Big data is an intimidating concept for publishers who don’t collect much info about their customers already. A first step is to think about “data-driven decision making,” said bitly chief scientist Hilary Mason at a Tools of Change roundtable this morning.
In the next couple of months, URL shortening service bitly is launching a real-time search engine that would let publishers see how their content — including books on Amazon — is being shared. And TOC’s Joe Wikert thinks publishers should track data like how far readers get in an e-book before they stop reading. “Amazon is probably gathering this data already,” he said. “At some point publishers will probably be able to buy access to that information,” but it’s even better if they can collect it themselves.
Audience member Dominique Raccah, publisher of the innovative Sourcebooks — whose e-book revenues were up 795 percent in 2011 — said publishers should use data to answer questions like how many of their books fail and how much it costs them to publish a book.
Sourcebooks is collecting data about their books and about how readers use their website. The biggest success they’ve had is using data to position books, Raccah said — “where we think the book is about one thing, and the consumer thinks the book is about something else.” Publishers can poll readers to find “what’s exciting somebody, what’s really turning them on — what it is that moves the needle for them.”
Bitly’s Mason agreed that data doesn’t have to be big or “require a huge amount of infrastructure” to be useful. And publishers can collect data through resources that weren’t explicitly designed for that purpose. For example, Mason said, if a publisher can’t decide what a book’s title should be, they should buy Google Ads for two different titles and see which title gets more clicks. The product you’re advertising, she said, “doesn’t actually have to exist.”
Correction: An earlier version of this story incorrectly stated Hilary Mason’s last name. Her last name is Mason, not Rosen. I apologize for the error.