Frictionless Sharing and the Interest Graph
This past Thursday, facebook announced some pretty substantial updates to their open graph and introduced the concept of frictionless sharing.
Facebook presumably thinks that they can use this frictionless sharing data as an additional input in refining advertising algorithms and making recommendations about what is important to users, but the question remains whether this data is in fact real “signal” or just additional noise. “Zuckerberg’s Law” states that the amount of content you share online doubles every year. As the recent TechCrunch article above indicates, the recent frictionless sharing introductions are only furthering that law.
This blog post gives an interesting discussion of the different types of graphs apart from the social graph of the facebook network. Of particular interest is the “interest graph,” which facebook seems to think they can capture as a result of keeping track of these “frictionless shares.” However, I submit that though the inclusion of these frictionless shares will help provide new context around the existing relationships on facebook, the facebook network will not subsume a user’s need to particpate in separate interest graphs outside of facebook.
For example, the recent Spotify integration with facebook has shown me the music that a number of my friends are listening to. I’m a fan of death metal. Almost none of my friends are fans of death metal (surprise, surprise!). As a result, a lot of the information about my friends’ listening habits feels like noise. I use other social networking sites (ex.fm, reddit, last.fm, tumblr and others) as my primary method of music discovery because facebook making recommendations about this is nothing short of entirely useless. As it stands today, facebook simply isn’t doing a great job of recommending new content to me or helping me with discovery.
I don’t think I’m the only one that thinks this way. Fred Wilson agrees that we won’t have one single social graph to rule them all. No matter how much data facebook captures and how good their algorithms get, I think it’s unlikely that they’ll be able to capture all the subtleties of different interest (and other types of) graphs.