Once you’ve defined the problem you’re ready to move to the next step: understand its root cause(s).
Understanding the root causes of a problem helps you focalize your thinking on solutions that can truly bring you closer to where you want to be; so, in general, it is a good idea to first diagnose the problem before looking for ways to correct it.
For complex problems, a logic tree—a why tree in this case—can help you diagnose by helping you organize and test your hypotheses. Diagnosing requires you to use inductive, deductive and abductive logic, and, in particular, you’ll need to organize the potential root causes in a mutually exclusive and collectively exhaustive (MECE) and insightful structure (even though the causes, themselves, might not be MECE but ICE).
That is, you’ll need to use critical thinking and creative thinking. To be creative requires you to diverge in your thinking by using brainstorming or other idea-creation activities; being creative, for instance using humor; and balancing judiciously optimizing and satisficing.
A useful tool for building why trees and applying MECE thinking is to think in terms of processes. Another useful approach is to use existing frameworks and trees when possible. If you prefer to stay away from issue trees, you may want to use an Ishikawa diagram.
Once you have identified all the possible root causes for your problem, you’ll need to summarize them in a set of hypotheses, prioritize their analysis, conduct the analysis using a hypothesis map, and synthesize your conclusions. The analysis requires you to link evidence—both confirming and disconfirming—to hypotheses and triangulate on answers.
As with defining problems, you can’t expect to be excellent at diagnosing problems without practicing it extensively. When you do, it is a good idea to engage with others to challenge your thinking. This includes using judiciously experts and novices, helping others be successful, and innovating when facing unfamiliar problems. If your problem is too complex, you may look at simplifying it, for instance by solving a simpler, related one, using the solution of an isomorphic problem, or converting your problem to a well-defined one.
To help you, here are case studies. The first is on whether Chris Froome cheated to win the 2013 Tour de France (see also part 2 and the conclusion). The second shows how to develop a why issue tree and test diagnostic hypotheses.
Want to learn more? Check out our slide decks on problem solving.