How to Structure a Problem - Consulting Interview

I follow a three-step approach. First, define the problem precisely — what exactly are we trying to solve, for whom, by when? A well-defined problem is half-solved. Second, decompose it into an issue tree using the MECE principle — each branch is mutually exclusive (no overlap) and collectively exhaustive (nothing missing). For a profit decline, my Level 1 branches are Revenue and Cost; Level 2 under Revenue is Price and Volume; Level 3 under Volume might be Existing Customers and New Customers. Third, prioritize — not all branches deserve equal analysis. I use a combination of impact potential and data availability to decide where to focus first. In my previous role, I used this approach to diagnose a 15% drop in customer retention. The issue tree revealed that the problem was concentrated in one segment — enterprise clients in the first 90 days — which led to a targeted onboarding fix.