Your product's DAU dropped 20% overnight. Walk me through how you'd investigate.
Model answer
Model answerI wouldn't panic — I'd diagnose systematically. First, I verify if the drop is real or a data issue. If real, I immediately segment the data to find patterns: which platform, which region, which user cohort? A 20% drop is almost always caused by one thing, and segmentation reveals it. Then I form hypotheses based on what changed around that time — did we deploy? Did a competitor launch? Did a notification break? I test each hypothesis with data, starting with the most likely. Finally, I communicate to stakeholders with both the diagnosis and a recovery plan with a clear timeline.
Framework · Systematic Root Cause Investigation
- MVerify the drop is real
Is this a real drop or a tracking/measurement issue? Check if analytics code is firing correctly, pipeline is working, and there are no reporting delays. - ESegment immediately to find the pattern
A 20% drop always has a pattern. Segment by: platform (iOS/Android/Web), region (geography), user cohort (new/existing), acquisition source. Which segment is most affected? - CForm hypotheses based on what changed
Timeline: What happened in the last 24 hours? Did we deploy code? Did a competitor launch? Did notification settings change? Did platform change their API? Form ranked hypotheses by likelihood. - ETest hypotheses with data
Start with the most likely hypothesis. Look for correlations: Did DAU drop coincide with the deployment time? Did it only affect one platform or region? Use data to confirm or eliminate each hypothesis. - ?Communicate diagnosis and action plan
Never communicate just the problem. Tell stakeholders: what caused it (diagnosis), what we're doing to fix it (action), and when they can expect recovery (timeline). Provide updates regularly.
Related questions
- How would you measure the success of a new AI customer support feature?
- How would you set up an A/B test for a new checkout flow?
- How would you define the north star metric for a food delivery app?
- How do you handle conflicting signals between quantitative and qualitative data?
- Our activation rate is 40%. Is that good or bad, and what would you do?