Estimation question

You have a grocery delivery service (similar to Blinkit or Instamart) that ships pan India within 24 hours. Without searching online for specific details, estimate how many 1-ton trucks you need to operate this service using the power of logic based on assumptions (e.g., assuming that India has 1.4 billion people of which you have a 25% market share).

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

Multi-step estimation using explicitly given assumptions, testing whether the candidate follows and builds on provided numbers correctly.

Use the given assumptions: 1.4 billion population and a 25 percent market share, giving roughly 350 million addressable people.
Estimate the share who actually order groceries via this service regularly, assuming 10 percent are active users, about 35 million people.
Estimate order frequency, assuming each active user orders about once a week, giving roughly 5 million orders per day.
Estimate orders per truck per day: each 1-ton truck carries about 150 orders per run and completes about 3 runs a day, so 450 orders daily.
Calculate: 5 million divided by 450 gives roughly 11,000 trucks needed nationally.
Sanity check the order of magnitude and state the final estimate as a range, roughly 9,000 to 13,000 trucks.

Builds directly and correctly on the assumptions explicitly given in the prompt rather than inventing different ones.
Adds the necessary missing steps, active user rate, order frequency, and truck capacity, with clearly stated assumptions.
Shows the full calculation chain transparently rather than skipping to an answer.
States a final range and sanity-checks it against a plausible national delivery fleet scale.

Ignoring the assumptions explicitly given in the prompt and inventing different ones.
Skipping the truck capacity and runs-per-day step, a key part of converting orders into a truck count.
Presenting an overly precise single number with no range.

How would this estimate change during a high-demand period like a festival season?
What would you assume differently for last-mile versus long-haul truck routes?
How confident are you in the truck capacity and runs-per-day assumption?