I'm currently practising a lot of market sizing. I usually get the logic and can structure it correctly.
However, my result is often twice as big or twice as small due to wrong assumptions.
I know my basics (household, life expectancy for people, different products etc.) however, I still tend to make bad educated guesses regarding some assumptions.
How important is it to be within 20 percent if the structure/logic is correct and most assumptions are correct too? Will the interviewer tell me directly that one assumption is wrong and help me to correct it? If yes, will this be regarded as a weakness?
For Bain interviews: how usual is it to get random, standalone estimation questions that do not really have something to do with market sizing directly? Case in Point author states that only smaller firms tend to do that, whereas MBB ask market sizing questions within the context of a larger case.
Market sizing: off by more than 20% due to wrong assumptions -error?
Overview of answers
I’d echo the previous response.
I’d add that when you size the market from the demand side, usually it’s good to be within +-30% range of the first input. For example, value size of market of tea in China equals to population of China multiplied by % drink hot drinks multiplied by... here you should start with a proper assumption on population of China. Ideally, you should break down the equation to as many steps as you can and your inputs should be within +-20% range of the actual values. It’s fine, however, if your final output is 2x times bigger than it is in reality. If the final number seems to be too high or too low, sanity check the response and change equation input values to lower or higher respectively
Precision in final results' importance is directly proportional to your access to data i.e. expectation to be correct is present as you have more access to the data. Hence, as long as we assume most of the numbers based on our reasonings, it is OK to answer within 1/10 to 10x range. What matters is your train of thoughts, so make sure you think out loud throughout the process.
Most interviewers won't interrupt the process even if your logics and/or assumptions are wrong. That said, they may ask give you a number saying that is the correct number and ask you what went wrong in your logic and/or assumptions. As long as you can point out which part of them could be wrong, it should not be viewed as a weakness.
I have not attend Bain interview before, others may enlighten you on that.
As always with market sizing, I recommand to split the conversation in 2 to facilitate the discussion with the interviewer :
- Structure of the calculation : This one is critical, and you must have it right. That's basically what is evaluated in a market sizing exercice
- Assumptions : This is far less important than the structure. Make sure you know the basic numbers (population, etc.) but relax on the other hypotheses, we don't expect you to be an expert of the market you will size.
As said by Vlad, if the numbers are critical for the case resolution, the interviewer will adjust your hypotheses as long as your structure is correct.
You shouldn't worry about that:
- If the exact numbers are not critical for the case - the interviewers will look for your approach and justification of the assumptions rather than the exact numbers
- If the numbers are critical for the case - the interviewers will help you with the correct assumptions. If you have a solid justification it won't be a mistake