Goodreads is now the Netflix of book recommendations, or at least the company thinks so, seeing as its CEO Otis Chandler said exactly that in a press release announcing Goodreads’ new recommendation engine yesterday.
It’s about time. Founded in 2006, Goodreads has gathered nearly six million users, a plump following for a niche social site, but while six million is a respectable number, the nature of Goodreads’ service requires these users to be fairly active to get much out of it. To grow the audience, Goodreads needed to appeal to a more passive audience. Some may call these people lazy. I prefer to think of them as busy. And yes, I’m one of them.
On paper, I’m the perfect potential Goodreads user. I love literature. A large part of my career is social media. But still, I’ve never been completely sold on Goodreads. I wanted to like it. I wanted to love it, but after a few weeks, it all started to feel like work. Sure, it’s nice to see which books my trendy college friend is waxing on about this month, or that some of my high school friends are flying their “Twilight” flags proudly, but I knew this all from Facebook. Goodreads had become just one more thing I needed to do everyday. It was a great community, sure, but what about my right to be a perfectly lazy consumer?
At the site’s inception, Goodreads prided itself on this idealistic human-to-human book discovery that the site offered. It was part of the “Aren’t you sick of robots telling you what you want?” digital boom. And it was appealing. The web felt like a community again – until we quickly realized that communities require a lot of upkeep. While peer recommendations are important, it’s hard to argue against math. As Pandora and Netflix gained notoriety for their algorithms, users flocked. What Goodreads needed was our information to sell ads. We became a little less inclined to give it to them without a means to an end.
The company took notice. In March, it bought Discovereads, an algorithmic book recommendation service, and plugged it into the information they had already gathered from users. While the service is overdue, it was smart of the site to approach its allegiance with machines lightly and waited until the “evil” became a necessity. And oh, how it has. Book discovery has become plagued with a flooding of the zone. By year’s end, more than 300,000 books will have been published in 2011, and the rise of self-published e-books has made it even harder to sift through the muck. “We have so much data on what people read, but we’ve never built algorithmic recommendations because it’s a really, really hard problem to get right.” Chandler said. Translation: He finally found his perfect cocktail of literary math genius.
(MORE: 5 Great Social Reading Sites)
While we still know relatively little about the final Goodreads algorithm, the company has not shied from the inevitable Netflix comparisons. Discovereads described it as a combination of multiple machine learning algorithms and graph analysis used to analyze book ratings and spot trends. Plug in the gold mine of user data inside the Goodreads vault and you’ve got yourself a recipe for a pretty powerful discovery channel.
While this service will ultimately serve users, it will also pad the company’s profits. No doubt authors and publishers will relish the chance to buy target ads sent directly to members who are recommended similar genres. (And because all six major publishers are active on Goodreads, this is pretty much a revenue sure-thing.)
Of course, there’s argument against the rise of machine-based discovery. Where’s the risk? Where’s the adventure? No doubt Chandler is still betting on users’ connections with each other as a safeguard against the boredom that can accompany the repetitiveness of our own preferences. At least now you’ve got the option.