Books: Brave New Data-Smart World

E-readers may be an innovative, eco-friendly alternative to pulp. But are publishers now looking over your shoulder while you read?

Have you heard about the latest innovation in book publishing? Neither have we. Like most monopolies, the traditional book publishing business resists change, even as retail bookselling has been disrupted almost beyond recognition by digital technology.

Launched in 1995, Amazon revolutionized the way books are marketed and sold. It offered readers a platform to share opinions and its algorithms gave them recommendations based upon collective browsing and purchasing habits. Yet book publishers’ basic content-­creation model (acquisition-editing-­design-production) remained in 2000 largely unchanged since the Gutenberg Press opened for business in 1436.

The 21st century ushered in on-­demand printing and mobile e-book ­devices, such as Amazon’s Kindle and the iPad. While these were not publishing­ innovations so much as retail ones, they have enabled competition from self-publishing and the “indie” press, and forced book publishers to rethink the ways they acquire and develop content.

To compete in today’s wired world, publishers now “pimp” their e-books with “enriched content”: adding multimedia formatting and interactive features. Tap a word or a phrase to watch a video, see a map, read a Wikipedia entry or even click a link to “sponsored content.”

Reading readers
The e-reader is not simply a device for consuming e-books. Like most online hardware, it is also capable of capturing a treasure trove of user-generated data – demographic information and personal reading habits, such as how and when e-books are opened, read, navigated, finished or abandoned.

“We can now measure the level of engagement a book generates and thus indirectly measure the level of emotional response,” claims Andrew Rhomberg, CEO of Jellybooks, a London-based startup offering e-book publishers and authors what Google Analytics provides webmasters – a way to parse data on actual usage, not just consumer behavior. Rhomberg (who grew up in Austria) sees e-readers as feedback “nodes” for an “Internet of Bookish Things.”

We can now measure the level of engagement a book generates and thus indirectly measure the level of emotional response.

– Andrew Rhomberg, Jellybooks

Jellybooks beta-tests e-books by offering readers free copies in exchange for permission to record and analyze their reading habits – their rates of completion, reading velocity, and likelihood to recommend. It turns out that only five percent of the books tested were finished by more than 75 percent of the readers (men tend to abandon a book sooner than women), and the average completion rate was between 40-50%.1 Business books are among the least finished e-books tested. Rhomberg concludes: “The 19th century approach of 100-page rambling introductions that lay out the background will turn off 21st century readers.”


A boon for publishing?
Rhomberg believes book sellers and publishers have yet to put such data to much practical use, unlike Netflix which parses its customers’ viewing data to help it (and them) make better programing choices.
For the moment, it seems publishers are using such metadata to make decisions about promotion rather than development of content. But for how long? Will publishers do A/B testing on manuscript variations, in other words statistically rate reactions to two different versions? Will they be able to scientifically predict if a manuscript will become a best seller? Will they abandon authors who don’t garner high levels of reader engagement?

In fact, these data may eventually eliminate the middlemen altogether. Reader behavior could be processed directly by an algorithm that will automatically acquire, edit, design, publish and sell an author’s work directly to the consumer. Who knows, maybe even the author will become obsolete! And we could all go on holiday. Now that would be innovative, indeed!



1Editor’s note: The version of this article appearing in the June 2016 print edition stated “only five percent of the books tested were finished by most readers.” This underrepresented the broader test results.

Leave a Comment