Instagram DAU Down 20%: Structured Metric-Drop Diagnosis
Company: Meta
Role: Product Manager
Category: Analytics & Experimentation
Difficulty: medium
Interview Round: Other
You are a product manager on Instagram, and daily active users (DAU) have fallen by 20%. Walk through how you would track down what caused the decline: what you would check first, how you would narrow the search, and how you would confirm the cause before recommending what to do about it.
```hint Real decline or broken number
Before looking for a product cause, ask whether DAU is still being counted the way it was before the drop, and whether the fall was a sudden step or a slow slide. The answers send the investigation down very different paths.
```
### Clarifying Questions
- How is a daily active user defined here: anyone who opens the app on a given day, or only someone who views or posts content? Are app and web usage counted together?
- Over what period did the 20% fall happen (a single day, a few weeks or many months), and against which baseline is it measured?
- Is the decline global, or concentrated in particular countries, platforms (iOS, Android, web) or app versions?
- Are Facebook, WhatsApp or other Meta apps showing a similar decline over the same period?
### What a Strong Answer Covers
- Confirming the drop is real before diagnosing it: the metric's definition, logging and pipeline changes, and a comparison against seasonality and past trends.
- Using the drop's timing and shape (a sudden step versus gradual erosion) to narrow which kinds of causes are plausible.
- Segmenting DAU along dimensions that each implicate a different cause, such as new versus existing users, platform and app version, country, entry point and user type, and reading where the loss is concentrated.
- A ranked set of internal and external hypotheses, each paired with the data that would confirm or rule it out.
- Checking that the identified cause accounts for the full size of the drop, then recommending a fix, a rollback or a monitoring plan depending on whether the cause is internal or external.
### Follow-up Questions
- The decline is spread evenly across every country, platform and user segment. What does that tell you, and where do you look next?
- A feed-ranking change shipped the same week DAU started falling, yet time spent per remaining active user went up. How do you decide whether to roll it back?
- If the cause turns out to be external and nothing in the product broke, what would you recommend the team do?
Overview: Instagram's daily active users have fallen 20%, and a product manager must explain how they would track down the root cause. The question tests structured metric-drop diagnosis: validating the data, reading the timing of the decline, segmenting users, and testing internal and external hypotheses.