10 Responses so far

1. 1. Paul S. kemp on Apr 4th, 2008 at 9:33 am

   Very interesting data. Thanks for sharing, Tobias.

2. 2. Michele Marques on Apr 4th, 2008 at 12:02 pm

   Thanks for the charts.

   I agree about Ragamuffin being difficult to find. The Canadian box stores didn’t seem to have it in stock (I only tried them first because I received gift cards), so I bought one at a specialty SFF store (Bakka). It was just as well, as it turned out to be a signed copy.

3. 3. Mark Terry on Apr 6th, 2008 at 7:51 am

   Very interesting. Of course, as you say, without the numbers we’re only seeing trends. If one week you sold two books and the next week you sold three, depending on the Y axis, you could have a significant jump in the graph.

   Nonetheless, I find this extremely useful data (and too bad publishers don’t share it with all their authors). It’s pretty amazing to me to think that you might actually be able to chart a cause-and-effect with your mixed sales by comparing them to things you or your publisher have done (aside from Amazon rankings, which are, to say the least, a bit ambiguous).

4. 4. Fred Kiesche on Apr 6th, 2008 at 8:09 am

   I think we need to re-do the experiment. Find a week where you won’t get any award nominations or see any reviews and release another one into the wild…

   

   Awards, free giveaways, or whatever, great to see the sales go up!

5. 5. Michael Capobianco on Apr 6th, 2008 at 11:17 am

   This is very interesting, and I hope you’ll keep us apprised of how the books continue to sell. Looking at these graphs, what I see is a repeating cycle of increasing and decreasing sales, with Crystal Rain cycling on a weekly basis and Ragamuffin on a roughly monthy one. Could these cycles be caused by the way Bookscan collects data, or how Amazon (for example) reports sales to Bookscan? What else could cause sales to fluctuate on a regular basis?

6. 6. tobias on Apr 6th, 2008 at 6:09 pm

   Could these cycles be caused by the way Bookscan collects data, or how Amazon (for example) reports sales to Bookscan?

   I doubt it, the fluctuations don’t exist quite like that in the previous months, CR doesn’t start fluctuating until Feb.

   What else could cause sales to fluctuate on a regular basis?

   A preliminary Neb. nomination and Instapundit mention, and then a final ballot nebula, and then a Tor giveaway as I explained in the post The times line up, and with only a slow and minimally fluctuating line before hand shows a pretty strong case for it being what it was I said I think it was in the post.

7. 7. Michael Capobianco on Apr 6th, 2008 at 9:30 pm

   Well, perhaps those -are- the most likely explanations , but wouldn’t you expect a smooth increase and decrease over a longer time frame? Maybe I’m still tied into an analog world where it takes time for information to spread and people to re-act, when the reality is that the effect of these things happening is virtually instantaneous, and the drop off of the effect is just as rapid.

   In any case, I don’t mean to throw a spanner in the works. I salute you for tracking this information and making it public.

8. 8. Tobias Buckell on Apr 7th, 2008 at 12:22 am

   but wouldn’t you expect a smooth increase and decrease over a longer time frame?

   Having followed bookscan week by week for almost a year now, what I see are quick spikes when some major bit of exposure happens, above and beyond the normal squiggles of ups and downs. If people are compelled to by your book by hearing you were an award nominee, or because the Instapundit linked me, it seems to be an activity that happens for a large number of people within that week.

   Now others may be buying me after a longer period, but they’re in the background noise because I don’t see those bumps.

   You can see more of the effect you were expecting to see with Crystal Rain, and that’s why I was comparing the two. IE: CR has a more consistent ‘up’ level, probably more people run into in a store, plus it’s cheaper, but with the more expensive, and pretty much orderable only online Ragamuffin, it’s peaks are tied directly to the spikes of the prelim nomination and Instapundit link, and then the date of the final nomination, and then during Tor’s freebie.

   While it does take time for things to spread and react I don’t think you’re going to see a slow build up, but maybe a slightly higher low bar after the incidents (ie, sales might be slightly higher than January) that the graph returns to.

9. 9. Mark Terry on Apr 7th, 2008 at 8:01 am

   I wonder if the spikes occur because so much about books falls into the category of “impulse buy.” When I hear of a book coming out by a favorite author or some new book that looks interesting, I tend to act on that rather quickly… or never.

   Now, that probably falls into the category of regular book buyers. The book buyer that only buys a couple books a year probably doesn’t matter in terms of predictability (new Grisham? Oprah pick?).

   But for those of us who already have too much to read, I tend to say, “Oh, John Sandford’s next novels coming out May 1. I’ll order it on May 1 or pick it up that week at the bookstore.” Because if I don’t, well, the books typically don’t get bought until they come out in mass market paperback… or not at all.

   These spikes may describe book buying behaviors in general.

10. 10. Cliff Burns on Apr 25th, 2008 at 10:18 am

    Monsieur Buckell:

    Can’t tell you how much I appreciate your sharing this with other writers. I’ve just posted my novel SO DARK THE NIGHT on my blog in its entirety. I received mixed messages from other writers and folks in the biz: it would hurt me, it would help, depending on who you asked and if they’d had their first cup of coffee that morning. I’m encouraged by your graph (feh, who needs numbers, I’m a wordsmith fer Chrissakes). What wins out in the end is the quality of your work and, on that point, you clearly have nothing to worry about. Thanks for this…