This post is about self organising models of organisational leadership … and is nothing at all to do with cloud computing. It will also raise more questions than answers as I don’t believe the jury is back in regarding the best way to structure and manage an organisation.
A while back I cam across this great post on cloud leadership by Richard Antcliff of the NASA Langley Research Center. The problem being solved is one in which the Langley Research Institute acts as a service provider to the rest of NASA, and needs to find a way of evolving those services over time whilst keeping the enormous levels of bureaucracy inherent in a near 100 year old government organisation.
Let’s put it another way. NASA decide they want widgets, so they ask the research people to do some research. Langley set up a widget department and get on with it. A year later, NASA want something else. How much red tape do you think the new widget department has managed to wrap itself in by now and how quickly will it be able to disband so the institute can do something else? I imagine it was a bit like wading through molasses.
The answer? Borrow from Semco and Gore to restructure around some fundamentally different principles:
- There is no fixed org chart: Teams form around interests and demand, and then disband when they are done
- There are no bosses: leaders are self elected. If no one wants to follow them, the team does not get off the ground. In short, leaders must lead
- There are no reporting lines: staff float around and work on whatever takes there fancy. If a team sounds interesting, they ask to join up
I was wondering about the balance between having no org structure at all, and having some broad functions or structural elements to enable the business to function. I ran across this extract form the Semco survival manual:
We believe that organizational structure is required to ensure good business processes. However, only people who have respect for their followers can be leaders. Situational leadership will always be stimulated and respected.
So it seems to me that some flexibility may be required as part of the business model, but there will still be particular functions that need to be performed, and particular business processes that need to be executed in order for that business model to work. So what is the ideal balance?
Gore describe their organisational structure as being a flat lattice. Key points include having sponsors instead of bosses and leadership emerging through people (or associates in Gore-speak) being followed. They also talk about associates being held responsible to their peers instead of to bosses (this form of personal commitment approach was discussed in the MiX recently, I forget exactly where). But again there does seem to be a need for some structure, it’s just a fair bit less hierarchical than the old school models.
So here is my question. Is any of this compatible with the view of the business as an entity designed to execute a validated business model or are we just making life difficult for ourselves by confusing startups and going concerns? In particular, at what point do we need to commit to a course of action and design efficient structures that just get on with it?