High expectations of social networking sites

A number of people, including Ton Zijlstra, Earl Mardle, and Martin Roell have written about the “failing” of social networking sites, Orkut in particular, to represent relationships between people in a meaningful way. They say that the sites are not doing this effectively because they are gathering insufficient information—a simple designation of “friend” or “connection” (Ryze/LinkedIn), a binary “good” rating (Ecademy), a slightly more granular variation on ratings (Orkut), etc., and call for the social networking sites to capture more meaningful things, like whether or not you’ve met the person face-to-face.

I’ll get to the face-to-face point in a later post, but first, let’s talk about our expectations of social networking sites. Marc Canter starts down the right path when he says, “Trying to lock everything into a black or white – “friend” or not – is the root of the problem (IMHO.)” From there, he goes into an explanation of how PeopleAggregator solves the problem with additional levels of granularity, ranging from “Know In Passing” to “Close Friend”.

I don’t argue that this is an improvement, and I’m glad to see it. My question is this… are we really going to make use of this information? And if we do, is it alone sufficient to be actionable? In other words, is it really useful? Interesting, yes, but useful?

I’m not going to do business with someone just because they’re someone’s “close friend”. I want an introduction or an endorsement. If someone approaches me directly, claiming a relationship with someone as a basis for trust with me, before I even consider any business with them, I’m going to actually communicate with the common connection.

I find the exploration of reputation systems an intriguing intellectual exercise, and I can see the need for them in communities where relationships are not developed and maintained, i.e., the transactions are short and clearly defined—eBay, Amazon, Elance, etc. But that’s not the case in a social networking community, or, at least I believe it shouldn’t be.

On Elance, I might narrow the field of web designers based upon their reputation, but on Ryze, I’m going to choose based on the recommendation of one person I trust. On Amazon, among several books on network science, I might pick the one with the most stars. But I’d probably rather contact Joi Ito and see which one he recommends. I could list some domains for sale on eBay, or I could contact some people I know on Ecademy who might be interested, and sell them to them (I did).

So to me, the most useful, actionable information will never be someone’s abstract “reputation” rating, or the list of someone’s “friends”, but specific, detailed endorsements and personal referrals and introductions.

I think we are simply putting too much emphasis on the technology in these sites, and not enough on the human aspects. The best relationship building takes place when people actually communicate, rather than attempting to objectify each other. Just because the medium is digital doesn’t mean that our relationships have to be quantified.


Great article from Terri Senft: Against Reputation, plus commentaryon it from Clay Shirky.