Sei sulla pagina 1di 3

CASE INTERVIEW

V.CHENG
Clippets from Chengs Emails
What interviewers pick-up in the first 5 minutes

The #1 thing I notice is whether or not the candidate has a problem-solving plan (or do they

just ask lots of questions in no particular order for no obvious purpose) BEFORE they start

the problem solving.

The second thing I notice is whether or not the candidate is efficient in their use of

questions. In a real live client engagement, each time you ask a question, it costs several

thousand dollars of time and travel to get the answer.

I'm immediately impressed by a candidate that only asks the bare minimum number of

questions needed to solve the case -- no pointless questions, no questions where the

answer doesn't impact the overall recommendation, no questions asked without a specific

intent behind why the question is important.

Then a candidate solves a case quickly, it is NOT because they speak fast. It is because they

zero in with an intellectual intensity on THE key questions to ask that reveal the answer to

the case.

These are the things I notice in the first five minutes of a case.

Calculations
The case interview is NOT a computational test. It is a thinking test that happens to involve

computations. So the first decision to make (with regards to when to perform a calculation
during an interview) is: Do you actually have to compute anything at all? How do you make

that decision about whether or not to make a computation?

Knowing when or when not to perform a calculation is only the first part of how to master

the math portion of your case interview. There are many other strategies and techniques to

learn and understand in order to successfully pass the case interview.

Significance of the numbers

The math itself is not that complicated. What gets tricky for people with no business

background is how the math is intertwined with commonly-used business concepts. I've

found the candidates without business backgrounds (e.g., science, engineering) can easily

do the arithmetic required in a case, but don't really grasp the relevance of their

computations.

For example, when they calculate market share for a client (the share of total industry sales

that the client generated), and the answer is 35% market share -- they get the number

correct, but don't know if 35% is good or bad or something else.

Hypothesis Statement: 6min rule

If you meet a client for the very first time and immediately tell them your hypothesis as to

what is wrong with their company BEFORE you ever ask them any questions, they tend to

look at you with suspicion and distrust.

But, it is DEFINITELY better to state your hypothesis immediately up front than it is to forget

to state the hypothesis at all.

"The 6 Minute Hypothesis Rule"

2
In a 30 - 40 minute interview, if you have not stated a hypothesis by the 6th MINUTE of the

interview, you're probably making a mistake and at serious risk of forgetting to state the

hypothesis entirely. My suggestion is: Regardless of what you have discovered by the 6th

minute of the interview, state your best guess hypothesis at that point and just work with it.

Otherwise the risk is just too high that you'll get distracted by some surprising aspect of the

case and forget the hypothesis entirely.

Keep in mind, you can and SHOULD revise the hypothesis as you progress in the case.

So it is not THAT critical that the initial hypothesis be perfect.

One caveat to this.

In the McKinsey interview format that most offices are using or are migrating towards, the

interview process is artificially broken up into semi-independent sections.

The whole case study interview consists of about 5 sections (about 6 minutes each), each

focusing on a different aspect of the same case - sort of jumping around from one part

of the case to another.

Usually there is a section on "What's your gut intuition as to what is going on here?".

Another section would be the interviewer giving you his/her hypothesis, and then you

structure it with an issue tree.

Potrebbero piacerti anche