I've seen quite a few people try to explain how Amazon arrives at its sales rankings, but in the end most give up, because no explanation quite seems to fit.  It's a bit like trying to pick up raw egg white with your fingers.  It's there; you can see it; but it constantly escapes your grasp.  One reason is that the rankings change every hour, so that no explanation will be fully accurate long enough to write it down.

Here's what we do know, and what we don't.  It seems entirely possible that Amazon now carries over 10 million books, although less than six million have ever sold a copy.  Once the first copy of a book sells, the book gets a ranking, and that ranking will start out at somewhere around #6,000,000.  We don’t know exact numbers because books are constantly being added and withdrawn.  The same is true for the number of Kindle editions.  There are around 900,000 of  them, but the numbers are growing every minute.

It is fairly clear that the rankings are based on number and frequency of sales, not on price. The frequency appears to be more important than the numbers.  If that were not so, the Bible would be #1 and stay there.  It's not, and neither is The Works of Shakespeare, although those two books have outsold anything you and I can ever hope to do. Instead, some think that the Amazon ranking depends on "What's Hot, Right Now?" 

A book that sells 90 copies within an hour will shoot right to the top of the best seller list, but it will have to keep on selling 90 copies an hour to stay there. As long as it does so, it will outrank a book that has sold 900 copies in the past 24 hours (that one was only selling at a rate of 37 per hour.)  And yes, that means that a clever author can manipulate the rankings by asking every friend he knows to order the book at noon on the same day. The trouble with that plan is that the author will run out of friends before Amazon runs out of books, and then the high ranking falls with a thud.

Some other factors enter into the equation as well.  Hourly and daily sales are much more important than weekly sales, and weekly sales are much more important than monthly ones. That's where the mysterious formulas come into play.  The rankings also seem to be affected by the number of reviews your book gets, the number of stars those reviews award, and, to a lesser degree, by the number of "likes" on those new buttons.

There are probably other factors as well.  No two books ever hold the same rank, so there have to be many ways to identify differences among  6,000,000 books. If you watch the rankings closely, you may see that Amazon responds to a high sales rate by putting the book on one of its recommended lists: "Best Books of the Year or Month," "Fall/Winter/Spring/Summer Reading Lists," "Best sellers in . . . ," "What's Hot," "Movers and Shakers." These lists go on and on, and they become cause rather than effect.  Many Amazon customers rely on those lists topic their books for them, so if Amazon lists your book somewhere, your figures jump, thus perpetuating your sales position, at least until Amazon decides to take you off the list.  Then sales drop, and your rankings follow suit, proving that the book should have been taken off the recommended list.

It's all an enormous game, which you need to remember before you get caught up in it.  You can manipulate the rankings to some extent by lowering the price of your book  Lower cost does translate into more sales at some point, although even the experts are not sure where that point lies. Free books, as I've demonstrated above, can turn you into a best seller, but don't plan on getting rich.