From veteran publishing consultant Mike Shatzkin:
When publishers decide whether or not to buy a book, or look back at a book to evaluate its performance, the analysis is done in a way that could be, and often is, misleading. This is almost universal practice, has been for at least a century, and may never change. But it is worth a few minutes for anybody interested in understanding the profitability of a publishing house to contemplate what seems to me to be rampant misunderstanding.
The problem is the practice of constructing P&Ls on a book-by-book basis. The concept itself is a logical fallacy. The idea of an individual book making a profit or loss only makes sense if there is no publishing house. That is, if you decide this afternoon to take time off from your career as a truck driver or a banker and invest a little cash in publishing a book, your exercise could at some point be measured and a profit or loss could be calculated.
Doing that would be very straightforward if you were doing it on a cash basis. You’d add up all the revenue you got from publishing the book and all the expenses you incurred in publishing it and pursuing that revenue, mash them together, and have your calculated profit or loss. You would have to account for unsold inventory if you didn’t use a print-on-demand strategy. You might have net positive cash (profit) or net negative cash (loss) and some unsold inventory (potential additional profit) that you might have to pay storage fees on (potential additional loss).
Standard accrual accounting methods would call that unsold inventory an “asset”, essentially adding to your “profit”, but that would only be true if you could actually sell it.
. . . .
But if you decided to grow your little operation and do ten books, even on a cash basis the accounting for each book now wouldn’t be quite as simple. Your production or marketing or sales team works on your whole list and how much of their time should be “allocated” to each book might be hard or even impossible to calculate. If you were being precise about it, you’d have to account for the reality that the books are not all the same. They take different amounts of effort to develop editorially and do not place consistent demands on your production and distribution overhead costs. You couldn’t actually just “add up” all the expenses for each book to subtract from the revenues to produce a profit calculation.
And imagine how much harder it would be to be precise about assigning those costs if you were dealing with hundreds of books in an organization each year. Or thousands.
Because many publishing decisions are made one book at a time and because accounting is done to the government to pay taxes and (sometimes) to shareholders as well, it is common to do the accounting on a “per title basis” and ultimately on a “per unit” basis (where we aren’t just trying to understand the profit — or loss — by title, but literally score things down to the individual unit transacted). Title P&Ls and unit cost accounting are part of the operating fabric of every large publishing house.
I’d argue publishing would work better if they weren’t.
. . . .
Here’s how book publishing economics actually works. A publishing house has overheads which are reasonably fixed: primarily rent and salaries but also including travel and entertainment, insurances, legal and accounting, and the costs all businesses have to keep operating and keep their doors open. Unless there is some conscious expansion or contraction of the publishing program, those expenses do not fluctuate appreciably based on the number of titles a house publishes or the revenues it generates from selling books and rights.
Then each book has two kinds of costs: the investments required to publish it at all (author’s advance and what used to be called design and typesetting but which would now be better described as “creating a print-ready file”) and unit production costs, the “paper, presswork, and binding” of the actual printed units. There is no unit production cost for ebooks.
When books or rights sell, the publisher banks a “margin”. For rights, that is all the revenue not paid through to the author. For book sales, it is “contribution margin”, the difference between the revenue the publisher receives from accounts and the actual direct costs required to complete the sale, which for most books requires subtracting the unit printing cost and any incremental sales commission and royalty due to the author (if the advance has earned out). The margin earned on each book has to “pay back” the book’s direct investments first but then gets applied to address overhead.
When the net positive margin generated by all the books, frontlist and backlist, in a fiscal year have covered the house’s overhead, the additional margin constitutes the house’s profit. One might say the book’s “profit” is the margin it generates, but no publishing house I know looks at it that way.
Instead, the standard practice is to assign each book its “share” of the house’s overhead. (Sometimes this is made even more complicated by assigning different overheads to books from different reporting units — imprints — within the house.) How the number to apply is calculated in each house is not transparent, and almost certainly varies, but the simplest form would be for the house to calculate what the fixed overheads were last year in relation to total sales and “allocate” each book that percentage of overheads. (The overhead number often ends up being 35 to 45 percent.) That overhead allocation pushes many, if not most, books from being scored as”profitable” to being calculated to be “unprofitable”.
. . . .
Before the rise of indie publishing enabled by Amazon, it was much easier for the big houses with their big sales and distribution capabilities, to be sure they’d get thousands of copies out on just about every book they did. Now it happens — and it really didn’t back then — that even a big house can have frequent abject failures: books that don’t even recover their direct costs (even without a massive advance against royalties). That was a much rarer event in bygone decades.
Link to the rest at Mike Shatzkin and thanks to Jan for the tip.