Root-Cause a 10% Decline in Messenger Daily Active Users
Company: Meta
Role: Product Manager
Category: Analytics & Experimentation
Difficulty: medium
Interview Round: Other
You are a product manager at Meta who owns Messenger. One morning you check the dashboards and see that Messenger's daily active users (DAU) are down 10%. The interviewer asks one thing: why? Explain how you would track down the cause, treating the interviewer as the person who can answer your questions about the data, and say what you would do once the cause is confirmed.
```hint Locate it before you explain it
A tenth of a very large user base did not show up. Before offering theories, work out when the change began, where the missing people are concentrated, and whether the number can be trusted at all.
```
### Clarifying Questions
- Down 10% compared with what: yesterday, the same weekday last week, or a forecast? Did it appear in a single day or build up over several weeks?
- What counts as a Messenger daily active user: opening the standalone app, sending or reading a message, or any Messenger activity including chats inside the Facebook app and on the web?
- Is the decline spread evenly, or concentrated on certain platforms (iOS, Android, web), app versions, countries or age groups?
- Did Facebook, Instagram or WhatsApp see a similar drop over the same period?
### What a Strong Answer Covers
- Confirming the drop is real before diagnosing it: logging and pipeline health, metric-definition changes, and agreement with independent signals such as server-side message volume.
- Characterizing the drop by its shape over time and by segment (platform, app version, region, user tenure, entry point) to shrink the search space.
- A structured hypothesis list spanning internal causes (releases, bugs, outages, push delivery, experiments, changes in the Facebook app that send people into Messenger) and external ones (seasonality, OS or app-store changes, competitor moves, regulation, press).
- Decomposing DAU into new, retained and returning users, and using companion metrics to tell a reach problem from an engagement problem.
- Ordering the checks by likelihood and cost, and naming a concrete action and owner once a cause is confirmed.
### Follow-up Questions
- Suppose the whole decline sits on Android, on the newest app version. What do you check next, and who do you pull in?
- DAU is down 10%, but messages sent per remaining active user went up. How do you interpret that?
- If the cause turns out to be a gradual loss of younger users to a rival messaging app, what would you propose for the next two quarters?
Overview: A product execution question about Meta's Messenger: daily active users are down 10%, and the candidate must explain how they would find out why. It tests validating the metric, segmenting the drop by time, platform and user group, building internal and external hypotheses, and choosing a next step.