We’ve mentioned e-book statistics company Jellybooks a time or two before—most recently in August of last year, when we covered a piece by founder Andrew Rhomberg talking about how his company gathered statistical information on how customers read e-books. Now the New York Times has a longer piece looking at this practice in depth.
Jellybooks sends e-books to a few hundred volunteers and asks them to read them then click a link when they’re done to send along statistics that show how much of the book they read and when. The results it’s been getting have publishers in a tizzy:
On average, fewer than half of the books tested were finished by a majority of readers. Most readers typically give up on a book in the early chapters. Women tend to quit after 50 to 100 pages, men after 30 to 50. Only 5 percent of the books Jellybooks tested were completed by more than 75 percent of readers. Sixty percent of books fell into a range where 25 percent to 50 percent of test readers finished them. Business books have surprisingly low completion rates.
Publishers have been using these results to adjust how much they spend on marking books, reducing the marketing budget for books people largely don’t finish. And those publishers don’t want their names to come out, lest their authors be upset to find their books are undergoing this sort of analysis. They don’t seem to be using the information to make actual changes to manuscripts yet, at least not so far as the New York Times knows.
I’m not entirely sure these are valid or meaningful results, though. For one thing, the sample size is rather small, and consists of people who are getting the book for free and are fully aware that their reading habits are being tracked. These factors could influence the way in which they read. Do they read the same books, or read them in the same way, as the thousands and thousands of people who spend their actual hard-earned money on them?
And in a larger sense, I’m not sure the idea that most people don’t finish most books they start is anything new, either. I’m pretty sure people have been putting books down (or throwing them at the wall) for centuries now—it’s just that before e-books, there never has been any way to track those habits without putting readers in a lab with lots of cameras pointed at them. (And that’s not exactly a naturalistic reading environment either.)
The more interesting thing, though, is considering what sort of information Amazon might have about the way people read e-books. After all, Kindles dutifully phone home where you are in the book you’re reading so that Amazon can set you right where you left off when you open the same book in another device or app. Amazon doesn’t have a selected sample of volunteers who click a link to send that information in—it has that information about everyone who uses its platform, automatically. (Or at least those people who aren’t sufficiently paranoid as to disable Amazon’s position sync feature.)
As secretive as Amazon is, there’s no way to know how or even if it’s using this information. It’s certainly not sharing any of it with the publishers. But given how much stock Amazon puts in statistical information—some even believe that’s the real reason it opened its Amazon Books store, to gather information about customers’ physical book-browsing habits—it’s seems likely it’s doing something with it.
Amazon has its own publishing division. Is Amazon requesting changes to its writers’ books based on the data it’s getting? Again, there’s not really any way to know. But my suspicion is that even if it is, or even if Jellybooks’s client publishers started using its information in that way, it’s not likely it would make any major changes to the way publishing works.
Despite the New York Times comparing this to Moneyball, the book and movie about a baseball team manager who used statistical analysis to build a better team, this data-gathering doesn’t represent as big a change as all that. Publishers have always run advance copies of their books by focus groups and listened to their feedback. It’s just that they never could get feedback this finely-grained before. So, really, it’s only a difference of degree.
Still, it will be interesting to see if this data reshapes the publishing industry over the longer term. We’re already seeing some changes in the world of television, where Netflix and Hulu can gather the same sort of information about how people watch shows. Netflix has been releasing complete seasons of the shows it produces, so people can binge-watch them. Books are “binge-readable” as a matter of course, though some publishers have tried going the other way and serializing them.
Whether these statistics are truly meaningful or not, it’s a powerful reminder of just how much more data our new digital lifestyle produces. It’s worth keeping in mind, though, that not all that data will necessarily be useful to everybody.