Over the weekend I read a post from Michael Sampson (@collabguy) on orchestrating accidental collaboration and it resonated strongly with me, so much so that I wanted to share a couple of thoughts on the subject.
The increased transparency and collaboration that accompanies a social business comes at a price and that is an increasing information flow (Activity Streams: Information Overload or Gold Dust?). While this potential deluge presents some challenges, it also represents a huge opportunity in understanding what’s going on across the business. This redistribution of knowledge has the potential to allow “serendipitous collaboration”, enabling people who don’t know each other to cooperate, share knowledge and work more effectively together.
Most of today’s social networks focus on the notion of explicitly following “things”; people, content, or applications. They talk about filtering the stream, but what they really mean is that you can tell the system exactly which events you care about and they will add those to the stream updating you when there is any update. So essentially you need to know in advance exactly who or what you might be interested in, which to me sounds like the opposite of serendipity.
Now there are clearly scenarios where this explicit tracking is useful, such as when I create a file and want to be notified when someone reads, tags, comments, or updates it; however it doesn’t address the accidental or serendipitous. This is where I need to apply analytics to identify those facts that may be interesting to me, when they are interesting, and even when the connection may be unobvious. This is where the knowledge graph, derived from social and non-social data from across the organization and beyond, could play a key role. I wrote a somewhat controversial blog post last year which went a step further and even suggested that these manually constructed networks were less than productive and all the interactions should be analytics-driven. Perhaps a step too far, but I suspect the right answer is somewhere in between; Managing Communities & Social Networks @ Web Scale: Pushing water uphill?
Regarding the “when they are interesting” reference above, one key characteristic of good social analytics is that its not just about identifying interesting information, its also about the context and timeliness. I want the system to tell me about a great sale on baked beans when I’m preparing to go shopping. I don’t want it to tell me when I am in the middle of reading a good book. And its exactly the same in the workplace.
I’ve been giving this topic quite a bit of thought recently and will be writing more on this in the coming months, but I just wanted to share this short note since it was rattling around my head :-)