Social Discovery and the Implicit Graph (explicit too!)

Recently I co-hosted an event with Ro Gupta from Disqus, Marc Leibowitz from StumbleUpon, Shaival Shah from Hunch, and Mark Coatney from Tumblr called Social Discovery and the Implicit Graph.

We hosted the event at the new Union Square Ventures office as part of an ad-hoc Internet Week event in NYC.  All the proceeds went to HackNY.

We invited folks to the event with a simple description and had a great conversation.

The term “social graph” was coined originally to describe the network of connections we already knew we had such as friends on Facebook or professional contacts on LinkedIn.

More recently, graphs that are inferred as a result of other primary activities – e.g. taking a photo in a bar (Color), commenting on a website (Disqus), expressing a taste preference (Hunch), sharing a new website (StumbleUpon), etc. – have emerged in a big way, particularly in advancing discovery and recommendations. Even platforms for which explicit connections are core – e.g. Twitter, Tumblr, Foursquare and others – are realizing the value in the implicit relationships that form over time.

From this intro we assembled about 40 people for about 2 hours of solid discussion.  There were many more questions than answers and we were able to dig into many of the issues surrounding both implicit and explicit social graphs.

I got a ton of great feedback on site and even online.  Theres probably too much to go into in one post but its a topic I definitely want to dive into more.  Hearing the opinions of the group and from the companies mentioned above, we were able to tackle some of the larger questions that arise.  There are a ton of nuances when dealing with implicit actions in a web service, and even more when it comes to deciding what to do with the data.  We think about this a lot at foursquare and it was good to step back for a awhile and talk some of these issues through with a large group.

One of the big takeaways was the need to continue the conversation – we setup a google group to do just that: http://groups.google.com/group/implicit-and-explicit-graphs

If you are interested in talking about implicit or explicit social graphs or other actions, come join us.