Metrics question

Let’s say we want to figure out for any Facebook user, who their best friend is so we can prioritize showing their content in the News Feed. What product signals and metrics (max 5) would you use to determine who a user’s best friend is?

Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.

Start a mock interview on this question · Mock interview from a job description

What this question tests

Tests translating a vague closeness concept into measurable behavioral signals and prioritizing them for a ranking model.

How to approach it

  1. Clarify the goal: best friend should predict content the user wants to see first in News Feed, not just who they message most.
  2. List signal categories: messaging frequency and recency, profile visits, photo co-tags, comment or reaction reciprocity, mutual friend overlap.
  3. Pick the top 5 by signal density: weekly message count, message reciprocity, profile view frequency, photo co-tags, comment reciprocity rate.
  4. Weight recency higher than raw volume so a friendship that cooled off decays in the ranking.
  5. Validate with survey data (asking users directly who their best friend is) and check precision against the signal-based ranking.

What a strong answer includes

Common mistakes

Likely follow-up questions

More metrics questions

More questions from Meta

Learn the skill behind it

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

Preparing for a specific role?

Book summaries for this kind of question

Browse all 4,000+ questions in the bank