Some Number Crunching on Career Sales, 2024

Tucked into yesterday’s announcement about the new 10-book contract with Tor is this notation:

John Scalzi has become one of the most popular science fiction authors of his generation with 4.5 million books in print in the United States. 

Let’s discard the “one of the most popular science fiction authors of his generation” bit as PR fluffery — which is easy for me to do since I wrote that bit — and focus on the “4.5 million books in print in the US” bit. What does that figure actually mean, in terms of my real-world sales?

Well! Let’s get nerdy with this stuff, shall we:

1. “In print” does not mean total sales. This much should be evident when you go into a bookstore and see copies of my books on the shelves. If they’re on the shelves, axiomatically, they haven’t been sold yet. I’ve published 17 novels with Tor since 2005 and none of them are out of print, so at any one time there are (at least) tens of thousands of physical copies of my books in stores, waiting to be purchased. If they aren’t sold and the bookseller wants the space for something else, they go back to a warehouse owned by Macmillan (Tor’s parent company), where they either go back out again to a different bookseller, remaindered (i.e., sold at a steep discount at the end of a publishing run), or pulped.

Back in the day it was not wholly unusual for a publisher to print a large number of copies of a book, have the sales be a modest fraction of the print run, and then remainder or pulp the rest of the print run, simply as the cost of doing business. These days, however, print and distribution technology are sufficiently advanced that print runs rather more precisely correlate with the number of units sold. I was recently offered the chance to buy remaindered copies of one of my hardcovers (it had moved into the paperback phase of its life), and the total number of remaindered copies available was… a few dozen. That’s a pretty impressive estimation of a print run on the part of Tor.

(Additionally, “in print” means something different for ebooks than for physical books. For physical books, it means actual printed copies of a book; for ebooks, it means having a digital file available for sale on a server. The correlation between “copies in print” and “sold copies” for ebooks is very nearly one to one.)

My professional life did not overlap the “print a bunch of copies” era of publishing; Tor has always been fairly conservative in my initial print runs (not stingy, merely attentive to previous sales and estimating from there) and then would do additional print runs when necessary. Also, ebooks have been a significant percentage of my sales — there’s a reason I more regularly hit the New York Times “Combined Print & Ebook Fiction” best seller list than its “Hardcover Fiction” list. What that means is I suspect my “in print” and “total sales” numbers are fairly close… as they would be for most authors these days.

So, no, I have not sold 4.5 million books from Tor in the US, given both the current number of printed-but-unsold copies on bookstore shelves, and the print run overages since 2005 that were either remaindered or pulped. I suspect of those 4.5 million, I’ve probably sold 4.3 to 4.4 million of them. Which is still, you know, okay.

2. I’m going to use the “4.4 million copies sold” estimation for this next part:

Over 17 novels and 19 years, 4.4 million averages out to roughly 259,000 copies sold per novel, and roughly 231,000 copies sold per year. Which does suggest that what Tor currently advances me per book on average ($261,000, going up to $300,000 when the new contract kicks in) is… pretty spot on in terms of my sales profile. This is not, you should know, horribly surprising: Tor knows how many books I sell for them, and the general rule of thumb for book advances is that publishers will advance you an amount that they think you are likely to earn over the life of the book, because writing actual royalty checks/sending royalty direct deposits is a bit of a pain (Update, 8pm: I meant to write “life of the hardcover” and had a brain fart, sorry). So, on average, my publisher hits the mark on my advances. Well done, Tor’s accounting department!

The reality of copies sold per novel and how many copies of all novels sold per year is, of course, much messier. I regret to say that I did not sell 230,000 copies of Old Man’s War in 2005, even though it came out on January 1st and I literally had the entire year to register sales. I sold, at best, 7,700 copies, because that was the entire run of the hardcover (and OMW didn’t initially have an ebook; the first Kindle wasn’t released until 2007). On the flip side of this, the Tor press release notes that each of my last three books, The Last Emperox, The Kaiju Preservation Society, and Starter Villain, respectively, “have each broken [the] previous record for the fastest-selling Scalzi novels of all time.” So my sales velocity, I am happy to say, has been increasing as I go along.

With that noted, it’s also no secret that the majority of my sales are in my backlist. Old Man’s War may have only sold 7,700 copies at most in its first year, but once it hit paperback in 2006 sales really took off… and have remained essentially unchanged, year in and year out, for almost two decades. It is unquestionably my best selling title across my entire publishing career. This in turn feeds the perennial sales of the rest of the OMW series, and when the seventh book in the series is released in 2025, we’ll likely see another sales bump across all six previous books in the series.

I now sell well out of the gate, which is fantastic — but selling backlist, i.e., where the expenses for the publisher are already priced in and now the title is generating pure profit, is what makes them happiest of all. Again, all of my novels for Tor are still in print and are still selling briskly, some more than others, but all well enough. Those annual backlist sales figures are likely to climb the more books I write for them.

So, no, not every novel of mine has sold the same, and not every year has seen the same amount of sales. “Average” is not useful here. But when it’s all added up it’s pretty clear (to me, anyway) that I’m compensated fairly by Tor in terms of advances, and that Tor, in turn, is getting a fair return on what they pay me up front.

3. It’s also important to note that the numbers offered above by Tor represent the floor for my sales (and print runs), not the ceiling. For example, the majority of my audiobooks are published by Audible, and those sales numbers aren’t part of Tor’s estimates. I’m happy to say that Audible has done very well by me in terms of marketing and sales (heck, they made a TV ad for Starter Villain) and that my sales in audio, especially for recent novels, are directly comparable to my sales in print/ebook.

Likewise, foreign language sales are not covered by Tor’s numbers, because the company doesn’t publish me in any language other than English. Most novels of mine are picked up for translation, and while each individual edition in a foreign language may or may not rack up huge sales (it’s difficult to shift a massive number of units in, say, Latvia, which has only 1.8 million citizens total), in aggregate they can add up nicely. At this point my work is available in three dozen languages, a fact I am continually delighted with (thank you, my foreign language publishers!).

There’s also my non-Tor writing, which includes but is not limited to non-fiction titles, novellas and specialty titles mostly released through Subterranean Press, collections, anthologies and individually-sold short stories, and so on. Fun fact: For a while there, my best-selling book was not one of my novels but Book of the Dumb, a 2003 humorous collection of stories about real-life people doing regrettable things. It sold 150,000 copies the first couple of years it was out. You can still get it (but why), so those sales numbers are probably still ticking upward very slowly. The SubPress titles are mostly limited editions in print but are available indefinitely as ebooks, and I’m happy to say some of them, particularly the “Dispatcher” series, are chugging away quite nicely.

Further muddying the waters in terms of “sales” are the titles that are available as part of a subscription service. For example, the audio versions of the “Dispatcher” novellas are part of Audible’s subscription service, so people who are Audible Plus members can listen to them for free, although they are also available for sale as well. They’ve sold enough to be New York Times bestsellers in the Audio Fiction category, and I know they’ve been downloaded hundreds of thousands of times, but I’m not entirely sure what the ratio between sales and subscription listens is. Likewise Amazon sent me a nice little trophy commemorating the fact my short story “Slow Time Between the Stars” got a quarter of a million readers, but it’s available as a free read for Amazon Prime subscribers. Do those quarter of a million readers count as sales? As the story being “in print”? Who knows! I got paid for all of them, which is the important part.

Maybe they and sales could be covered by another umbrella metric: “published reads,” in which work for which I was paid by a publisher was then accessed by a reader, who paid for it either directly (via sales) or indirectly (via subscription). A “published read” metric would then also cover much of my foreign-language audio, which is accessible via subscription services rather than available for individual sale.

4. Add all of this up, what does it mean? Again, mostly that the 4.5 million “in print” figure offered by Tor is a floor, not a ceiling. Without torturing my agent by making him collate every single foreign print run and sales figure, or digging out my own records on pre-2008 non-fiction titles (for whom I had a different agent, who has since left the business), I would conservatively estimate my total sales figures, in all genres and formats worldwide, are close to double the Tor “in print” estimate. If we throw in subscription readers and listeners into a “published read” figure, it goes up another million or so, easy.

So, maybe 10 million in total sales/published reads over nineteen years. For various reasons, that feels reasonably accurate to me.

That’s… pretty good. And so long as I’m not eaten by a bear/hit by a bus/contract brain worms, etc, there will be another fifteen years, at least, before I turn in my last novel and coast on royalties into retirement. More than enough time to add to those numbers.

I’m going to get to that. Uhhhh, after I get back from this year’s Worldcon.

— JS