McKinsey / Accenture / Got all BIG3 offers / More than 300 real MBB cases / Harvard Business School
Book a coaching

What's the best way to lead a case? Is it strictly following the case or allowing anything that has a logical approach?

Anonymous A

I have been discussing with a guy about how to lead a case. I think that considering the fact we are not consulatant we should follow what the case says. This guy let me do the case with different info e different math saying that it was logic anyways. Who is right?

  • Upvotes
  • Date ascending
  • Date descending
Vlad replied on 04/10/2018
McKinsey / Accenture / Got all BIG3 offers / More than 300 real MBB cases / Harvard Business School


Several things regarding interviewee-led cases:

1) It's even more important to ask the clarifying questions in the beginning:

- Clarify the business model (i.e. how the business works and what are the revenue streams / core products or business lines)

- Clarify the objective both in money terms and timeline (e.g. Our objective is to increase profits by 5M in 5 years). When you have a to select from several options in a case - clarify the selection criteria

- Clarify other possible limitations if you feel that it's necessary

2) You lead the case through the structure you've prepared a) asking questions and trying to identify the root cause of the problem in the branch of your structure b) making a transition to the next branch c) proactively calculating the data and making data-driven conclusion from the data they give you d) Making a conclusion when they ask you to finish a case

The general algorithm may be something like this:

  1. Ask for a piece of data / info you've defined in a structure
  2. Compare the data with historical trend / benchmarks
  3. If you find something interesting ask for the root cause or state the hypothesis
  4. If no root/cause at this point available - ask for segmentation to drill down further. Make a new structure if needed to identify the root cause
  5. Once you are done with the analysis in one branch of your framework (found the root-cause / found nothing interesting) - summarize what you've learned so far and move to the next one or give the conclusion


Anonymous B replied on 04/10/2018

Keep doing more cases and try different partners and you’ll soon answer your own question.

Currently non-active expert
replied on 04/10/2018

Hey anoymous,

In interviewee-led cases, it's quite common for the interviewer to allow the candidate some degree of liberty and to go in a different root compared with what's in the guidelines - as far as it may make sense! That said, if you are practicing with other candidates, such practice can be really risky, as candidates may not have the perfect sense to evaluate whether that's a root which makes sense or not (and the interviewee might happen going completely off track).

All in all, none of you is completely right or wrong, it's just about style... but there's definitely more risk in the way your colleague is proposing (in case he/she is not very experienced in conducting interviews)