Estimation question

How many coffee shops are there in San Francisco?

Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.

What this question tests

Whether you can build a quick, defensible density-based estimate for a familiar local business type.

Confirm scope: independent coffee shops plus chain locations like Starbucks and Peet's within San Francisco city limits.
Start from San Francisco's population, roughly 800,000 to 850,000 residents, plus daytime workers and tourists who also buy coffee.
State a density assumption, such as roughly 1 coffee shop per 1,500 to 2,000 people, reflecting a coffee-dense city.
Divide the population estimate by the assumed people-per-shop ratio to reach a total count.
Sanity check against a known reference, like the fact that San Francisco is widely known for a high concentration of independent cafes.
Present the result as a range, and flag the density assumption as the most sensitive input to the final number.

Uses a density-based method, population divided by people per shop, rather than a guessed round number.
States the density assumption explicitly and explains why San Francisco likely skews denser than an average US city.
Accounts for daytime workers and tourists, not just residents, since coffee shop demand isn't limited to locals.
Sanity checks the result against San Francisco's known reputation as a coffee-dense city.

Giving a single number with no visible method or stated density assumption behind it.
Using national average coffee shop density instead of adjusting for San Francisco's specific urban density.

How would this estimate change if you excluded large chains and counted only independent shops?
Which assumption would you most want real data on to sharpen this estimate?