Market Sizing Case Question: How big is the market for physicians in the US?

My favourite approach (there are many as Guennael highlighted) is similar to Michael's: demand driven.

Why is it better than going for a more top-down one? Because it uses building blocks that you are more confident in estimating: you know reasonably well how often people get ill and how many people there are in the US. You know less well though how many people graduate as physicians/doctors per year. It all depends however on how much time you have for the market sizing case (demand driven solutions are more time intensive).

To add on top of Michael's answer I created a simplified tree for solving this case:

Remember, trees should always be your favorite tool for tackling market sizing cases as this. See why on our tutorial "Crack the Case (5/9) - Market Sizing Framework".

Notice in our simplified tree that:

1. There are other types of doctor visits than just when people get ill. They also resort to a doctor for routine check-ups (prevention) or when they are just suspicious of being ill
2. Always add an "Others" segment. This will make sure you account for any other cases you may not have thought of: in this case it could be doctors in the back office who assess X-ray exams and never face any patients
3. If you don't have enough time to go through the detailed calculation of all the segments in your tree, just calculate one of them thoroughly and extrapolate the others based on the first one. For example, if you calculated that there are 3 million doctors who account for ill people, add another 50% (or any other reasonable percentage) to the total to account for prevention doctors

At the end, don't forget to sanity-check your answer. You will make your sanity check more credible if you can think of a reference to compare your results against: "I known that in my city there are 10,000 doctors and it's population is 0.1% of that of the country's, so 7 million doctors doesn't seem like a bad estimation at all."

Agree with the general remarks above comments by Guennael.

Specifically, my approach would be:

1. Start by clarifying what is meant by "physician". We will just assume Physician = general medical practitioner AKA family doctor working at general practice / clinic.
2. Make and clarify with interviewer the general assumption that at any one time supply of physicians in the US will be enough to meet demand, .i.e. there are no supply side constraints given the well-functioning educational system / level of prosperity in the US etc.
3. Identify key “direct” drivers of number of physicians in the US. Not indirect drivers.    
4. Two key direct drivers that need to be broken down are the 1) the demand for physician services divided by 2) the capacity per physician
   1. Demand for physicians visits per year = Size of population * # average number of illness events” per person per year * % of such illness events, which are serious enough to require a physician * % of such illness events that actually result in a visit to physician (why would people not go to a physician if they require it? 1) Financial reasons 2) Some are too busy and just cope somehow.
      1. Increase final result a bit to adjust for the fact that 1) especially the very young and the elderly part of the population visit physicians for routine check-ups even though they are not ill and 2) people might visit physicians for guidance on a broader set of “personal health matters” than just specific illnesses.
   2. Capacity per physician per year = Average #of patients that can be attended to per physician per hour * working hours per physician per day *opening days per week * 4 weeks per month * 12 months per year
      1. Decrease final result a bit to adjust for “seasonality” / that the physicians also go on vacation / get ill etc.
5. Perform reasonableness check of final result.  

Hope you find it useful and that I am not completely wrong :)

You could approach it a varieties of ways - we don't care which one you chose, as long as it is consistent / MECE and you quickly drive to a reasonable answer. For example (off the top of my head in the time it takes me to write this):

- How many people in the US, how many clients can a MD see?

- How many people in the US, average salary of physician is X, average profit per visity is Y

- How many physicians in my town, how many people in my town, how many people in the US

- How many medical schools in US, how big are classes, how many years a typical physician practices medicine before retiring

I'm sure you can think up a few more ways to address the question on your own. Just make sure you end result is neither "10 doctors in the US" nor "250 million physians" :)

Cheers,

Guennael

ex-BCG Dallas