7

 

By Brian Howard

Last week, following a soft-launch the week prior, Random House marched out BookScout, a Facebook app designed to link readers with books they’ll like but might not have discovered on their own.

The recommendation engine draws on a user’s “likes”—both on one’s Facebook timeline and then directly through the app. Intriguingly, BookScout is not purely a Random House recommendation engine—it’ll tip readers to any book in print, regardless of whether it was published by its own imprint Knopf, Big Six rival HarperCollins, indie McSweeney’s or even Amazon Publishing.

Though the app’s early reviews have been mixed (I’ve found its recommendations to be uncanny: How did I not know about Michael Pollan‘s Second Nature previously?), we’re told that Random House is taking an iterative approach to BookScout, with plans to tweak the product based on user feedback.

We asked Amanda Close, Senior Vice President, Digital Marketplace Development at Random House, to take us through the steps of getting an app like BookScout into the world, and to give us a peek at how it might evolve going forward.

Brian Howard: So anyone who looks at my Facebook would know that I like, say, Michael Pollan, John Hodgman, gardening and brewing beer. How does BookScout know this? How does it work?

Amanda Close: After you install the app, BookScout takes a look at your Facebook timeline to learn about what you like—it looks for previous books you’ve liked on Facebook or other websites, as well as general interests, and uses that information to recommend titles you might enjoy.

Brian Howard: I know I’ve posted about certain authors and certain books, so I’d imagine those interests are easy enough to parse. But how does it get more topic-based info?

Amanda Close: The app gathers information from Likes on your profile, rather than status updates or other information.

Brian Howard: So it’s my understanding that BookScout does not just recommend Random House books, but basically all titles in print by pulling title data from Ingram, Bookscan and BISAC. Is that correct?

Amanda Close: The recommendation engine includes titles from all publishers—our development team worked with Ingram to gather title data on all their in-print book listings to power the app. Bookscan and BISAC information are also used in the recommendation algorithm we developed.

Brian Howard: What’s the thinking behind investing in a technology that could just as easily sell a Random House book as, say, a Harper Collins book? Or an Amazon Publishing book?

Amanda Close: We know that word of mouth is the number one way readers find out about new books, so book discovery is incredibly important—making it easier for readers to find great books and integrating conversations about books into Facebook, where so many people already go every day, is beneficial to the entire industry.

Brian Howard: Does BookScout learn? Do your recommendations get better as you view more books?

Amanda Close: The more books you Like within BookScout, the more information the app has to go on about what you like to read. It also incorporates Likes from other websites, and you can mark books “not interested” as well, that helps BookScout learn about your tastes as well.

Brian Howard: How long was Book Scout in development? 

Amanda Close: BookScout began over a year ago, as my team began to research what kind of tool we could build within the Facebook platform. We surveyed a number of book club members to see what sort of tools they would find useful, and our research told us that what readers really needed was an easy way to find which book to read next and share that with their friends.

Brian Howard: Why the decision to go with a Facebook app? Were there other platforms you were considering?

Amanda Close: Facebook is the number one platform for word of mouth discovery. We look at all major platforms and aim to embed books and authors into the discussion, but this initiative was always intended to make the most of the Facebook platform.

Brian Howard: I’m told that this is an iterative process. What sorts of tweaks and enhancements are in the works?

Amanda Close: We’ll continually be responding to feedback from readers, now that the app is up and running, on how to make it better and more useful, and are already planning to create a mobile experience, as well as the ability to display customized topical or timely recommendation ribbons.

Brian Howard: What kind of feedback have you received?

Amanda Close: We’ve received great feedback from readers who’ve tried out the app so far—the most popular comment being that once you start adding books to your shelf, browsing and Liking, it’s hard to stop!

Brian Howard: I know it’s early, but can you share any initial results? Sales bumps, backlist titles being rediscovered, etc.? 

Amanda Close: We are happy with the launch and are seeing a healthy number of sign ups each day. The adoption is encouraging enough for us to continue building a better app.  And we are stimulating more discussion about books and authors on Facebook.

Brian Howard: Could you foresee a time when this platform could be the front door of its own e-commerce site?

Amanda Close: BookScout currently links to many of our retail partners so readers have the option to buy the books they’re browsing within the app easily, but the tool was intended to be a driver of conversation about books, not a sales initiative.

Brian Howard: I’m told that The Lean Startup played an integral role in the development of this app. Can you briefly discuss that?

Amanda Close: The development team followed the Lean Startup model in launching BookScout—we wanted to create the app and launch as soon as we had the basic functionality ready, with the idea that we’d keep adding features and testing once it was out there and being used, having readers help inform the “final product,” rather than launching with our entire wish list of features completed.

 
7