4 ms·
They probably know how many flu cases end up in the hospital or urgent care, then extrapolate from that number. They have historical data and also geographic, a
by Lendal 5y ago
They probably know how many flu cases end up in the hospital or urgent care, then extrapolate from that number. They have historical data and also geographic, and census data, so they can make pretty good statistical estimates.
It's kinda the same math that hardware stores use to figure out how many shovels to stock in a given store each week. Nothing is a mystery if you have data.
- tombert 5y agoGenuinely curious; what kind of math are they using to stock shovels in hardware stores? I've never heard that before but it sounds interesting.
- carlmr 5y agoNot in that business, but normally you expect certain types of shovels to be seasonal. E.g. snow shovels probably sell more in winter, gardening shovels more in summer. Add to that you can usually look at weather data and refine the seasonal estimates, since in a warm winter people buy fewer snow shovels, in a cold spring, people buy fewer gardening shovels. Just put all that into some simple regression models and you probably get a fairly good estimate. The trickiest thing is figuring out how out of stock you are. How many shovels would you have sold if you hadn't run out of stock. So usually you have to be more careful with data points when you might have not registered customers that didn't get what they want in your store.
- greiskul 5y agoHuh, it's funny, I never thought about it, but if you are selling something, the best number of unsold inventory you want to have is 1, not 0. Cause if it's zero, you have no idea how many sales you lost, if it's one, you pretty much stocked perfectly.