Bloggy Bits

Are you prepared for The Big One?

Nov 12, 2010   [permalink]

Ebook Sales Curves — Sssssssss Hot!

Talking about the article saying Stephen King reading half digitally lead to a question about hard data on ebook sales, so I found that the Association of American Publishers reports ebooks are now over 9% of sales. They also present a chart showing their annual growth data from 2002, in dollars and percent of sales. It very much looks like the left/early side of a classic adoption S-curve.

I thought it would be interesting to curve-fit the AAP's data to an S-curve see how it fits, and see where it projects it's going. I remember reading somewhere a couple months ago that the major publishers thought ebooks would exceed 50% of total sales no later than 2015. I seem to recall they used the word "conservatively" in that statement.

See, this is an exponential function. It grows really fast. The early numbers may look small, but the year-over-year growth rate is still high. There is a significant increase in each of the years -- the numbers double or triple each year.

Those are the real data points. Note the nice exponential shape. Percent of sales going up, years since 2001 going across.

It obviously can't keep doubling or tripling. Only until you reach around the 70%ish mark; then it slows and eventually closes in on 100%. Adoption of products tends to follow the Pearl curve aka Logistics curve aka S-curve. It's a well known curve, and usually pretty accurate.

For grins I decided to do a regression analysis on the AAP's data points, fitting it to an S-curve. The r-squared correlation coefficient came out to an incredible 0.9677. (1.0 is a perfect fit. 0.9677 is an extremely good fit.)

Here's the projected curve. Year across the bottom, percent of book sales going up.

The predictions from this curve fitting are pretty astounding: Ebooks sales roughly tripling again next year (so 9% of sales to around 27%). By the next year ebook sales exceed 50% of sales, roughly projected around 2/3. By the year after that, ebooks could be as much as 90% of sales.

The sources of error in this curve-fit come mostly from the fact that we only have annual data, and from the fact that a lot of the fit comes from the early, small potatoes numbers. Quarterly data would be more useful. And of course human nature isn't a perfect curve fit; economic factors, luck of the draw on what ebook devices are out there, etc. all contribute.

But the data suggest that ebooks could overtake print book sales and become dominant within a very small number of years. Exponentials are not to be trifled with.

"That's FAST!" you say. Yes, well, let's not forget that DVDs overtook VHS in only three years.

It's always scary to make predictions, but this has all the signs of being "the big one."

True or not, sort of like Pascal's Wager, it makes sense to plan as if it might be so. Are you ready? :)

[ comments | add a comment ]