I understand that managers who have invested effort and money in training teams in Agile methods may want to see how much those teams are improving. There are a handful of reasonable measures to look at to see whether the organization is improving over all (which I’ve written about here and here).
You can apply some of those measures to team improvement. Are defects trending down? Is the ratio of value work to fixing work improving? Are teams improving their practices?
But, teams don’t exist in a vacuum.
Part of team improvement comes from the way a team works together, their approaches, and skills. Another part comes from how well the environment supports their efforts:
P=f(p,e) where P= performance, p=people, e=environment.
Teams can only improve so much, unless the environment also improves.
So, rather than look only at team results, also evaluate whether the environment that supports the team’s work is improving. When teams aren’t improving as hoped, this is where I start looking.
Team composition and stability. Are the teams appropriately cross-functional? What is the frequency of membership changes? If the team isn’t well designed, or isn’t a really a team, don’t expect team-level improvement.
How work flows into teams. Are teams pulling work, or is management pushing? Are stories/features well-formed and customer facing? How much new work gets added during a sprint?
Trends in dependencies between teams. Are POs working to reduce dependencies in the way they shape stories and order the backlog? Are the teams organized to reduce hand-offs and dependencies?
Clarity of the product visions and team mission. Is it clear what problem they are solving/benefit they are creating for which people? Are team missions articulated and independent?
Adaptive planning. Are POs and Management adjusting their expectations based on team capacity?
I expect team performance to improve with agile methods. But if you really want high-performance you need to improve the entire system–and that is management work.