Lean and Agile, Part 2
I’m going to continue the discussion I started in my last post.
A more pragmatic approach to discussing the relationship between Lean and Agile than a ”Lean versus Agile” debate with all its idealogical overtones is to ask when one approach might be better suited than the other to solve a particular kind of problem. Ralph Stacey’s complexity model, which is frequently referenced within the Scrum coaching community, helps frame this discussion. Stacia Heimgartner provides a good account of Ralph Stacey’s model and its relationship to Scrum in this post. She also references a Harvard Business Review article on leadership styles that in turn references the Cynefin model.
The Cynefin model breaks problems into five domains: simple, complicated, chaotic, complex and disorder. To jump to the punch line, what Ken Schwaber and others suggest (and rings true with me) is that Scrum is especially well suited for “complex” problems, while Lean is especially well suited for “complicated” problems. Complex problems are problems that are not well understood, so that it’s not clear upfront what constitues a good solution, and consequently solutions emerge over time. First you try something, and depending how well it works, you revise and try again. Eventually, you discover a good enough solution. Complicated problems have many moving parts, but its very clear up front what consitutes an ideal solution. For example, assembling all the parts of a Mercedes engine is a complicated problem. Designing the first steam engine, however, would be a complex problem.