Nobody tells you that the case interview is less about the answer and more about whether the interviewer wants to work through a hard problem with you. McKinsey, BCG, and the strategy teams at companies like Swiggy and Meesho run case rounds specifically because they simulate the job: an ambiguous situation, imperfect data, and someone watching how you think in real time. Getting to the right answer matters less than getting there without losing the thread. That's a learnable skill. But only if you practice saying your structure out loud before you're in a room where someone's evaluating it.
Every case can be broken into four steps:
- Clarify: Ask questions to narrow the problem scope
- Structure: Create a framework (don't force-fit MECE; adapt to the problem)
- Analyze: Work through each branch with data and logic
- Recommend: State your answer, the key driver, risks, and next steps
The biggest mistake? Jumping to step 3 without doing step 1 properly.
Example'How many electric scooters are sold in India per year?'
Approach• Start with India's population (~1.4B) • Urban population: ~500M • Two-wheeler households: ~35% = 175M • Annual purchase rate: ~8% (new + replacement) = 14M • EV penetration: ~10% = 1.4M electric scooters/year
Always state assumptions, check reasonableness, and note what data you'd verify.
FrameworkRevenue (Price x Volume) - Costs (Fixed + Variable)
Always ask• Is the decline in revenue, increase in costs, or both? • When did it start? What changed? • Is it affecting the entire market or just this company?
Then drill into the specific branch that's causing the issue.
Example'Should Swiggy launch a grocery delivery service?'
Structure1. Market attractiveness: TAM, growth, competition 2. Strategic fit: Synergies with existing business, brand alignment 3. Feasibility: Operational capability, investment required 4. Risks: Cannibalization, regulatory, execution risk 5. Recommendation with conditions
The most important habit is also the most skipped: practice out loud, with another person or a recording device, not in your head. Case interviews are oral exams and your brain processes them completely differently when you're actually speaking. Candidates who prep silently are often stunned to find they can't articulate their framework under the mild pressure of being watched.
Write your structure down before you open your mouth. Even thirty seconds of silence to organize your thinking is not only acceptable: it signals discipline. Interviewers at McKinsey and BCG are explicitly looking for candidates who don't just start talking.
Mental math is a real bottleneck and it's entirely fixable with daily practice. Not because the math itself is hard, but because doing it while narrating your reasoning to another person is disorienting until it isn't. Aim for round numbers, state your assumptions, and move on: a precise wrong number is worse than an approximate right one.
Business news matters because cases are drawn from real situations: a telecom company's declining ARPU, an e-commerce player's last-mile cost problem. Reading one business story a day and practicing "how would I frame this as a case?" is a better use of prep time than memorizing frameworks.