Amazon Rank

For those of you who have not experienced one of your own, an Amazon Author Rank is basically what might be displayed in the dictionary beside “noisy” and “black box”. It ranks you against other authors on Amazon, using sales for all your books and possibly quantum entanglement from electrons on Mars, for all I, the mere mortal author, knows.

Since I’m a scientist at heart, I finally got tired of staring at all the noise and decided to poke at the data in a somewhat scientifically un-rigorous way. Mine are at least simpler than others’–I’ve talked to someone whose rank is influenced by used sales of his out of print books, which obviously doesn’t help him on the royalties or sales numbers with his publisher fronts. When I captured this data, I had two novels in print (print and ebooks) and one in preorders.

With apologies for the small font, since all the graphs are so long in the original size, here is my ranking, captured early January 2014:

(Yes, there’s no scale. It’s sort of beside the point for what I’m doing here. I’m comparing me to me, not me to J.K. Rowling.)

You’ll see I’ve started out by annotating it with the “duh” pattern. New releases make the ranking jump, but new editions don’t do as much.

This is after curve smoothing. New releases still seem like the biggest drivers, but that peak on November 4, 2012 lingers. I’ll note that in all of this, I chose to investigate peaks rather than troughs–after all, as I understand it, even if my sales plateau rather than drop, some other author’s gangbuster sales can drop my rank. Besides, peaks seem much more likely to be linked to an actual event/action on my part.

So November 4. What’s that? Well, examining my schedule, that was the World Fantasy Convention. Neat, huh? Only…that doesn’t work for any of my other conventions.

We’re missing one on that graphic, but only because Miscon 2013 was the week after Tarnished released, so any patterning is indistinguishable from the new release bump.

So do conventions improve my rank? (Or my sales, for that matter–the link between rank and sales is not completely clear, so better to think of them separately). Looks like I’ll have to wait for more data.