Data on book sales is murky. More numbers may not be the best solution.
If you think about it, isn’t it a little strange that we, the public, generally have very little idea how many copies of individual books are sold?
Movies report the weekend box office grosses down to the dollar. Streaming services such as Netflix can be opaque, but regular broadcast television has the Nielsen ratings. Spotify reports the number of streams, and if you go to YouTube or TikTok, the number of views counts upward in real time.
Bestseller lists rank the order of books, but they don’t come with actual sales figures. BookScan, the book sales tracking service operated by Nielsen, is a subscription product primarily accessed by publishers which doesn’t count sales that happen outside of the retailers that are part of its network.
During the recent trial determining whether Penguin Random House and Simon & Schuster can merge, lots of weird data started floating around social media about how many copies typical books sell, including one nugget that 50% of the 58,000 commercially published books per year sell fewer than a dozen copies.
A dozen, as in 12, as in fewer copies than if you could get each of your close friends and family members to buy your book.
Writing in his “Counter Craft” newsletter, Lincoln Michel, author of “The Body Scout,” showed the ways this stat is dubious, further proving how difficult it is for the public to know how many copies a book has sold.
This brought Kristen McLean, the lead industry analyst for BookScan, into the newsletter comments to provide some of the best aggregate data I’ve ever seen, including that two-thirds of books published in a given year by the top 10 publishers sell fewer than 1,000 copies.
Only 163 books, 0.4% of the total, sold 100,000 copies or more. We can guess, but we do not know which books those are.
Writing recently at Public Books, Melanie Walsh goes looking for, and finds, other sources of data about what books are read and by whom. Walsh co-leads the Post45 Data Collective, a group that is trying to use open-source data to make cultural industries, such as publishing, more transparent to the public.
An early project involved collecting data about the race and ethnicity of authors published by the Big Five publishing houses (including Penguin Random House and Simon & Schuster), which revealed that, for example, only 11% of fiction published by the Big Five in 2018 came from nonwhite authors.
This kind of quantification of something we suspect is true — that nonwhite authors are disadvantaged in the big publishing marketplace — can help publishers be more reflective and thoughtful about what they put into the world and can help readers be more thoughtful about what we consume as well.
However, I am less certain that making the actual number of books sold or read transparent to all will necessarily change much.
In truth, while the public may not know these numbers, publishers have a very good handle on exactly how many copies their books sell because they literally must account for them in paying royalties to authors. The people who decide what gets published know exactly how many copies of a given book have sold.
I think there’s a sound argument that there are untapped and underserved markets out there, and for sure I’d love to be able to look up the sales figures for books by my friends and enemies. OK, mostly my enemies.
But focusing on the number of books sold is more likely to narrow rather than expand the range of what gets published.
Replicating what is already popular will become the safe bet.
I want publishing to be risky and interesting and as diverse as imaginable.
John Warner is the author of “Why They Can’t Write: Killing the Five-Paragraph Essay and Other Necessities.”