How to structure any case

Intuition

After enough cases, candidates split into two camps. One camp memorizes twelve frameworks and prays the case matches one. The other learns a single skill, how to build a structure for anything, and is never caught off guard. The second camp wins, because real cases rarely match the textbook, and the firm's actual job is structuring novel problems.

Structuring any case is just three habits applied to a fresh problem.

Framework

• Anchor on the objective. Every structure serves one decision. Write it first: "grow profit, " "decide whether to enter, " "fix the margin."
• Decompose toward that objective. Use an equation if money is involved (profit tree), a process if it's operational (the value chain), or stages if it's a decision (market → economics → execution → risks).
• Make it MECE and tailored. Buckets shouldn't overlap, should cover the whole problem, and should name the actual client.
• Prioritize and pressure-test. Pick a starting branch; ask yourself "if I'm wrong about this, where's the real answer hiding?"

Worked Example

Prompt: "A national library system is losing relevance and visitors. What should they do?" No framework fits cleanly, so build one. "The objective is to reverse declining visits. I'd structure around: who is leaving and why (segment the lost visitors), what alternatives are winning them (digital, bookstores, cafés), and what we could offer that's defensible (events, workspaces, digital lending). I'd start by segmenting the decline, since the fix depends on whether it's students, families, or seniors leaving." No canned framework, a structure built from the objective in real time.

Try another orphan prompt: "A city wants to cut traffic congestion." Weak move: reach for the 4Ps or a profit tree because they're familiar, neither fits, and the interviewer watches you force it. Strong move: "The objective is fewer cars at peak hours. Demand for trips can be reduced, shifted to transit, or spread across the day; supply of road space can be expanded or priced. I'd start with pricing, since it's the fastest lever." Built in real time, from the objective.

In the room

When a prompt fits no framework, say the anchor out loud before anything else: "Let me anchor on the objective: we're trying to reverse the decline in visits." Then name your engine and why: "Since this isn't a profit problem, I won't force a financial tree, I'll structure it around who's leaving, what's winning them, and what we can offer that's defensible." Telling the interviewer you chose not to use a standard framework converts a scary prompt into a demonstration of judgment. Finish with the pressure-test, unprompted: "If I'm wrong that students drive the decline, the answer hides in the alternatives branch, so I'll check that second." At this moment the interviewer is scoring one thing: can you build a structure, or only retrieve one?

Pitfalls

• Jamming a profit tree onto a non-profit or qualitative problem because it's familiar.
• Buckets that overlap or leave an obvious gap (forgetting risks, forgetting execution).
• A structure with no objective at the top, branches that don't add up to anything.