I just mapped all my LinkedIn connections onto a visual graph through a new web service called InMaps. It’s worth checking out; and even though it’s still not particularly useful so far, the service does provide an interesting look at were network analytics is likely to go in the not-too-distant future – and that’s going to get pretty interesting, I think. Let me know your take once you’ve had a chance to check it out.
After connecting InMaps with my LinkedIn account, it clustered all my connections into eight clumps that you can see by clicking on the image to the right. The whole thing took a just a minute or so. The result is a graph that maps pretty well to the various layers of my life. InMaps forms the clusters and then it’s up to you to name them based on what you think they are. For instance, the big blue cluster at the bottom of my map seemed pretty much like it was my “Microsoft cluster” at first blush, but then I looked a little more closely and realized it also included connections from other technology companies. Most of my technology folks just happened to be Microsoft folks because I worked there for almost ten years, so that makes sense.
InMaps isn’t just segmenting people by industry sector or some other category in LinkedIn; it’s far more subtle than that. For example, I used to run a mission-driven technology consulting group called Groundwire that specializes in helping organizations that work to protect the environment. InMaps picked up on an interesting split in the environmental movement that showed up as two distinct clusters in my network. After looking at the people in each cluster, I named one cluster “dark green” and the other “bright green” (a term coined by my friend Alex Steffen to describe the technology-oriented sustainability wing of the green movement). Just to be clear, I didn’t name the clusters and then go find the people that belonged there. No, InMaps found the clusters and when I looked at them, I saw that they mapped to this fairly subtle distinction between the people I had worked with in various environmental organizations. It wasn’t perfect, but I have to admit I found it surprisingly nuanced.
Where the clusters sit on the map can also tell you something about the clusters of people in your network. For instance, the “bright greens” (a fairly small, distributed cluster in light blue above) is nestled in between my for-profit technology sector folks (dark blue), my non-profit technology folks (orange) and my “dark green” folks (tannish yellow). It makes intuitive sense that the tech-savvy “bright greens” would be located between other environmentalists and my for-profit and non-profit technology clusters. Again, this isn’t anything I told InMaps to do. It just emerged from its analysis. This will make more sense to you once you try it yourself and see real people you know showing up in your own clusters.
Of course, what I really want to see this evolve into is an interactive interface along the lines of something like the Visual Thesaurus. Play around with this amazingly cool tool for interacting with the relationships between words, and you’ll instantly get what I’m talking about. Imagine being able to drill into a particular LinkedIn connection and see their relationships bloom into the primary focus on screen. Or just think of being able to drag a particular person with your mouse and have that gently ‘tug’ at their web of connections with other people in the network. I’ve little doubt that this is where these kind of network analysis tools can go – if given the chance.
The question is – will they be given the chance? I still don’t have a good sense for LinkedIn’s philosophical approach to its API. Twitter has opened its API in such a way that third party developers have had the freedom to generate a smorgasbord of tools for the service. I’m not just talking about things like TweetDeck, for managing tweets. Facebook allows that too. I’m talking about the hundreds of other nitty-gritty tools that provide all kinds of slices on the Twitter data set. Will LinkedIn keep the goodies of its network graph just for itself? Or will it really open it up for third party development? Will it follow the Twitter playbook or the Facebook playbook? The answer depends on whether LinkedIn sees itself as a “social network application” or a “social network utility” on which other applications can be built.
One can only hope it will be the latter, and that LinkedIn takes its cue from Twitter so that we end up in a world with more robust competition for the interfaces onto the amazing information buried in our own social graphs.