Book sales: What that viral Substack post gets wrong.

This essay was adapted from Lincoln Michel’s newsletter Counter Craft. Subscribe here.

Last week the article “No One Buys Books,” by Elle Griffin, went viral, topping Substack categories and being shared widely on social media. It’s easy to understand why. Publishing is an opaque industry, and Griffin’s piece—which collects quotes and statistics from the 2022 Justice Department suit against Penguin Random House, in which the government successfully blocked PRH’s $2.2 billion purchase of Simon & Schuster—is filled with shocking claims. Over 90 percent of books sell fewer than 1,000 copies; 50 percent of books sell fewer than 12 copies. The article paints a nearly apocalyptic portrait of traditional publishing, in which nothing works, few make money, nobody reads, and the whole industry might go poof at any moment. This vision is appealing to many people, including writers who (fairly or unfairly) feel stymied by the industry, popularists who think reading itself is a snobby hobby, and tech types who mock “paywalled dead trees.” The only problem is, the picture isn’t true.

Let me say first that Griffin does an admirable job collecting these quotes, and much of the overall thrust is correct: Most people don’t buy books, sales for most titles are lower than many think, and big publishing works on a blockbuster model where a few hits—plus perennial backlist sellers—comprise the bulk of sales. It’s also true that books don’t occupy as central a place in the culture as they did before film, streaming TV, video games, and social media fragmented culture. (That ship sailed many decades ago, however.)

Still, many of the statistics here are wrong or misleading in part because they were intended to be. PRH’s legal strategy was to present publishing as an imperiled, dying industry beset on all sides by threats like Amazon—a characterization intended to convince the court that a merger was necessary for publishing’s survival. There may be truth to that idea, and I’m not saying any of the quotes are lies. I’m saying the quotes and statistics are tailored toward a specific narrative in the context of a legal battle. So, read them with a $2.2 billion–sized grain of salt.

Before we dive into numbers, though, let’s step back and look at the biggest question: Do people buy books?

How many books are sold in the United States? The only reliable tracker we have is Circana BookScan, which logs point of sale—i.e., customer purchases at stores, websites, etc.—for most of the print market. BookScan counted 767 million print sales in 2023. It collects this data directly from reporting bookstores, chains, and online retailers that combine for the majority of the market. BookScan claims to cover 85 percent of print sales, although many in publishing think the percentage is smaller. Many books sold at book festivals or in-person at readings are not reported to BookScan, for example, nor are library sales. (Almost every author will tell you that their royalty reports show notably more sales than BookScan captures, sometimes by orders of magnitude.)

Still, I’ll be conservative and assume that 85 percent is the correct figure. This means that close to 900 million print books are sold to customers each year. Add in e-books and the quickly growing audiobook market, and the total number of new books sold is over 1 billion. Again, this is a conservative estimate. And it does not include used book sales or library purchases. Americans do buy books.

Is 1-billion-plus a lot of books? This depends on your point of view. For comparison’s sake, there were 825 million movie tickets sold in the U.S. and Canada in 2023. So, more books are purchased than movie tickets, two comparable entertainment options in terms of price. On the other hand, that’s “new movies in theaters” vs. “all books available for sale from throughout history.” Movies are vastly more expensive to make, so far fewer movies are released each year than books are published. Individual movies are watched far more than most individual books are read.

Whether or not this “1 billion books” number is big, it’s fairly stable. Print book sales have not been decimated by digital sales/streaming in the way of so many other industries. Despite the introduction of e-books, various attempts at “Netflix for Books” services such as Oyster and Scribd, and endless cries about the death of publishing … print sales have increased in recent years, before even accounting for the millions of e-books sold and the audiobook market. Publishing is an industry with plenty of problems—believe me, I know—but there’s no reason to think it’s on its deathbed.

In the world of lies, damned lies, and statistics, book stats are especially easy to manipulate. I’ve already mentioned that BookScan doesn’t track all print purchases and that the stats typically don’t include e-books or audiobooks. But there are lots of ways—if, say, you are trying to persuade a judge to allow a big corporate merger—to manipulate these numbers to mislead:

Many of the quotes from the trial seem to reference calendar year sales—meaning they include books that have been out for only a week or a month—instead of a reasonable time frame like two years, much less lifetime sales.

Overall book sale stats apply to all books from throughout history that are still in publication. A bestseller from 1924 or 1824 might sell only 10 copies in calendar year 2024, but it might have reached 1 million in lifetime sales.

BookScan tracks by ISBNs, meaning what most people think of as a single book—e.g., one novel—actually counts as multiple books. My novel The Body Scout came out in four formats with different ISBNs, and so it counts as four separate books: hardcover, paperback, e-book, and audiobook. The same is true for almost every book published.

Certain types of books, like some academic texts, are barely sold in the general marketplace at all, even if they have an ISBN.

“Books” as a category includes everything from novels and memoirs to crossword puzzle collections and adult coloring books.

“Publishers” includes everything from big corporate publishers to small presses to academic presses that don’t really sell in bookstores to serious self-published authors to amateur self-published authors who may have no expectations at all of sales but just want to put a collection of cat poems out for their grandkids.

I hope you can see where I’m going with this. What the average Big 5–published book sells in one calendar year is a completely different number from what it sells over a lifetime. That number is completely different from the average book sales of all publishers, from micropresses to big corporate publishers. And that number would be completely different from the average sale of books, including the millions of self-published books.

And within any of those numbers, the statistics vary wildly by type of book. Think about movies again. When people talk about movie revenue, they’re probably restricting themselves to feature films released in theaters or movies streaming on big platforms like Netflix. (The closemouthed way the latter handles its data has already famously mucked up movie stats.) But “books” encompass everything from niche academic monographs and Sudoku puzzle collections to bestselling novels and memoirs. It would be as if “movie” statistics included educational videos, online porn streaming, and TikToks.

My point isn’t that there’s anything bad about self-published cat poems, crossword puzzles, TikToks, or anything like that. My point is that “average book sales” tell you almost nothing about, say, how well the average novel published by a big press, and widely available in bookstores, sells. (To be fair, this also means that the billion-plus number above includes lots of things we wouldn’t normally think of as books.) I went into greater depth about book stats here.

There’s also a lot of confusion in book publishing because the economics are both complicated and obscured. Confusion is common. I’ll look at just one example from the original article, because it illustrates two common misunderstandings:

If I look at the top 10 percent of books … that 10 percent level gets you to about 300,000 copies sold in that year. And if you told me I’m definitely going to sell 300,000 copies in a year, I would spend many millions of dollars to get that book.

Publishing houses pay millions of dollars for a book that sells only 300,000 copies??? Well, because books don’t sell a lot of copies, they don’t make a lot of money. …

According to [consultant Nicholas] Hill, 85 percent of the books with advances of $250,000 and up never earn out their advance.

McIntosh said 300,000 copies “in a year” rather than in lifetime sales, so we are talking calendar-year sales, not lifetime sales. (See above.) A book that sold “only” 300,000 in one year is going to sell many more copies in other years. It has a good chance of being a perennial backlist seller, even. That is the goal of any publisher, since they sell with almost no overhead or marketing costs. Such a book would certainly make many, many millions for a publisher. What company wouldn’t pay millions if they were “definitely going” to make even more millions back? Business 101.

I think the above quote also might give the impression that books “don’t make money” when they don’t earn out advances. This is a common misconception. A book can be profitable for a publisher well before the author earns out the advance. (If you don’t know, an advance is money paid upfront to an author. Authors don’t earn royalties until a publisher has made back that advance money out of the author’s per-sale share.)

The actual math of bookselling is complicated since royalty rates vary by format, price points can fluctuate, and big publishers have different distribution deals from small ones. That said, I’ll do a bit of simplistic napkin math to explain.

Let’s say the average book sells for $20. Of that, about half goes to the retailer and distributor and the other half to the publisher. Of the publisher’s $10, about $1 goes to pay back the advance—or to royalties when the book “earns out.” (Again, I’m simplifying. Standard royalty rates are between 7.5 percent and 25 percent, depending on format.) So, if a book sells 1 million copies, then the publisher makes $10 million, with $1 million going toward earning out the author’s advance. If the publisher paid an advance of $3 million, the author hasn’t earned out … but the publisher still made $7 million in revenue.

That’s not profit. Publishers have expenses to pay back, including the costs of printing the book and marketing. Economies of scale come in here. But you can see how books can be profitable well before advances are earned out. Indeed, here’s another quote from Griffin’s article, from Simon & Schuster CEO Jonathan Karp, that supports that: “About half of the books we publish make money, and a much lower percentage of them earn back the advance we pay.” Along these lines, it’s a myth that a book that hasn’t earned out is a failure. It’s even possible to have a long and lucrative career without ever earning out.

You might think that only 50 percent of books making money for the publisher is a bad thing, but most comparable industries operate this way. In entertainment, a few big winners have always funded the rest.

I assume most people reading both Griffin’s article and this one don’t care about the sales of adult coloring books and pocket copies of the Constitution and so on. What they’re interested in are the sales of what most people think of when they hear the word books: novels, biographies, essay collections, story collections, and the like. So, cutting through the noise, how do those sell?

Sadly, you can’t really say. The answer depends on the type and genre of book and the type of publisher. Someone from a prestige Big 5 imprint whose books are often award contenders and bestsellers once told me any book that sold fewer than 25,000 in print was a failure for the publisher. On the other hand, many literary fiction writers have told me that anything more than 5,000 sales is a success. So, some small-press editors might be happy with 1,000 sales.

What I can say is that no book you’re likely to have heard of is selling only 12 copies or anything near that. Most new books in bookstores are selling, at a minimum, hundreds of copies and most of the time thousands. The popular ones are selling tens of thousands, and the breakout hits are selling hundreds of thousands or even millions.

I would add here that it’s not true that only movie star memoirs and James Patterson–type commercial thrillers are breakout hits. The NYT bestseller list often includes award winners, and so-called literary fiction titles like The Goldfinch and All the Light We Cannot See can sell in the millions. And many authors earn a living—or, if not a living, then a significant chunk of a living—from their books without ever becoming as famous as Stephen King or Danielle Steel.

Again, I’m not saying all is great in publishing land. Certainly, I wish that the novels I love were selling millions instead of thousands of copies. But I imagine few are shocked that books in 2024 do not occupy the place in the culture they did 100 years ago, before video games, cellphones, social media, streaming TV, and the like dominated our time.

This seems like a fair look at the relationship between advances and big publishers’ expectations. The lowest sales target is in the thousands. Obviously, many books will fail to hit their targets, though typically not by 4,988 copies. On the other end, anything selling more than 75,000 is a rare hit. If you’re on a big publisher, you are likely to sell a few thousand copies. If you’re lucky, tens of thousands. If you’re exceptionally lucky, hundreds of thousands or millions.

The above are, I think, just print sales. E-book and audio sales are trickier to handle because of subscription services and because prices vary wildly for individual titles. The Body Scout—like most novels—has routinely been put on Kindle Daily Deals and the like, selling for $1.99–$3.99 for limited periods of time. Industry professionals have told me that for most literary fiction titles, e-book sales are about half of print sales. But for some genres, especially science fiction and fantasy, e-book sales can easily exceed print sales. (For what it’s worth, I just looked up the sales of The Body Scout, and it’s almost exactly 50 percent hardcover/paperback and 50 percent e-book/audio.)

But it’s good to remember that everything talked about here is sales of new books, a category that is not the same as reading. Many people purchase used books or check out books from the library or borrow books from friends. And that’s not even getting into reading magazines, Substacks, and more. (Of course, people also buy lots of books they never read …)

There’s a lot I think publishers could change, including focusing on big advance books and celebrity authors—Griffin is quite right that those acquisitions often go bust, especially with social media celebrities. Instagram and TikTok followers rarely translate to book sales. And books are certainly not at the center of culture in the way they were before TV and social media and all the rest.

Still, books endure. Publishing is still chugging along, not just in the Big 5 publishers—which account for most of these stats—but also in indie and small presses, where much of the best writing is published. If you want to keep the books coming, well, keep buying them and/or checking them out at your local library.

Slate is published by The Slate
Group, a Graham Holdings Company.

All contents ©
The Slate Group LLC. All rights reserved.