Chang is the founder of The Case Room. She spent five years at McKinsey, where she was an interviewer, left as an Engagement Manager, and earlier got offers from McKinsey, BCG, Bain, and PwC Strategy&.
The most common way a first case goes wrong is not a bad framework or a wrong number. It is silence. The candidate thinks hard for two minutes, then says one sentence: "It's too big to reach." That might even be right. The interviewer still has almost nothing to score.
When I interviewed candidates at McKinsey, I could only give credit for thinking I heard. A right answer with no reasoning behind it looks like a guess. A wrong answer with clear reasoning often scores better, because I can see exactly where it went off and how you would fix it. Thinking out loud is not rambling. It is letting the interviewer see your route.
What to say while you think
You don't need to narrate every thought. You need four things on the table at all times.
What you are solving for. "So the question is whether this market is big enough to reach $50 million in five years." Saying it keeps you honest, and it catches a misread question before you spend five minutes on it.
Where you are. "I have the total market. Next I need the share we could realistically win." One sentence like this lets the interviewer follow you, and it lets you notice when you have drifted.
What you are choosing, and why. "I could size this by households or by stores. I'll use stores, because the client sells through retailers." The interviewer learns more from the reason than from the choice.
So what, and what's next. "That means the market alone can't get us to $50 million, so next I'd look at whether an acquisition closes the gap." Every number or insight should end with what it tells you and how it moves you closer to your recommendation.
Compare "It's too big to reach" with "To reach $50 million we'd need about 40% of the market. Leaders in this category hold 15 to 20%, so it looks too big to reach unless we buy a competitor." Same conclusion. Only the second one shows a candidate who can do the job.
How to ask for a minute
Asking for time is normal; in fact, interviewers expect it, especially before the structure. Say it plainly: "Can I take a minute to structure my thoughts?" Then take the minute. Write, think, and come back with the answer first: "I'd look at three things."
Keep it to a minute or two. The clock keeps running, just as it does in a real interview, so a short pause is smart and a long one costs you the rest of the question.
On The Case Room, you'll hear "Sure, take your time," and then quiet. The interviewer checks "Are you still with me?" only after about two minutes of silence. If it can hear you working through something under your breath, it treats that as thinking, not silence.
How to narrate a calculation
Math is where silence hurts most, because a wrong number with no working gives the interviewer nothing to help with. Use the same pattern every time.
- Say the equation in words before any number. "Revenue is customers times visits per year times spend per visit."
- Say the numbers and how you'll round. "20,000 customers, about 4 visits, $25 each."
- Say the steps as you go. "20,000 times 4 is 80,000 visits. Times $25 is $2 million."
- Sanity check. "$2 million for one location sounds reasonable."
- Say what it means. "Against a $10 million investment, that is five years of revenue before we count a single cost, so on these numbers it looks slow."
Step 5 is the one candidates skip. A number is not an answer until you have said what it means for the client.
Narrating also protects you. If you slip, the interviewer can see where. On The Case Room, if a step looks off by more than about 10%, the interviewer may ask "Can you double-check that last number?" It never gives you the right value; finding it is your job. If you stop at the number, you may hear "What does that tell us?"
How to recover from a wrong turn
Everyone takes one. What the interviewer remembers is what happens next.
Say it, fix it, and say what changes: "Actually, I used monthly numbers there, and the table is annual. Let me redo that. That makes it $24 million, not $2 million, which changes my answer: the market is big enough." No apology, no spiral. A clean correction is a good moment in a case, not a bad one.
The same goes for your structure. If an exhibit shows something your framework missed, add it out loud: "This changes my view. I'd add a fourth area, the cost of switching suppliers, and look at that first."
If you are stuck, say where. "I can see costs went up, but I can't tell which line drove it" gives the interviewer something to work with. Silence doesn't. On The Case Room you can also press the hint button: each hint points you one step further (what to think about, then where to look, then the approach) and none of them states the answer.
The habit in one line
Say what you are solving for, where you are, what you are choosing, and what it means for your recommendation. Do that on every question and your thinking gets credit whether or not the final number is perfect.
The best way to build the habit is to practice it with someone who waits for you and listens. Your first full case on The Case Room is free: sign up, open a case and try thinking out loud from the first question.
Remember: if you don't say it out loud, the interviewer can't give you credit for it.