The New York Times is the most commonly mentioned object across books, ahead of World War II, the Bible and Coca Cola.
At least, it’s the most commonly mentioned in the 7,000 books that have been analyzed by a startup called Small Demons . The company has agreements with four of the biggest six publishers to index the people, places and things in their books, and it’s been do so at rate of about 1,000 books per month.
Pinpointing books--and the passages in them--that mention the song “Hey Jude,"Doc Martins or your hometown makes for an entertaining website . But what makes Small Demons more interesting is what the New York Times and other brands could do with the same kind of information.
“Instead of saying your friends like this, we’re able to say, here’s something this character wears or something he’s drinking,” says Small Demon's VP of Content and Community Richard Nash.
Welcome to, as the startup puts it, "the storyverse."
Small Demons plans to release an API that allows all developers to leverage book references. Imagine a fashion website organized by what specific characters wear, a travel site that displays relevant quotes from novels next to destinations or advertising targeted based on book preferences. Small Demons can power it.
Of course, the million-dollar question (possibly literally) is whether consumers will actually be influenced by book references in the same way they’re influenced by social recommendations.
Small Demons' CEO, Valla Vakili, tells a story about changing his music purchases, drink of choice and even vacation destination based on his reading of Jean-Claude Izzo's Marseilles Trilogy. But do books have that purchasing influence for everybody? Does every book?
Nobody knows. Most efforts to mine data in books have focused on publisher marketing efforts and book recommendations. A project called Booklamp , for instance, uses what it calls “StoryDNA”to power its Pandora-style book recommendation engine. But its real customers are publishers who want to refine their targeting and marketing based on a book’s contents.
Small Demons, like Booklamp, wants to help publishers sell books. In exchange for access to the digital files in publishers' libraries, it’s handing them a list of cultural references for each of their books. They’ll use the data to tag their electronic files with new terms that help surface them in more searches.
But Small Demons also conducting an experiment that hasn’t really been tried before.
"Once we started to do this, one of the things we started to realize is books contain worlds," Nash says.
Now they’ll see if those worlds carry any clout outside of their containers.
[Image: Flickr user Visibleducts ]