I can’t possibly do justice to his great post, but Andrew Chen has been doing user interviews to understand how people think about their web of personal relationships, and his conclusion matches up well with my own intuitive sense for where the useful social graph really lies: it’s not Facebook, it’s not the inbox, it’s not any one datastore, it’s the aggregated and normalized patterns we have in all our channels of communication, from phone to email to IM to socialnet. We as users own that data, and when we can pull it all in and remix for our own benefit (and not just so any single data provider can market to us more effectively), we’ll really be on to something.
