Katkat
Back

How would you drive product growth?

MetaProduct Analyst · Onsite
MediumProduct AnalystOnsiteAnalytics & Experimentation

Last updated

This question evaluates product growth analytics competencies including experimentation design, funnel decomposition, metric definition and guardrails, segmentation, and causal inference within the Analytics & Experimentation domain for social media and mobile product features.

Assume you are interviewing for a Product Growth Analyst role at Meta. Answer the following product growth and analytics cases. For each case, clarify the goal, define the primary metric and guardrails, break the problem into a funnel, identify what data you would inspect first, propose hypotheses, prioritize one intervention, and explain how you would validate impact.

  1. Improve Instagram Stories viewer engagement.
  • How would you improve the Instagram Stories viewing experience?
  • What data would you examine first?
  • What product ideas would you generate?
  • Pick one idea and explain how you would test it.
  1. Double Boosted Posts MAU in 6 months.
  • Assume Boosted Posts MAU is defined as the number of distinct users or businesses that launch at least one boosted post in a calendar month.
  • How would you determine whether doubling MAU in 6 months is realistic?
  • Which levers would you target across acquisition, activation, retention, and resurrection?
  • What trade-offs and guardrails would matter?
  1. Grow the number of active Reels creators.
  • Define what an active creator means.
  • How would you diagnose supply-side bottlenecks in the creator funnel?
  • What interventions would you prioritize to increase durable creator growth rather than only short-term posting spikes?
  1. Increase logged-in users on the mobile Facebook app.
  • Logged-in user count on the mobile Facebook app is declining.
  • Focus your analysis on the login funnel, session persistence, password reset, and authentication friction.
  • How would you determine whether the issue is a measurement problem, a product bug, a security or anti-abuse trade-off, or a real behavioral change in users?

In all cases, discuss segmentation, possible confounding factors, and when you would use an A/B test versus a quasi-experimental or diagnostic approach.