Estimation question

Estimate the market size for Google Fi.

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

Tests market-sizing methodology for a niche product (an MVNO), requiring a top-down or bottom-up funnel rather than a guessed headline number.

Define the addressable market top-down: US mobile phone users (roughly 300 million), narrowed to those open to switching carriers or already on a Google ecosystem device (Pixel, Android).
Narrow further to Fi's realistic target: tech-forward, often single-line or light-data users, and travelers who value Fi's international roaming pricing, estimated at maybe 10-15% of Android users.
Apply typical MVNO market share assumptions: MVNOs collectively hold a modest single-digit percent of the overall wireless market, so Fi likely captures a small fraction of even its addressable segment.
Estimate revenue: multiply an assumed subscriber count (e.g. 3-5 million) by average revenue per user (roughly $50-70/month) to size the market in dollar terms.
State the result as an order-of-magnitude range, e.g. a multi-billion dollar addressable opportunity but a much smaller realistic near-term revenue figure, and flag the biggest assumption driving uncertainty.

Uses a clear top-down funnel (total market to addressable segment to realistic share) instead of guessing a number directly.
Anchors the addressable segment in a real behavioral insight, Fi's natural fit is with Android/Pixel users and international travelers, not the general population.
Converts to a dollar market size using an ARPU assumption, since 'market size' for a paid service should ultimately be expressed in revenue terms, not just user count.
Flags which assumption (like MVNO market share) carries the most uncertainty, showing awareness of estimate fragility.

Sizing the market in users only, without converting to a dollar figure, which is usually what 'market size' means for a subscription product.
Not narrowing from the general population to Fi's actual realistic addressable segment before estimating.

How would this estimate change if Fi expanded eSATs support to non-Android devices more aggressively?
Which assumption in your estimate would you want to validate first with real data?