I found this question difficult to tackle. Even though I could segment the possible channels where rats are sold, estimating the number is really difficult. Would love to hear answers from everyone and understand how to tackle such unconventional questions.
Market Sizing - Guess the number of rats sold in the US?
Overview of answers
This seems like an interesting one as you could actually make the discussion quite extensive. So this is how I would go about it (if this does not make sense I'd gladly be corrected):
At the moment I can see four different market segments for rats:
- Research (univeristy labs and companies)
- Food (for other pets)
I will just do this one for now, maybe you can think of the others yourself. There is a lot of assumptions here, but I think the main point they want to see is that you can get to a number by logical reasoning. This number could be way off in reality.
With education I assume that biology and medical students at universities will dissect animals as part of a course. Let me assume that the percentage which has to dissect an animal of biology students is 100% and medical students 50%. At my university, of the 20,000 students about 3,000 are in medicine and 1,000 are in biology. This makes 2,500 animals being dissected for education at my university. Let's assume 20% of that are rats, so 500 rats being dissected. I will assume that the rats/student ratio of 500/20,000 = 2,5% is constant throughout the country. Let us assume that there are about 20 million university students in the US. By these students collectively, 500,000 rats will be dissected during their studies. Assuming the average university study period to be 4 years, this amounts to 500,000/4 = 125,000 rats per year.
Great answer by Denis.
The purpose of these questions is to somewhat throw you off balance and see how you react. The actual answer is much less relevant - I doubt that anyone knows the actual answer. The key is how you get to that answer...
So if you tackle the problem like any other market sizing, you'll be fine. So Structure, structure, structure... Denis has given a great example. Pick halfway thought-through assumptions and off you go...