Estimation question

How would you estimate the bandwidth of Gmail for a community in Mars?

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

Creative, structured estimation under an unusual constraint, testing reasoning process over factual recall.

Clarify the scenario: a small community on Mars needing Gmail-style email bandwidth, likely limited by Earth-Mars communication constraints, not local bandwidth.
Estimate the community size and typical email usage per person per day, similar to Earth usage patterns as a baseline.
Estimate per-email size including attachments, and multiply by emails per person per day and community size.
Account for the real constraint: Earth-Mars latency (4 to 24 minutes one-way) and limited relay satellite bandwidth, which dominates the problem more than raw data volume.
Propose a batching/store-and-forward design to work within scheduled communication windows rather than assuming continuous real-time bandwidth.
State a final bandwidth estimate as a range, acknowledging relay bandwidth is the true bottleneck, not community email volume.

Identifies that the real constraint is interplanetary relay bandwidth and latency, not the community's raw data generation, which shows systems thinking beyond a naive estimate.
Uses Earth-based per-person email volume as a grounded starting assumption, clearly marked as such.
Proposes a batched, store-and-forward architecture as a practical design response to the latency constraint.
Delivers a defensible range rather than false precision, given the scenario's inherent uncertainty.

Estimating only local data volume and ignoring the dominant real constraint of interplanetary latency and relay bandwidth.
Treating this as a pure math problem without engaging with the physical communication constraints.

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