A Peek Into Amazon Sales Figures

By John Gilstrap
Since its creation a decade or so ago, amazon.com has provided one of the very few sources of sales feedback for authors. At a glance, we’ve been able to tell what our sales rankings are, as determined by whatever top secret algorithm they use to determine such things. If you’re obsessive about it, you can watch your numbers fluctuate wildly, from number 3,500 in the morning to number 350,000 in the afternoon. The trick is to decipher how rankings translate into sales.

As I write this on Thursday, January 13, the Kindle sales rank for my book Hostage Zero lies at 2,007—its first foray above the 2K mark that I’m aware of. Given that there are a bajillion books out there available to be ranked, though, one would think that that still represents a fairly robust sales velocity.

By contrast, the Kindle sales rank for Tom Clancy’s new book, Dead or Alive, sits at 28. Given the distance between the rankings—and the fact that we’re talking Tom Clancy—I wonder what that translates into in terms of actual copies sold.

As luck would have it, I’ve recently stumbled upon a website called novelrank.com, which uses algorithms of its own (or maybe just old fashioned spies) to monitor Amazon sales data and translate it into useable numbers. The results are interesting.

The Kindle Store has sold 73 copies of Hostage Zero thus far in the month of January, which earned it an average January sales rank of 1,718. Clancy’s Dead or Alive, with a month-to-date average sales rank of 21, has sold 253 copies in the same period. Last month, beginning with the book’s December 14 laydown date and a debut sales rank of 6, Amazon sold 493 copies of Dead or Alive from the Kindle Store. (I don’t have December figures for my book because I didn’t trigger the tracking function until January.)

For what it’s worth, the hardcover edition of Dead or Alive saw an average sales rank in January of 75 (it’s 105 now), with 1,010 hardcovers sold.

If the results gleaned from novelrank.com are reliable—and I have no way of knowing either way—my first impression is that the gross numbers seem low, though I admittedly have no basis on which to judge such things. The spread surprises me, too, with Clancy selling only 12-15 copies more per day than I do, despite a four-figure difference in ranking. Admittedly, that all must add up to a seven-figure delta between our incomes, but I would have thought his numbers to be several times what they are.

If nothing else, novelrank.com is a fun tool.  It lets you type in the title of any book, and if someone has already triggered the tracking function, you can see how it’s doing.  If they haven’t then you can trigger it yourself.  It’s a shame that my friend and blogmate Michelle Gagnon is weaning herself from the Internet, or she’d be able to see how her sales are doing.
Okay, that was mean. . .

13 thoughts on “A Peek Into Amazon Sales Figures

  1. John, I don’t know what formula they’re using, but it can’t be a very good one: I can tell you that the Clancy figures are way, way higher than that.

  2. Like I said, I have no way to judge the validity of the numbers. I’m as techno-challenged as they come, but I must say that the explanation offered by novelrank in their front matter projects confidence in their numbers.

    Neil, if Clancy’s numbers are way, way higher–and therefore my numbers are way, way higher–then God bless the math error.

    Kathryn, you just had to return the favor, didn’t you?

    John Gilstrap

  3. Last week I ran an ad for my ebook 65 Below and watched the general Amazon rank of the ebook leap from 120,000th in the morning to 4,000 in afternoon. Even more impressive was the in-genre leap from 30,000 to 4 in Kindle/War, where it stayed for most of the week.

    Looking at sales reports on Amazon’s DTP site it turns out that massive jump only meant 300 books sold for the week. That is a ton more than the 5-10 per week for December, but not as impressive as the ranking jump seemed to indicate. By week’s end, when the advert was long forgotten sales were back down to 2 or 3 per day.

    I think that when it comes down to it all this shows me is that I have to keep writing until I hit it bigger than this.

  4. I don’t know this for sure, but it seems like Novelrank totes up one sale every time your ranking goes up by a specific margin, so I’m not surprised Neil says that the Clancy figures are way off.

    However, I know that Amazon print sales are very low compared to in-store sales. I bet any author here on this blog sees a very small number on the Amazon line if their publisher shares detailed sales information with them. I know that was true for me.

    Kindle sales, however, are a different matter. Two years ago (has it been that long?) when I was self-publishing, I got to see my sales on a minute-by-minute basis. In Mar-Jun 2009 a ranking of about 1000 translated to Kindle sales of about 20 copies per day. A ranking of about 200 translated to about 75 copies per day. I’d guess those figures would be much higher now since the number of Kindles out there has exploded.

  5. Thanks, JaxPop. Kind words really matter.

    Basil, the wisdom of book advertising is a whole different topic. Every day on the subway I see the mug of some writer I know–whether David Baldacci or Brad Meltzer or Jeffery Deaver–staring out at me from a book advertisement, and I always wonder if the bucks spent on what has to be expensive advertising is really earned back in sales that wouldn’t have happened without the adverts. My guess is that the real intent of those ads is less to entice people to try a new author than it is to let the existing fan base know that a new story is available. If advertising is a science, it’s a damned imprecise one.

    Thanks for sharing, Boyd.

  6. Hey,
    Creator of NovelRank here. Clancy is selling a lot more, but the closer you get to a sales rank of #1, the harder it is to get accurate numbers. The sales are not derived from the sales rank figure itself, as Timothy Fish suggests, but rather the percentage change from the previous rank as well as the starting point (starting at 150,000 is different than starting at 10,000).
    Anyway, NovelRank also tries to estimate multiple sales in an hour as well. Check out the blog post on MYTHS and the FAQ page for some additional info.


  7. Novelrank is ridiculously inaccurate. I looked up their estimates my self-published Kindle sales of my out-of-print books and their numbers are off by more than 50%.

Comments are closed.