If you are lucky, you have had the opportunity to work with a high performance team - a small group of people that accomplish in a few months what larger groups have been talking about for years. Once a team like this gets going, it delivers results on its own, like it or not, support it or ignore it, lead, follow, or get out of the way. This team doesn't require much direction because each of the team members will step up to lead some aspect of the project. My friend Doug Smith wrote a book called "The Wisdom of Teams" which analyzes the circumstances where these teams form.
What's the recipe?
The key thing that ties these teams together is that all of the team members share a single goal. It helps if the goal is concrete, and clear to people inside and outside the team - for example, a product release.
Team members should be allowed to lead. If you structure a team so that there is a strong project manager, and enforce centralized task assignments, you won't get a high performance team. You will get a bunch of people who help a "single leader". This explains why outsourcing teams with strong project and task management are rarely self-managing. Smith and Katzenbaum note that both approaches - the team approach, and the single-leader approach - can deliver good results, but the single-leader approach will not give you an independent, self-managed, high-performance team, and it puts a lot of pressure on the quality of your single manager.
I add one more ingredient to the mix. I think it is important that all of the team members actually be working on the same thing. That is why I ask for daily commits and daily builds. You won't get a team if people can work independently for more than a few days. Work on the same thing.
The book's analysis reveals factors that do not seem to influence the formation of a high performance team. Team members don't have to like each other. They just need to share a single goal. In my experience, team members don't even have to know each other.
From this perspective, team building exercises that focus on creating interpersonal interaction are a waste of time. You will get better results from a clear and compelling goal. Success, as recognized by teammates and the world at large, is the ultimate interpersonal motivator.
Agile development methodologies like the one we run at Assembla are pretty good at developing high-performance teams. If everyone knows they are going to be judged by the quality of the daily build and release iterations, they start to share the same goal. If team members can pick their own tasks, they can exercise initiative, and you are on your way.
I don't have a lot of time today, but I was thinking about this and wanted to share it quickly.