An estimate is only useful if it survives a quick test. Interviewers often hand you a number, sometimes their own and sometimes a "colleague's", and ask whether you believe it.
- Per person or per household: divide the total by the population. $100 a year per American on board games would mean most people buy several games a year. Does that match what you see?
- Against a known anchor: compare with a company's revenue, a country's spending, or the number of stores you can picture.
- Two ways: size it top-down and bottom-up. If the answers are far apart, find the driver that explains the gap.
- Hunt the outlier: go driver by driver. Usually one share or frequency is far too high.
Say what you'd change and why. Fixing the broken assumption matters more than the exact number.