How long is a bestseller? Two studies, two sets of books
Published · Last updated
Ask whether bestsellers are getting longer or shorter, and the answer depends on which years and which books you count: the two studies most often cited cover different periods and different lists.
In February 2022, the analytics firm WordsRated published an analysis of 3,444 hardcover fiction and nonfiction titles that had reached the top three of the New York Times weekly bestseller lists between 2011 and 2021. Average length, they found, had dropped from 437.5 pages to 386, a fall of 11.8 percent.
The share of bestsellers running over 400 pages fell too, from 54 percent to 38 percent of the list. And since 2016, short books have simply outlasted long ones on the chart: an average of 7.8 weeks for the shorter titles against 5.9 weeks for the longer ones.
Six years earlier, a different company reached the opposite conclusion on a different set of books.
Flipsnack, in a study published in December 2015 and reported by The Guardian and Smithsonian Magazine, examined 2,515 fiction books from the New York Times Best Sellers and Notable Books lists and Google's "Most Discussed", spanning 1999 to 2014, and found average length rising, from 320 pages to 407, a jump of 25 percent.
Different books: fiction and nonfiction hardcovers at the top of the chart on one side, fiction from bestseller lists, critics' picks and a Google list on the other. Both count pages. Opposite headlines.
Both studies describe their samples, and the samples differ: fiction and nonfiction top-three hardcovers on one side, fiction from bestseller, Notable Books and Google lists on the other, over windows that overlap only in 2011 to 2014. On those shared years the two series don't even meet: WordsRated's 2011 average is 437.5 pages, while the highest point in Flipsnack's series is 407, in 2014.
A third data point complicates things further, in an interesting way. In May 2025, a writer named Ray Tabler examined a smaller, differently built sample: 118 of the most-read novels published between 1722 and 2012.
Rather than a single steady climb or decline, he found a dip. Bestselling length falls sharply after 1875, bottoms out around 1925 to 1930 (almost nothing in that stretch clears 80,000 words), and then climbs back up in the decades that follow.
That shape says nothing about the two earlier studies: Tabler's list shows a single dip, from 1875 to 1930, followed by a rise, and it has almost no titles after 2007, one from 2012, so it is silent on the years WordsRated counted.
None of the three studies is wrong, exactly. They're measuring different populations, over different years, with different definitions of what counts as a bestseller in the first place, and the resulting numbers simply don't reconcile into one tidy verdict.
The three reports don't even speak with the same confidence: WordsRated states a single downward trend; Flipsnack reports a rise but asks whether other factors are at play; Tabler calls his curve a "roller coaster". Length isn't a single, stable line moving in one direction. It's more like weather than climate: real, measurable, and true for the period it covers, but a poor guide to what the next decade's numbers will say.
WordsRated's own data adds a date that is easy to misread. The firm marks 2016 as a turning point, the year after which short books started outlasting long ones on the list.
That pivot is about weeks on the list, not page counts, and 2016 falls after the end of Flipsnack's window, so it can't account for the disagreement between the two studies.
A reader curious about "the" length of a bestseller today is really asking three separate questions at once: which years, which list, and which two points on that list you're comparing. Answer those first, and the two contradictory-sounding studies stop contradicting each other quite so loudly. They were just never describing the same thing.