With only a few days to prepare for the 1st round interview with McKinsey, I listened to your material (LOMS) and was invited to the 2nd round.

Now, I’m struggling with the change in the interview format (interviewer-led x candidate-led). I’m having difficulty prioritizing the areas of analysis (in the 1st round I didn’t have to prioritize which areas should be analyzed, instead, I only mentioned what should be analyzed).

How should I prepare for the 2nd round interviews? (I have a week ahead).

Thank you!

My Reply:

The key in the candidate led case, assuming there is a format change, is to let the hypothesis determine what gets prioritized. You ask yourself which factor will likely have the biggest impact on the hypothesis.

If it is unclear, you gather some data and then re-assess. Here’s a simplistic example: Profits have declined by $100 million, and you find sales are down $99 million and costs have gone up by $1 million.

While the decline in sales and the increase in costs both contribute to the decline in profits, they do not do so equally (it’s extremely rare in a multi factor situation for all factors to contribute equally). In this case, it makes more sense to focus on the factor that’s making a bigger impact on the problem — in this case, focus on the decline in sales.

In cases that aren’t as mathematically “neat” as a profitability case, try to use common sense to determine where to start first. In many cases, it’s not initially obvious what is most important. So the starting point is somewhat arbitrary.

But, at some point in the case, it becomes clearer what is and is not the main contributing factor. At that point, stop the case, restate the revised hypothesis, revise your factors needed to test the revised hypothesis, and start again with a revised hypothesis and issue tree (see the chapter on issue trees in my book Case Interview Secrets ) for a more detailed explanation.

Other times, there’s information that implies one factor might be more important than another. For example, for a new technology, a big factor is: do customers actually want to buy it? That is a fundamental issue. For a mature market, where costs have skyrocketed recently, analyzing customers often is not a priority, as they have nothing to do with costs. More likely, you want to analyze the client’s costs and compare that to competitors’ costs to see if it is a client-specific issue or an industry-wide issue.

If it’s client-specific, there’s no point in analyzing competitors first, better to understand the client’s operations first and only analyze competitors as needed as a comparison point. If it’s an industry-wide problem, no need to analyze the client’s operations first because the issue is impacting every competitor. It makes more sense to analyze the industry as a whole first, then perhaps circle back later to analyze the client once you understand what’s going on in the industry.

Good luck!