How book publishers make or lose money

TOR employee Anna Louise gives a vivid portrait of the inside workings of a fiction publisher and how they make or lose money on individual books.

In part I, she writes about a complete mass-market failure:

The publisher tells you to get some in house reads, because she isn't sure this is a wise use of resources -- without blurbs, you're going to have a hard time. Plus, you're on your second cover -- the art department just can't get it right. You spent $4,500 hiring an artist. Now the art director is working on the cover himself, using stock art. You still have to pay for stock art -- it costs $1,400.
[...]
On the initial profitability and liability statement, the excited, committed editor theorized it would print at least 50,000 and sell at least 30,000, and paid the author an advance of $12,500. She didn't want to go all the way up to $16,000, just in case she was slightly off the mark -- and normally a first time author would get something like $5,000 (just in case! and also leaving room to grow!), but this was out with four other houses, and the agent had a $10,000 offer from NAL, and, damn it all, the editor really wanted it, so her publisher let her pay an exorbitant amount.

Oops.

In part II, she writes about a successful hardcover book and its move to the mass-market:

10,000 copies. Everyone gets a 50% discount off the cover price of $24.95, which means we're starting out with $12.48 per copy.

The author gets the standard 10% through 5,000 copies sold, and 12.5% for the 5,000 after that, and 15% thereafter. This means the author makes $2.50 per copy through the first 5,000 copies sold, and then $3.12 per copy on the next 5,000 copies sold. This particular project sells 8,000 copies, which means Rygel makes a total of $21,860 on the hardcover sales alone.

Part III still to come.

The fascinating thing is that book failures are completely normal for publishers, and not a black mark at all. (If you're a first-time author, on the other hand, failure to sell will turn your name into mud.) Publishing seems to be like professional gambling: what matters is the long-term success, not any individual wins or losses. Of course, a big enough win or loss will make a big difference to whether you get promoted or fired, but in general, it expected that editors will promote their share of turkeys. There are so many things are completely out of the publisher's control that nobody cares when a book fails to sell -- except the author.

That's not to say that publishing is completely random. Like a gambler, the skill is in doing your sums correctly so you know how much risk you can afford to take for the expected payoff. Unlike gambling, where the payoffs are usually known in advance, publisher's sometimes stumble accross a book or two that the public falls in love with, and then there is no limit to the payoff they can get.