There are any number of statements about numbers, including those that cite “lies, damned lies, and statistics,” which was a statement Mark Twain attributed to Disraeli, but which appears in none of Disraeli’s written statements. And then, there are statistics that are accurate, but which misrepresent when applied to smaller segments of whatever’s being characterized by those statistics.
Several weeks ago I was talking to my editor, who read me something from some publishing statistics that indicated that ebooks now represent about thirty percent of all book sales, but the rate of increase in ebook sales has supposedly slowed. After thinking about this for a while, I decided to analyze my own royalty statement [and given the way the figures are presented, it does take a certain amount of raw mathematical number crunching before one can analyze, because the figures are broken down book by book]. If I’m at all remotely representative of the F&SF field, that thirty percent number is way off for fantasy and science fiction books, since for my last royalty statement, sixty-five percent of my sales were in ebook format, and if one eliminates new releases the percentage is even higher.
In terms of revenue, especially for new releases, however, the story isn’t quite so clear. For new releases my sales in the first year are around sixty percent hardcover, and forty percent ebook. In addition, on average, I receive about sixty percent more for a hardcover than for an ebook – and that’s for the initial $14.99 ebook. So while ebooks are a good deal for buyers, even at the higher initial price, they’re anything but a good deal for the author in terms of new hardcover releases.
In the case of backlist books, though, the calculus reverses, especially in my case, because my backlist is so extensive that no bookstore, even the F&SF specialty stores, carries anywhere close to a significant percentage of my backlist, which means that readers can easily purchase ebook versions of books that are difficult, if not impossible, to obtain in print versions.
Generalizing from a sample of one is extremely risky, fraught with danger, and often highly inaccurate. Even a sample of seventy books [roughly the number of separate titles of mine in print – including omnibus editions] is an incredibly small sample, given the millions of books out there. And, on top of that, I have to admit that I’m probably not the average F&SF writer in terms of sales, because I have a pretty substantial backlist, and quite a few books on that backlist are hard to obtain in print format, which will pump up the ebook numbers to some degree. But other authors also have titles that are hard to find in print, and when my numbers come out at twice the supposed industry average, I have to suspect that what’s happening is that the sheer volume of cookbooks, how-to books, and other “genre” books that don’t lend themselves to ebook format or whose readers aren’t as interested in ebooks, for whatever reasons, are overwhelming fiction numbers, and especially F&SF numbers.
I don’t doubt the statistics, but I do doubt their applicability to fiction, and especially to F&SF, and that illustrates the danger of applying “industry-wide” statistics to a sub-set of an industry, because using correct, but misrepresentative statistics… well, that tends to fall into the category of statistics that Twain was describing.
Nonetheless, the numbers I’m seeing personally suggest that brick and mortar bookstores specializing in fiction are facing a very uphill struggle to survive… unless the present trends slow or change rather dramatically… or unless I’m incredibly unrepresentative.