Monday, October 15, 2012

A New Sort of Publishing Model

by Stephanie Osborn
with Dr. James K. Woosley

There's been a lot of discussion lately about the publishing industry as a whole and how it is or isn't growing, and whether writers are or are not competing. I've shared that discussion with friends and family, debating both sides of it. In general, I think that, especially with small presses and indie presses springing up, the growth is actually good and a positive thing. In theory we have a tremendous growth capability. In the short term, however, and in practical terms, I'm not so sure it isn't like saying we're going to fly to the Alpha Centauri system next year because some scientists have managed to develop a quantum-scale warp bubble in the lab. Theoretically it's possible, but practically, there's a few bugs yet to be worked out.

My friend Dr. James K. Woosley is a physicist and a Heinlein essayist; you'll meet him in more detail next week. But he's a hard-core, long-time science fiction/fantasy/specfic fan, and one of the friends with whom I've been discussing this issue. He's also skilled at modeling systems, down to the quantum level. Leave it to him to come up with a way to quantify this...


“It's partially an application of what I call a Sturgeon filter (because if 90% of everything is crap, then 10% of everything is worthwhile, and successive application can get to fairly decent estimates of demographics, as in 10% of people are readers, 10% of those are readers of SF/Fantasy literature, 10% are readers of more successively refined tastes, etc.

“Still, your potential reader base is X. Any individual author's actual reader base is a fraction f of that, or total number fX. Your base X has a total amount of disposable income C, and an average amount C/X, which is used to purchase reading material. The number of authors competing for that money is A, and the average number of works each author has for sale is n (where n is some conglomerate of new press and backlog, which complicates the analysis), such that the total competing number of works is N=nA. The average price of any work is P.

“Total individual items sold is thus C/P.

“Average sales per item is C/NP.

“Sales per item can be assumed (first order) to be Maxwell-distributed [an asymmetrical bell curve with one side steeper than the other --Steph] about that mean.

“The second constraint is time to read each book, moderated by the number of people who will buy attractive books that they want to read but lack the time to read, but also controlled by the fraction of people who re-read books. Assume an average reading speed (including breaks) of 15,000 words per minute or 6 hours for a typical novel, and assume further that the average reader will read one novel per week or roughly 50 novels per year, if they have the resources, but that an adult steady state reader will on average re-read 25 novels per year, so that new novels account for half of their reading.

“Assume that the book stockpiling factor is also 2. (Speaking for myself it's probably closer to 290, but...)

“The bottom line is that the non-cash-constrained market is about 50X novels per year, so that average sales computed that way is 50X/nA (where, again, n is some conglomerate of new press and backlog, which complicates the analysis). Or more simply, Sales(Avg)=50X/N.

“We'll assume cash parity initially, so that C = 50XP.

“I think a good working assumption is that X=1 million for SF/fantasy, so that about 50 million genre books are purchased each year. That is probably a maximum.

“At any time, A is about 500, with n=10 books (current plus backlog) in active publication.

“That yields a Maxwell-distributed average of 10,000 copies per book sold.

“At that level of the analysis the game is zero sum. The situation changes when either the fan base is increased, or disposable income increases so that people are more willing to stockpile books.

“And that's about it; danged if I know what if anything the analysis buys.”


Note that that's an initial analysis, and that the AVERAGE number of copies per book is 10,000. That isn't the best-seller numbers, and it isn't the bottom of the list numbers. This is how many books the mid-listers should be selling a year.

Food for thought.

~Stephanie Osborn