- Despite a gaggle of tools such as liking, following, and friending, enterprise social networks still struggle to derive meaning from the ever evolving nature of actual, human associations.
- In order to truly capture a user’s social graph, our software must accommodate the more malleable aspects of our human nature, such as bias, contradictions and even multiple personalities.
All movie actors are separated from Kevin Bacon by no more than six films. If you don’t believe me, just try your hand at the “Six Degrees of Kevin Bacon” game. However, as staggeringly interesting as this may be, how valuable is such knowledge?
Does the fact that Wil Wheaton has a Bacon Number of 1 really tell us anything about the nature of their relationship, aside from the fact that they both starred in “She’s Having a Baby?” What does this tell us about their relationship? Do we know whether or not Wil would be willing to spot Kevin a ten spot for a cab ride home? Can we derive that knowledge somehow from the fact that Mr. Wheaton appears in cameo just at the end of this film, in the credits, rather than staring opposite Mr. Bacon? Certainly not. And yet we pretend that our enterprise social networking software can do just that while still relying on what can only be described as a ‘Baconian’ approach to capturing the true nature of our workplace relationships.
It is not enough to define our corporate associations via likes, dislikes, follows, and unfollows. This notion was interesting and even valuable back in 2007, when Facebook began talking about the value of what it termed as “The Social Graph.” The shorter the path between two people (their Bacon Number, if you will), the more value you can ascribe to their association. If user Kevin uses a given file frequently, then it is likely that user Wil, who is directly connected to user Kevin, will also find that file valuable.
Don’t get me wrong – this basic, foundational element is currently driving some very subtle and useful capabilities such as expertise location. Our current crop of enterprise social software can both filter out noise and prioritize information that is likely to be of value to us based upon the many, always changing filaments that make up our individual social graph. What we need is a deepening focus on a few of the more malleable aspects of social networking. Here are a few examples.
- Account for Change. The values used to calculate features like recommendations must evolve in response to our current responsibilities, the past, current and future projects we work on, the travel we undertake, even the time of day (e.g., Friday crunch time).
- Reflect Multiple Personas. Our workplace social graph must literally absorb our external, personal social graphs from the likes of Twitter, Facebook, Google+, LinkedIn, etc. The same goes for external corporate graphs established by partners and customers.
- Accommodate Contradiction. A truly useful social graph must be able to understand that a relationship can be of less or more value as the context of that relationship changes (e.g., project to project).
- Consider Personal Bias. When a user “mutes” someone’s event stream posts, that action most certainly influences future algorithmic decisions.
With advances in enterprise data and analytics espoused by market leaders SAP, TIBCO, Oracle and IBM, collaboration platforms are on the cusp of being be able to move well beyond basic near and far calculations in constructing a more nuanced corporate social graph. But doing so will for the foreseeable future extract a significant cost in licensing, IT support and user training. And given the numerical nature of such endeavors, a truly smart, emergent, self organizing social graph may not even be an option for smaller companies without a statistically valid sample size. In the meantime, I would encourage customers to think of their social graph as an effective means of calculating the nearness of people, documents and groups. This may not be as nuanced as we analysts would like, but such capabilities are already a very valuable commodity for business.