Telepocalypse

Social outcasts

Via Stuart Henshall comes news of StumbleUpon. This is a bookmark-cum-clippings social network service. So unlike, say, blogdex or del.icio.us, it’s as much about who the bookmarks come from as what they are.

My farewell patent frenzy at Sprint led to a filing along these lines. I won’t disclose the details — you’ll have to wait for it to get published via the usual channels. But I’ll give you some clues. There’s a hidden power in the dark heart of the smart network traditional telco. A central aggregation point of all traffic — voice, email, IM, or SMS — gives one entity a complete view of who talks to whom, for how long, who ends the conversation, who returns whose calls, and so on. In other words, there’s a latent social network in there.

If you call me and we speak for an hour, that tells you something about or relationship. If a telemarketer makes thousands of calls and they typically last 10 seconds, and are terminated by the callee, that tells you something too.

Assume opt-in from users to overcome privacy concerns. What the telco is sitting on is potentially the biggest social networking goldmine east of the Yukon. As I’ve said before, the core asset of a telco is the data it acquires, not the network. Networks are loss leaders to create foot traffic and encourage people to engage in transactions.

Another consequence of StumbleUpon and its ilk is that social networking works best when it is achieved via recovery of the social pattern from an unrelated adjacent system. If I decide to take you off my buddy list, you might get offended. If I don’t call you and the relationship fades away, nobody cares. These proxy social network systems are also very hard to game. Bill Gates isn’t going to spend an hour on the phone to me no matter how many will-you-be-my-pal invites I send to his secretary.

For pure stupid network services, this type of latent social network analysis is tougher. The more decentralised the application’s control mechanisms (e.g. users have a choice of browsers, or IM clients) the bigger the coordination problem in obtaining and aggregating the data in a central point.

Now, any ideas of how a low-profit VoIP service provider like Skype might turn a few cents from their virtual network…? I do!