Estimation question

Estimate the number of groups on Facebook

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 structured estimation using known user-base scale and reasonable group-membership assumptions.

State the goal: estimate the total number of active Facebook Groups worldwide.
Start from a known anchor: assume roughly 2 billion monthly active Facebook users.
Estimate group participation rate: assume 50% of users are in at least one active group.
Estimate average groups joined per participating user: assume an average of 3 groups per participating user.
Estimate average group size: assume a typical group averages 500 members, spanning small local groups and larger communities.
Compute: 1 billion participating users times 3 memberships divided by 500 average size, roughly 6 million groups, then sanity-check the order of magnitude.

Anchors the estimate in a known, defensible number, total MAU, before layering assumptions on top.
States a clear formula, total memberships divided by average group size, rather than guessing a group count directly.
Uses round, explainable assumptions, 50% participation, 3 groups per user, 500 average size, instead of false precision.
Sanity-checks against Facebook's own public statements about Groups to validate the order of magnitude.

Guessing a total number directly with no formula connecting it to user base or membership patterns.
Ignoring that group sizes vary hugely, from a handful to millions, without an averaging assumption.

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