Estimation question

Estimate the number of piano tuners 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

The classic Fermi estimation structure: decomposing an unknown quantity into estimable, multiplied factors.

Start with San Francisco's population, roughly 870,000, and convert to households, assuming average household size 2.3, giving roughly 380,000 households.
Assume the share of households owning a piano, say 3 percent, giving roughly 11,400 pianos in homes, then add institutional pianos from schools and concert halls, adding a smaller estimate like 1,000.
Assume each piano needs tuning once a year on average, giving roughly 12,000 tunings needed annually.
Assume a piano tuner can complete about 4 tunings per day, working around 250 days a year, giving roughly 1,000 tunings per tuner annually.
Divide total annual tunings by tunings per tuner: 12,000 divided by 1,000 equals roughly 12 piano tuners.
State the final estimate as a range, roughly 10 to 15 piano tuners, acknowledging the sensitivity to the piano ownership assumption.

Breaks the problem into clear, labeled steps: population, ownership rate, tuning frequency, tuner capacity.
Uses round, defensible numbers rather than false precision at each step.
Adds institutional pianos as a secondary source, showing more complete thinking than households alone.
States the final answer as a range and names the most sensitive assumption.

Skipping institutional pianos entirely and only counting home pianos.
Giving one number with no shown decomposition, which is the entire point of this classic question.

Chapters of the AI PM course that teach what this question tests.