Agile CMMi or Agile Competing Competencies
In the early 1990’s, when I was at the Univ of Michigan, I had the good fortune to be selected to write the initial software that helped Bob Quinn and his team (including Dan Dennison) in the development of their Competing Values Model for assessing organizational effectiveness. The results of their research culminated in their book Becoming a Master Manager: A Competency Framework.
Working on that software gave me a strong appreciation for the notion of strong, empirically validated models for assessing organizational effectiveness. A key part of this is the notion that effective leaders must balance and manage competing values and competing comptencies. An example of competing roles are the Producer Role and the Mentor Role. When acting as a Producer, a leader must, well, produce results. When acting as a Mentor, the leader must mentor others in the production of the results. Both roles take skill. Choosing which to employ and to what degree requires wisdom.
Which is one of the reasons that I was drawn to Agile in the first place. It is helpful to periodically reread the Agile Manifesto (do it now if you can’t recite it from memory!). Note that the awareness and apprecition of competing values is present in the manifesto, and that the principles of the manifesto can be realized through many different kinds of practices. We’ve also learned that some of the practices matter more than others (e.g., making sure that each spring or iteration is the same duration is pretty important). Organizing these as “smells” instead of formalized assessments may be useful for some tribes within the Agile community (e.g., Mike Cohn’s Toward a Catalog of Scrum Smells).
We should also consider the CMMi. Briefly, I’ve been reading some of the recent literature on Agile and CMMi, notably David Anderson’s work (early work here, his much more recent, jointly developed SEI Technical Note here). I find the CMMi’s structured collection of practices that describe the characteristics of effective processes quite useful. And I think there there is a way to leverage this kind of framework in Agile projects.
The assessment concept of the CMMi has also been discussed in the Agile community. Consider, for example, the Nokia Test, initially developed by Bas Vodde and tuned by Jeff Sutherland. I think the test is quite useful, and, without realizing, created a similar test for our clients (I should have searched for theirs before developing ours!). I especially appreciate its directness and simplicity for “essential” activities. It appears that there are assessment and response to assessment relationships that are forming between CMMi and Agile/Scrum, and that’s good. I prefer assessments to trying to detect smells or identify anti-patterns because the detection of a smell requires judgement when judgement is likely lacking, and the notion of cataloging all of the ways that you can fail in anti-patterns seems like a task that might never end, as there are an infinite number of ways to fail.
At present, these assessments seem to be focused on “core” or “essential” practices. While the obvious place to start, I think the maturation of the Agile Community will quickly draw us into more sophisticated assessments that enable Agile teams to improve their application of Agile practices that we know to be useful based on data. And we’ll want to have guidance as to which practices, and to what degree of skill, a team should exectute.
This requires wisdom / judgement along with an appreciation of when/how to execute specific practice areas. And for me, it will also require the integration of some of lessons that Quinn and others have shown us about competing values and competing competing roles and how these impact Agile teams.
Thoughts?
