Estimation question

How would you go about finding out the number of red cars in China?

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 methodology combined with data-gathering strategy, since the question asks how you would find out, not just for a number.

Clarify the two interpretations: estimate the number directly, or describe a data collection methodology, and address both.
For estimation, start with China's total vehicle count, roughly 300 million registered vehicles, a commonly cited order-of-magnitude figure.
Assume the share of cars that are red, based on general color popularity data, roughly 5 to 8 percent of cars globally.
Calculate: 300 million times 6 percent gives roughly 18 million red cars.
For methodology, describe validating this via sampling traffic camera or dealership sales data across regions.
State the final estimate as a range, roughly 15 to 20 million, noting regional variation in color popularity.

Explicitly addresses the how-would-you-find-out framing with both an estimate and a real data-gathering method.
Uses a defensible color-share assumption grounded in general automotive color trend data.
Proposes traffic camera or dealership sales sampling as a genuinely feasible validation method.
States the final range with an explicit uncertainty caveat about regional color preference variation.

Answering only with a number and ignoring the methodology angle of the question.
Using an unfounded color-share assumption with no reasoning.
Not acknowledging regional variation in car color preference.

What data source would you trust most for validating the color-share assumption?
How would you adjust for newer versus older vehicles having different color popularity?
How confident are you in the total vehicle count assumption?