Estimation question

How much storage space do you need for Google Maps?

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

Technical estimation: reasoning through data volume for a real-world mapping system.

Clarify scope: are we estimating raw map data (roads, imagery, POIs) or per-user cached/offline data. State the assumption.
For raw global data, break into layers: road network data, satellite/street imagery, and points of interest, estimating each separately.
Estimate road network data using assumed total road length worldwide and bytes per km of vector data.
Estimate imagery as the largest component, using assumed tile counts and resolution across zoom levels.
For per-user offline storage, estimate a typical offline city download size and multiply by typical number of saved regions.
Combine into a total range and sanity check against known facts, like Google Maps offline downloads being a few hundred MB per city.

Separates global backend storage from per-device offline storage, since the question could mean either.
Uses a known reference point, e.g. an offline city map download being roughly 100 to 500 MB, to anchor the estimate.
Shows the multiplication chain explicitly rather than jumping to a final number.
States the dominant cost driver is imagery, not vector road data, and explains why.

Not clarifying whether the question is about global backend storage or a single device's offline cache.
Ignoring that satellite imagery dwarfs vector data in storage size.

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