I am at the Graphing Social Patterns conference today, which promises to be one of the most interesting events I have been to in recent memory given the topic and the diversity of the audience.

Reid Hoffman is doing the keynote address, which is interesting on it’s face because it begs the question why Reid is keynoting a conference largely about Facebook. Having said that, Reid is a wealth of knowledge about social networks, and given his experience with Linkedin he has legitimate claim to being a pioneer in this category.

The slides Reid used were really text heavy and detailed so rather than just give you all my notes I will try to condense it down to a couple of key bullets.

Reid reminded the audience that in 2003 there was a similar debate about web 2.0 social networks around features vs. applications vs. platform. Myspace evolved the notion of widgets that are hacked into the system, Ning focused on a platform for building social network applications, and lastly, and Facebook is the first platform that integrates access into a very broad social graph.

What is interesting about this comparison is that Facebook enables key features that Myspace explicitly prohibits, such as access to relationship data, key user profile data, and of course, access to platform features. Reid keeps mentioning Ning, which from what I have heard is achieving some real traction, but it’s kind of odd to include them with FB and Myspace in this discussion.

One relevant comparison given the speaker is Facebook and Linkedin. Reid offered up a slide on this and called out Linkedin’s robust search capabilities as a key differentiator. This I will certainly agree on, Linkedin does have a very effective search function that goes beyond name search to roles, companies, pictures, etc.

The Linkedin Answers vs. various Facebook questions applications are another point of differentiation and on this point I would suggest that for what Linkedin is trying to do, their Answers service is a good fit. I’m not sure it is an apples-to-apples comparison.

Messaging is another differentiator but on this point I think Linkedin is operating at a disadvantage. Their messaging service really doesn’t encourage repeat use, but on Facebook I am finding that it is increasingly more useful for me to Facebook message someone as opposed to email them.

The last point I want to highlight from Reid’s keynote is the question will there be one massive social graph that rules them all. I agree with Reid that it is inevitable that  social graphs fork and divide to become more granular, while at the same time scaling in overall size. This mimics how networks in general operate, nodes divide and multiply which grows the network as a whole. A network doesn’t require every node to have complete knowledge of every other node.

So how will social graphs delineate? According to Reid they will fall into different semantic lines that capture relationship data, connection types (e.g. personal vs. professional), and lastly, have different baseline rules for engagement.

Tags: GSP07