Style of Bain cases

Question

Hey everyone - preparing for Bain interviews and from conversations with a few friends who have gone through the process last year, it appears Bain interviews do take on some elements of being "interviewer led", e.g. for cases where the focus is on one specific issue / requires a deep dive on one branch of the expected issue tree. Can anyone who has gone through their process within the last 2 years verify or comment on this?

Also, if anyone is preparing for Bain, feel free to drop me a msg if you would like to practice.

Cheers.

Answer 1

Hi,

First of all - it is a myth that at McK and BCG all cases are Interviewer led. I know a lot of partners who don't hesitate to ask the interviewee lead the case.

Secondly - indeed almost all Bain cases are interviewee led. Thus you have to drive the case and you have to come to a certain number and the conclusion in the end.

It may seem to you that these 2 types of cases are different, however, the interviewer-led type is just a simplified version of the interviewee-led case. My advice is to prepare a lot in the interviewee-led format so that you could solve both easily.

Now, how to approach them?

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:

Ask for a piece of data / info you've defined in a structure
Compare the data with historical trend / becnhmarks
If you find something interesting ask for the root / cause or state the hypothesis
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
Once you are done with 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

Best!