Question:
Quick question: when is the best time to do the synthesis + hypothesis combination in interviewer-led interviews? I know that you would normally do it after closing each branch/moving to the next framework in candidate-led cases, but what about the interviewer-led cases? Do you do it before each new question is asked? After? At another interval? I tried to do the synthesis + hypothesis combination in a few of my practice interviews, but it seemed very forced, e.g.:
Me: “We know that the market conditions are favorable for McDonald’s to enter region X, so my hypothesis is that McDonald’s should enter, however I would like to test the hypothesis by looking at the risks.”
Interviewer: “No, let us discuss the products with which McDonald’s should enter region X.”
The type of example I said occurs with almost all the hypothesis statements I make in interviewer-led cases.
So, when would you recommend to do the hypothesis statement? Would you recommend it at all for interviewer-led cases?
My Reply:
For an interviewer-led interview, you do not want to force it. The time to do it is typically at the beginning and anytime they say something like, “what factors would be important in this decision…” or “what are the key issues…” or “how would you approach this problem…” or “how would you structure this problem”… or “how would you determine what makes sense here…”
Within the interviewer-led interview, this often comes up in the first 5 – 10 minutes of the interview. After that, the questions asked are not problem-structuring questions, but rather “point” questions that are super specific. In those cases, you do not state a hypothesis, you just answer the specific question.
Also in an interviewer-led case, sometimes the interviewer will tell you or imply the hypothesis they want you to prove. In those cases, it is not necessary to restate the hypothesis (which conversationally can feel forced). Instead, silently think to yourself what’s your hypothesis and then structure it.
The main point of the hypothesis isn’t to say it, but rather to force yourself to make sure your structure will actually prove or disprove the hypothesis. Too many candidates just list a bunch of factors of things that might kinda, sorta be a little important… without ever critically thinking if each factor can logically prove or disprove the hypothesis. This leads to extremely inefficient problem solving. Now I err on having candidates state the hypothesis out loud, not because the interviewer needs to hear it, but because the candidate needs to be reminded of it — so the rest of the process in the case goes more smoothly.
As for the synthesis, at the end of each question, it is useful (but probably not mandatory) to tie your answer back to the specific question the interviewer just asked, to the overall question the client was concerned about in the entire case. It can be just a sentence or two, just to show that you still remembered the overall objective the client had in mind and didn’t get lost in the details.
Give that a try and let me know if it feels more natural for you, or if anything else seems forced. It should definitely not be forced ever.
0 thoughts on “Hypothesis Statement in Interviewer-Led Interviews”
Hi Victor,
I was wondering, in profitability interviewer-led cases, is it ok if our hypothesis is just “I think profitability is declining because of an increase in costs”, or should we include both options: “I think profitability is declining because of an increase in costs and decline in revenues”.
I think, the first option is much easer to analyze and structure, but the problem is that I feel like in interviewer led cases I will never have the opportunity to try another hypothesis if the one I state is wrong. Usually it starts with the question “what factors would you analyze” and once I state my approach we pass into another completely different question. But what if my hypothesis was that costs had increased but the reality is that revenues have decreased? will the interviewer think I’m wrong and eliminate me from the process?
An explicit hypothesis is more appropriate in a candidate led case where you have a few cycles to find the dead-end and revise your hypothesis to go down a different branch. In an interviewer led case, I’d do the following. If they ask you explicitly for a hypothesis, you need to provide one. If they don’t ask and only ask for what factors would be important, then I’d use a few silent hypotheses in my head and figure out everything I need to know to disprove the various hypotheses I have in mind.
So if I think it’s cost, I would need to know change in cost and change in revenue in order to disprove that hypothesis. If I think it’s fixed costs, I would need to know changes in fixed costs and per unit costs to disprove the hypothesis that it’s driven by fixed cost. In essence, they want you to layout the entire issue “tree” with all of its branches.
My profitability framework laid out visually is an example of an issue tree.
-Victor