This entry cross-posted to my new blog location.
Luis Solis of GroupSystems moderated a panel on Group Intelligence Across the Enterprise: 5 Keys to Success, with David Marshak of the Unified Communications & Collaboration group at IBM, Alex Pentland of the MIT Center for Collective Intelligence and John Abele of Boston Scientific. One interesting thing was that Solis invited us to SMS him with questions during the presentation; not sure that that's all that efficient, but maybe will pull a few of the introverts out of the closet.
Solis stated that personal productivity is (so) yesterday, even though that's what a lot of this conference seems to have focussed on, and that we need to focus on group intelligence. He feels that collaboration is a too generic concept, and breaks it down into ad hoc (socializing, and 2 peers notifying or discussing) and structured group collaboration (lecturing, and many-to-many exploration/innovation). He sees the many-to-many exploration/innovation is the interesting part of collaboration for enterprises, since that leads to problem-solving, but that it's hard work because it relies on trust within the group, a supportive culture, and confident leadership. Interestingly, he sees that American culture can be anti-collaborative because of a natural command-and-control mentality, whereas other countries (he mentioned Sweden in particular) have a more inherent bent towards collaboration in their national culture that spills over into their corporate culture. It would be interesting to see where Canada lies on this spectrum.
Pentland spoke about sensor networks, and the management of face-to-face communications. Using an on-body sensor (a sort of smart name tag), they measured who interacted with whom over the course of a project (which happened to be a banking marketing campaign) by detecting proximity as well as who was speaking. Rolling this together with the email traffic that went around for the same project, they were able to look at the total communication model and correlations between face-to-face/email traffic and factors such as the quality of the group interactions; the final result is that face-to-face communication communication is better than email. Okay, we really already knew that, but for the gadget geeks, there's now a way to actually measure it. Apparently, these smart tags can also measure things such as physical gestures, which can have an impact with things such as animating an avatar in an environment like Second Life.
Marshak discussed IBM's internal community tools, which I've heard about from friends of mine who work at IBM. With more than 300,000 employees, IBM has a few internal communication issues, especially around having people with common interests or complementary skills find each other. Their intranet allows people to join groups that are geographic- or interest-based, then use that to harness the community knowledge. His example was about when he upgraded to a new version of Lotus Notes and his Blackberry no longer worked, and he wanted to find out what to do (I resisted the urge to shout out "Buy Microsoft Outlook"); he pinged the internal Blackberry group and was soon in an interactive text chat with others who either had the same problem or who had a solution. The nice thing is that there's no anonymity: everyone is accountable for their internal communications, and those involved in the communication can rate responses and even report someone else if they are using that communication method improperly. Even their Second Life group requires that you register your real identity in order to join. There's an ongoing debate about many social networking sites about using your own identity, and I strongly feel that you should use your own identify if you want to engage in a public conversation, since hiding behind a pseudonym often results in completely inappropriate behaviour (I notice that those who oppose this view are almost always the ones engaging in such behaviour).
Abele discussed methods for creating the right environment for group collaboration, which he sees as requiring diversity (in culture, experience, bias, age, geography), a means of gathering input from everyone, and a means of aggregating information. He discussed impediments to collaboration: silos and different cultures, strong egos, messenger killers, pontificators, group think, off topic time wasters, diverse levels of understanding, a range of bias (although he also listed this as part of the diversity that was required), lack of leadership, and poor organization. Many environments actually reward silo behaviour, which manifests in enterprises and academia as information or skills specializations: the better that you get at a particular specialization, the more you get paid; unfortunately, focus on a specialization usually happens to the detriment of other knowledge and skills.
We ended up with audience questions, which started some interesting discussion threads. Marshak said that he doesn't think that the goal of online communications should be to emulate face-to-face meetings: it's just different, and those differences can be leveraged.
They also discussed the problems of group-think that can occur both from cultural/gender biases and when corporate organizational structures are allowed to influence how people interact, and how anonymity can be useful in some cases to allow people to ask questions or voice an opinion more freely. However, ultimately, its much more effective to actually know who has which concerns and asks which questions. This moved to a discussion of collaboration (which typically requires that you know who's who) versus the broader area of group intelligence (which can include anonymous contributions).
Pentland has some interesting comments on how people's personal goals are no longer fully aligned with that of their employer: when people worked for the same company for life, there was near-perfect alignment, but now, almost everyone has their own career interests that may go in different directions. This can impact people's willingness to share in a variety of ways, depending on the particular corporate culture: they may be rewarded for bringing forward innovative ideas by being assigned to "cool" projects, or they may be punished for not toeing the company line.
Solis ended up with five keys to success to group intelligence:
- Be sensitive to what nature of group work you wish to improve/optimize, e.g., many-to-many
- Leadership commitment
- Measurement of collaboration and its impacts
- New tools create new behaviours, so make sure that the tools that you choose are likely to create the outcomes that you want in terms of behaviours
- Beware obstacles: think back to the list of impediments above and which are likely to occur in your organization
I'll be running a collaboration group process design session next week for a client, and you can be sure that I'll be watching for a lot of the things that I'm hearing here today.