Estimation question

How many dentists are there in New York?

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

A classic market-sizing estimation testing top-down decomposition with clearly stated, reasonable assumptions rather than a memorized number.

Clarify scope: assume 'New York' means New York City (roughly 8.5 million people), not the whole state, and state that assumption.
Estimate population per dentist ratio: recall or assume a reasonable ratio, e.g., roughly 1 dentist per 1,500-2,000 residents in a well-served urban area, stating this as an assumption based on typical US dentist-to-population ratios.
Compute: 8.5M people / 1,750 (midpoint assumption) = roughly 4,800 dentists.
Cross-check with an alternate method: estimate via dental practices, assuming roughly 1 dental practice per city block-cluster in commercial areas, with 1-3 dentists per practice, and sanity-check that this produces a similar order of magnitude.
Adjust for NYC-specific factors: as a dense, high-income city with many private practices, the ratio might be somewhat better (more dentists per capita) than the national average, so the estimate could land higher, e.g., 5,000-6,000.
State a final range rather than false precision: approximately 5,000 dentists in NYC, clearly framed as an estimate.

Uses a clear top-down method (population divided by a stated ratio) and cross-checks with a second method (practices times dentists per practice) for sanity.
States every assumption explicitly (city population, dentist-to-population ratio) rather than pulling a final number from memory.
Adjusts the baseline ratio for NYC-specific context (dense, high-income, likely better-served than national average) showing situational judgment.
Presents a final answer as a reasoned range, not false precision to the exact digit.

Stating a final number with no visible reasoning or assumptions.
Not clarifying whether 'New York' means the city or the state, a common and consequential ambiguity.
Ignoring cross-checking, leaving the estimate unvalidated by a second approach.

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