Estimation question

Estimate the number of hours wasted in traffic each year.

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

Estimation skill: can you build a clear top down decomposition with stated, reasonable assumptions rather than guessing a final number.

State the scope, assume this means US commuters and errands, not global.
Estimate the driving population, assume 200 million US adults drive regularly.
Estimate average extra time lost to congestion per driver per year, assume 50 hours based on typical commute delay estimates.
Multiply: 200 million times 50 hours equals 10 billion hours per year, and state this openly as the core calculation.
Sanity check against a known reference point if you recall one, and adjust if the result seems off by an order of magnitude.
State the answer as an estimate with the assumptions listed, not a falsely precise number.

Shows the full chain of assumptions out loud rather than jumping to a number.
Uses round, defensible figures and labels them clearly as assumptions, for example 50 hours per driver per year.
Sanity checks the final order of magnitude instead of stopping at the first multiplication.

Giving a final number with no visible reasoning.
Using implausible inputs, like assuming every American drives in heavy traffic daily.

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