The Ten Awful Truths about Book Publishing

Well, THERE’S an eye-catching post title. This is a post at Berrett-Koehler Publishers, which I saw via the Passive Voice blog — The Ten Awful Truths about Book Publishing. Fine, what are these truths and just how awful are they?

–-The number of books being published every year has exploded. [N]early 1.7 million books were self-published in the U.S. in 2018, which is an incredible 264% increase in just five years.  By 2019, the total number of books published in the U.S. exceeded 4 million in that year alone.

Yes, that isn’t so far an awful truth, just a truth.

–Book sales are stagnant, despite the explosion of books published. U.S. bookstore sales have declined 42% from their peak in 2007.

That’s admittedly not great. That’s a lot worse than stagnant. That’s falling off a cliff. Although bookstore sales doesn’t count ebooks, so … not sure this is actually super relevant to overall book sales.

Ah, here we go, this is more relevant:

–Despite the addition of e-book sales and downloadable audio sales, overall book sales have shrunk. Even adding in e-book sales and audio sales, the total book publishing pie has shrunk since its peak in 2007—yet every year it is divided among millions more titles because of the explosion of new titles and because most titles ever published are still available for sale.

But then we go back to excluding e-books!

–Average book sales are shockingly small—and falling fast. The average U.S. book is now selling less than 200 copies per year and less than 1,000 copies over its lifetime. According to BookScan—which tracks most bookstore, online, and other retail print sales of books (including Amazon.com)—only 690 million print books were sold in 2019 in the U.S. in all publishing categories combined, both fiction and nonfiction.

Look! This says “print sales.” It’s ridiculous to exclude ebooks and then draw conclusions about book sales overall. If you don’t count ebooks, you’re just pretending to address sales. [Also, they mean FEWER. Just didn’t want you to think I didn’t notice.]

–-A book has far less than a 1% chance of being stocked in an average bookstore.

I think most traditionally published fiction titles are, right? But not for very long, probably.

–It is getting harder and harder every year to sell new titles.

Could be, but I’m not convinced yet.

–Most books today are selling only to the authors’ and publishers’ communities. There is no general audience for most nonfiction books, and chasing after such a mirage is usually far less effective than connecting with one’s communities.

Bold is mine. This post is emphasizing nonfiction, as far as I can tell, though I imagine many of the broader points could apply to fiction as well.

–Most book marketing today is done by authors, not by publishers.

–No other industry has so many new product introductions.

I think I disagree with this definition of “product.” I wouldn’t treat each individual book as a new product. I think it would make a lot more sense to treat each individual author as a product. It doesn’t matter what cover or title a publisher puts on a book by Nora Roberts; all her books will sell no matter what the publisher does.

–The book publishing world is in a never-ending state of change.

That isn’t an awful truth. Like the first point, it’s just a truth.

I will say, that thing about the average US book selling fewer than whatever number of copies, well, even if we correct the numbers by including ebooks, “average” is a word that disguises an awful lot of what is going on. Do they mean mean or median? I hope median, as megabestsellers would haul the mean way far over one direction, while a ton of books selling in the single digits will haul it way back the other direction. Neither would be useful in seeing what’s really going on.

Also, it’s an obvious truth that a large proportion of all books published are on one playing field and the rest are on an entirely different playing field. I don’t mean self- vs traditionally published. I mean extremely terrible books vs readable-to-good books. The former should be discounted before asking how the picture looks.

Here are some actual numbers from traditionally published authors who are all clients of a particular agency:

Let me give you some real numbers from real royalty reports received by our agency without revealing the author name or the publisher (note the different genres and number of books):

Author 1: novelist – 3 books – avg. lifetime sales per title = 8,300

Author 2: novelist – 12 books – avg. lifetime sales per title = 19,756

Author 3: novelist – 3 books – avg. lifetime sales per title = 7,000

Author 4: novelist – 7 books – avg. lifetime sales per title = 5,300 (two different publishers)

Author 5: nonfiction devotional – 5 books – avg. lifetime sales per title = 10,900

Author 6: nonfiction – 2 books – avg. lifetime sales per title = 5,300

Author 7: novelist – 4 books – avg. lifetime sales per title = 29,400

Author 8: nonfiction – 3 books – avg. lifetime sales per title = 18,900

Author 9: fiction – 7 books – avg. lifetime sales per title = 12,900

Author 10: nonfiction – 5 books – avg. lifetime sales per title = 6,800 (three different publishers)

That’s all very well, in fact it’s quite interesting, but it raises a different question: What in the world do you mean by lifetime? If Author 1’s debut novel came out three years ago, then their numbers are 8000 or so copies per book over one to three years. If Author 2’s debut novel came out twelve years ago, then their numbers are about 20,000 copies over one to twelve years.

8000 copies per three years is 2600 copies per year.

20,000 per twelve years is 1666 copies per year.

For all we can tell, Author 1 is selling better, not worse, than Author 2! Lifetime is a stupid measure unless you define lifetime! Right? Or am I missing something obvious? It sure looks to me like the relevant measure MUST include time interval: how many sales per year, or how many sales per ten years, not how many sales per lifetime when at any moment, the lifetime of any random collection of books has to vary from less than one to over fifty years.

Regardless, the picture may still be bleak in some ways, I know. In a lot of ways, really. Authors Guild Survey Shows Drastic 42 Percent Decline in Authors Earnings in Last Decade. That was in 2018. The Authors Guild’s 2018 Author Income Survey, the largest survey of writing-related earnings by American authors ever conducted finds incomes falling to historic lows to a median of $6,080 in 2017, down 42 percent from 2009. This included traditionally, hybrid and self-published authors who have commercially published one or more books. When discussing median incomes, the survey looked at both full-time and part-time authors.

Now that’s bleak.

On the other hand … I’m surprised to see no pushback on this from The Passive Guy. I think it’s quite obvious that a rather large absolute number of authors, including a fair number of self-published authors, are earning some reasonable income via writing. Once you define the playing field to exclude the hordes of really terrible, unreadable books, then the proportion of authors earning some reasonable income is probably higher than it looks at first.

This reminds me in some ways of that famous post Slushkiller by Teresa Nielson Hayden. When you submit a query to an agent or a book to a publisher, you’re not competing with ALL the query letters or ALL the books. You’re only competing with whatever fraction of query letters are not terrible, or whatever fraction of books are not terrible.

I’m not aware of anyone who has ever worked out even the roughest possible average for book sales or author income AFTER excluding unreadably terrible books. But this would be, in a way, easy to do. I mean, this is a quick-and-dirty method that would lump some unknown percentage of perfectly decent books in with the unreadably terrible ones. In fact, really this is a division that isn’t about quality as such, though unreadably terrible books would almost all fall into just one of the categories, which is the point. This division is about cutting the world of available titles into two categories: Books That Aren’t Selling At All vs Book That Are Selling.

Just exclude all titles that haven’t sold a single copy in a whole year. Or the equivalent in KU reads, of course. Or you could make it six months or whatever you preferred. No matter what else may be true, books that aren’t selling at all are just not relevant data points if you are interested in sales numbers for books that are in fact selling some copies. Then you can cut both categories again: Nonfiction vs Fiction. Sales of nonfiction are not relevant to sales of fiction and vice versa. I think most writers who have the remotest clue, and who want to succeed, and who are writing readable books, can get almost any readable novel to sell enough copies to move from the first category into the second. That category is the one that actually matters.

Once you exclude books that aren’t selling at all, you can actually look at median copies sold per year for books that are selling or median income per book for novelists who are selling books and various other interesting data. Maybe that would make it easier, or at least possible, to draw conclusions that are somewhat valid.