Strategy question
Instagram feed impressions have dropped by 50% day-over-day. What actions would you take?
- Strategy
- Hard
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 incident response and root-cause triage under pressure: can you structure a fast, systematic diagnosis of a sudden metric drop rather than jumping to conclusions.
How to approach it
- Treat this as an incident first: confirm the drop is real, not a measurement or logging error, by checking the data pipeline before assuming a genuine product issue.
- Segment the drop to localize it: check if the decline is uniform across platforms, regions, and segments, or concentrated in one area, since that points to a different cause.
- Check recent changes first: review deployments or configuration changes in the last 24 to 48 hours, since a sudden 50 percent drop fits a broken release better than a behavior shift.
- Check content supply: verify whether content creation itself dropped versus impressions dropping from a ranking or serving bug with supply unchanged.
- Escalate appropriately: if this points to a serving bug, loop in engineering immediately for a rollback given the severity, rather than a slower iteration process.
- Define the immediate metric to watch as impressions recovering to baseline, with a post-mortem afterward covering root cause and prevention.
What a strong answer includes
- Treats a 50 percent day-over-day drop as an incident requiring fast triage, recognizing the scale rules out normal organic fluctuation.
- Checks for a data or logging error first, a step many skip, before assuming the underlying metric truly changed.
- Prioritizes checking recent deployments as the most likely root cause for a sudden, large drop, rather than slower behavioral hypotheses.
- Proposes segmenting by platform and region to localize the issue quickly, a standard incident-diagnosis technique.
Common mistakes
- Jumping straight to a slow, exploratory root-cause analysis (like surveying users) instead of first ruling out a technical bug or logging error, appropriate for the urgency implied by a 50 percent drop.
- Not checking recent deployments or changes first, missing the most likely cause of a sudden, large drop.
- Treating this as a strategy problem to solve over weeks instead of an incident requiring immediate triage and escalation.
Likely follow-up questions
- What would you do in the first hour after noticing this drop?
- How would you distinguish a serving bug from a genuine drop in user behavior?
- How would you prevent this kind of incident from happening again?
More strategy questions
- What should Facebook's next acquisition be?Meta · Strategy · Hard
- Identify a new monetization strategy for Instagram.Instagram · Strategy · Hard
- How would you grow the number of Instagram users?Meta · Strategy · Easy
- Where do you think Reels fits in Facebook's long term strategies? Why choose Instagram to launch Reels and not Facebook?Meta · Strategy · Hard
- You’re a Product Manager at Instagram and you've been asked if Instagram should stop supporting the verified accounts feature—what do you do?Instagram · Strategy · Hard
- Would you build a travel product on Instagram?Meta · Strategy · Hard
More questions from Instagram
Learn the skill behind it
Chapters of the AI PM course that teach what this question tests.
- Chapter 4: Discovery and strategy for AI products
- Chapter 9: Prove it paid off: outcomes, economics, and pricing
- Chapter 14: Get the job: the AI PM interview loop