So long as the data can be found (sadly, the AAP no longer does their own press releases; thankfully, other sources with access to BookStats do post the monthly ebook numbers) I’ll continue to update the graph.
I’d also like to remind the folks at AAP/BISG/bookstats.org that posts to RocketBomber.com are available freely under a Non-commercial CC license: they are welcome to include my analysis (or adaptations of my analysis) in their own reports. I’ll even waive the non-commercial requirement (which would be a sticking point, since they charge big bucks for the data now) so long as they still include an attribution. You share with us, I share with you, new things get created, new analysis and viewpoints proliferate: internet!
Those of you new to this blog should check out:
The Original Projection: June 2011
And Update 1: September 2012
The formula for the sales projection curve is
ebook(t) = k * (1 + tanh(-π+((π/80)t)))
Where t is the time variable, counted by months, and k is a constant one selects out of one’s ass (a surprising number of scientific constants work that way) equal in this case to $130 Million. The constant k is also the assumed value of ebook sales at the inflection point in the graph.
Using fractions of pi — (π/80) above — is how we “stretch” the s-curve to match the observed growth over time. My first projection used (π/60), an assumed dynamic growth phase of about 120 months. To get the projected graph to match reported sales, however, I had to slow things down a bit — (π/80) translates to a “dynamic” phase of 160 months, about 13 years. For the graph below, our starting point (t=0) is the month of August, 2005.
This also means we hit the inflection point back in May, 6 months ago.
My [*] on the Projected Sales is the same disclaimer as last time:
- The only data available to me are ebook sales as reported by the Association of American Publishers: so these correspond only to US ebook sales from established publishing houses and does not include self-published ebooks.
- Merely looking at a dollar sales figure (again, the only data available) glosses over the fact that ebooks are sold at lower price points: unit sales of books will be higher than the dollar figure might suggest
- My projection is not the only interpretation – but I’ve tried some other models and ebooks sure look like they’re following a fairly common sigmoid growth curve
- …however, if ebooks do not merely cannibalize sales of other formats but instead push books into new genres, new business models, new retail channels, and effectively blow up books as we know them: why sure, I guess there’s no upper limit & my projection is wrong. You can make any assumptions you like along those lines. My graph represents a fairly short future time frame (3-5 years out) and a relatively stable publishing industry. (Well, stable other than the disruption currently happening due to ebooks.)
As data becomes available, I’m happy to post future updates. I think the next e-book sales projection will be after the Jan-Feb ebooks sales numbers go public — covering this fall plus the post-holiday-ereaders-given-as-gifts bump we’ve seen in years past. I’m guessing that will be in May, about 5-6 months from now.