and deal with the satanic behemoth, I spotted this on twitter:
"As someone working w/ Amazon marketing, I can confirm that it is HIGHLY important to review your favorite authors' books. HIGHLY. Ideally a book should have minimum 15 reviews & at least a 3.5 star rating for certain algorithms to kick in."
Disclaimer: I have no idea whether this is genuine/correct.
AJ
It would make sense, there are two different criteria they're using. There is, of course the "also bought" one. Then there is the "suggested reading" which more likely than not is following breadcrumb trails left by people giving positive reviews.
So if a reviewer read and reviewed book Xyz and you just finished that book and were looking for next book suggestion, well Amazon is going to note that same reviewer also gave high marks to this other book so you might like it too.
Not so reliable when you're looking at a pool of 1 or 2 (or 12) reviewers. But if instead you had 200 people read and favorably review Xyz, then say 60 of them go on to also do the same with a second title, regardless of author or other factors, there is a very strong correlation there.
It would make sense, there are two different criteria they're using. There is, of course the "also bought" one. Then there is the "suggested reading" which more likely than not is following breadcrumb trails left by people giving positive reviews.
No, the 'also bought one' is exactly what it says. It has nothing to do with who reviewed a book (although it likely doesn't kick in until you've hit certain metrics), instead it's based on other readers actual purchase histories (i.e. what they bought immediately after purchasing the listed book).
I've used it to direct sales, running ads that said "similar to" the listed books, but it had zero traction (probably because all of my readers already had the books!).
No, the 'also bought one' is exactly what it says. It has nothing to do with who reviewed a book (although it likely doesn't kick in until you've hit certain metrics), instead it's based on other readers actual purchase histories (i.e. what they bought immediately after purchasing the listed book).
Which is why they're "different criteria" and their own category for suggestions. :)
I may have phrased that poorly, but I was trying to describe pretty much what you just did. "Also purchased" is its own thing.
The suggested reading however, likely follows the reviewers around.
The suggested reading however, likely follows the reviewers around.
Correct (both your clarification and the quoted line), the "suggested readings" are based on a combination of ratings and overall reviews.
Unfortunately, it's true. While I initially garnered several reviews for each of my books (nowhere near 15, though), now it's rare for readers to register their views on either new or my older books.
The other thing to keep in mind, is that the algorithms are continually churning, so they don't report how well your book has done over time, just over the last several week (I forget the interval, but I suspect it's around six weeks), after that, all your purchases don't amount to jack shit. They ONLY promote what's hot right now.
