Thank you, Clara and Ian! I gave it a try. First, I’ll be more specific here and say we’re talking about pharmacies/drug stores in the European sense (so no CVS, etc., but “pure pharmacies”). I think that I’m mixing things but let’s see. I came up with two ways to solve this. I think that the first one is more suitable for pharmacies as they’re not like other businesses, I’d approach it geographically, but I made it too simplistic. The other approach would be probably more suitable for restaurants and I feel like that’s not the way to go. 1) This is too simplistic. I could decompose the number of pharmacies as (average # per city) x (# cities in the country). The average might be around 15 pharmacies. Larger cities can have much more, smaller cities are more numerous and usually have fewer. For the cities, I will assume that half of the US population lives in cities and that the average population is 150-200 thousand. Again, small cities are much more common than large ones, which is the reason for my estimate. This gives me an estimate of (350M/2)/175k = 1000 cities So, I estimate 15,000 pharmacies in total. 2) 1. I’ll segment the population into young (0-18), middle-age (19-65), and old (65-100). The age distribution is approx. 25% of young, 60% of middle age, and 15% of old population. > we have 87,500 young; 210,000 middle age and 52,500 old (Population is 350,000) 2. Frequency of pharmacy visits per age: I’ll assume young need something in a pharmacy once every 2 months, middle-aged once every 3 months, old once every 1 month, I’m neglecting tourists. > 87,500/(3x30) = 1,458 young people need something in a pharmacy every day; 2,333 middle age and 1,750 old every day = 8,263 people need something in a pharmacy every day 3. Now I’m stuck. I don’t think I should calculate how many people can be served by a pharmacy in one day (as their distribution is not based only on profits but also based on “needs”, and they can be full in peak hours but not all the time). Does the first approach make sense and should I just break it more down? Or should I somehow try to develop the solution for the second approach? Thank you!

(edited)