I just came back from ScrumGathering in Munich, and one of the hot topics was teams. How they are structured, and how to make them most efficient.
Scrum states that teams should be self-organised, and this will allow them to form into the most efficient team possible given the set of people in them. But it doesn’t go into any detail on how this works.
This is where the talks at Scrum Gathering picked up. It was immediately apparent that there was not much agreement in what is meant by a self-organising team. Several talks defined this, and one focused on providing a specific and general definition.
Summarising what the general ideas seems to be, I get the following:
- The team should be given goals, and scope of work to be done. This cannot be self organised and is outside the scope of the team.
- When a team self organises, roles, behaviours and processes emerge naturally. These identities (to give them their generic name) are not possible to predict, and may or may not be efficient. Therefor this will not always be a good thing.
- Self organisation takes up energy – unless you apply more energy to the team, the self organisation will stop. It may not stop in a useful or efficient state.
- Changing external factors (timelines, scope, physical environment etc) or internal factors (team make up, rules and so on) will cause a change. This could be good or bad.
Several talks focused on how you can influence the emergent behaviours, roles processes, in order to make the team more efficient. I guess this means able to produce large amounts of well written code while still being happy at work.
One very abstract mode, ABIDE, was presented, as well as two more practical models, “Anderson’s Seven Levers” and “The Contains Differences and Exchanges Model” (both in Mike Cohen’s presentation). To my mind, both of these are implementations of ABIDE. They provide ways in which you can modify the inputs to a team, to generate new emergent behaviours, and the idea that you have to move very slowly. You can’t know what new behaviours will emerge as a result of your changes, so you have to constantly stop and observe if what your doing is creating good or bad effects.
There were also a couple of ideas how teams change and evolve over time. You should cook a team, but not burn them, as well as the classic, Storming Forming Norming and Performing
Personally I find this all very interesting, and there is no doubt that having been introduced to these models helps understand how your team dynamics are working (or not). But its not clear if that makes it any easier to create a team that works well. It seems to me quite surprising that we ever brought into the idea that leaving a team alone and not interfering will create an efficient team. At the very least it requires an outside, objective observations, which feed into the team, as well as active facilitation (intervention)
