graphs aren't social

At Ning, we often hear the following question from people outside the company: how do we see the ‘social graph’ from our perspective?

They don’t ask what people enjoy doing on the service. They don’t ask for stories of users, of finding long-lost friends, or of newfound friendships. They don’t ask what makes people on Ning join ten different social networks, as they often do, even if this flies in the face of conventional wisdom about how people are sick of joining social networks.

No. They ask about the social graph.

I am honestly amazed at this question – is that really what they care about? An abstraction? I am amazed, and at the same time I understand where they are coming from.

When I started at what would become Ning some three and a half years ago, I had spent a few years thinking about social networks, ad hoc group formation, and self-organization. I was completely immersed in the terminology with which we try to make sense of phenomena that is sometimes hard to grasp at an intuitive level. Groups of people are not just the sum of the individuals that compose them, and this creates a chasm that is hard to cross–for we, as individuals, project out of ourselves and into the world when we look for answers.

What I hear when I listen to someone talk about the social graph, what I think of when I read about the social graph as if it was a real, tangible thing –in fact the only real and tangible thing–, is a mathematician, talking to someone proudly showing off her new car, pointing at a wheel and asking: “how’s that circumference working out for you?”

The notion of a circle, how it is unequivocally described, what its characteristics are, is undeniably of great value to anyone interested in mathematics or geometry, and even to anyone interested in designing a car or producing a wheel. And it is of no value to that same person when using a car. Not in the mechanics of the action, mind you, in which the device is at work, but in the actual use, and even enjoyment of what the abstraction, applied to something real, can do. Likewise, the notion of a social graph is interesting to us in nerd-dom as part of our continued understanding of what it is that we’re actually doing, but it holds almost no sway in determining whether something has value for people.

Hundreds of social networks have come and gone. They all had ‘social graphs’ as a key part of their functionality. If this mythical graph is of such importance, why, then, did so many of them fail?

A thousand factors, to be sure, but one of the most important surely must be that they didn’t connect, in some meaningful way, people to each other, and to the service.

I choose words carefully when I say people and not users. Semantics matters. It’s easy to fall into abstractions to handle numbers or notions that seem too big to grasp, but we should resist the temptation. Because it’s when we start designing our software for abstractions, rather than for people, that things go off-course.

Don’t get me wrong – I live and die with abstractions. You will have to take my little diagrams and sourcetrees and formulas and tools from my cold, dead hands. But these are the things I build software with, not for. And that’s how it should be.

We must see past the abstractions and speak to what people really use, and crave, and communicate with. Let’s use the abstractions and obsess over them, but then translate that into something real and meaningful.

Graphs aren’t social. They’re just graphs.

And we should never lose sight of that.