Root-cause analysis when Facebook Stories usage grows but ad revenue shrinks
Company: Meta
Role: Product Manager
Category: Analytics & Experimentation
Difficulty: medium
Interview Round: Other
You are a product manager at Meta working on Facebook Stories. Engagement with Stories keeps climbing, but ad revenue is going the other way. Explain what you would do: how you would check the numbers, work out why the two metrics are moving in opposite directions, and decide what to change.
```hint Follow the money path
Engagement turns into revenue only through ad impressions and the price each impression sells for. Break revenue into those steps, and ask where the extra engagement is coming from.
```
### Clarifying Questions
- Is the falling revenue from ads shown inside Stories, or Facebook's total ad revenue?
- How large are the two changes, over what period, and did they start suddenly or build gradually?
- Which engagement metric is rising: people viewing Stories, stories viewed per person, time spent, or stories posted?
- Is the pattern global, or concentrated in certain countries, platforms or types of advertiser?
### What a Strong Answer Covers
- Pinning down which revenue and which engagement metric are meant, and validating both data sources before diagnosing.
- A decomposition of ad revenue into impressions and price, with each factor tied back to Stories engagement.
- Checking whether people's time moved into Stories from surfaces that monetize differently.
- Internal causes (ad load, ad delivery, formats, measurement) separated from external ones (advertiser demand, privacy changes, competition, currency), narrowed with segmentation.
- A concrete recommendation, including an experiment whose guardrails protect long-term engagement and advertiser results.
### Follow-up Questions
- Suppose a higher Stories ad load restores revenue but slightly reduces stories viewed per person. How do you decide whether to ship it?
- Advertisers say their Stories ads underperform because they only have horizontal Feed creative. What would you build?
- How would you set up a long-term holdout to learn whether a higher ad load hurts retention months later?
Overview: A metric diagnosis question about Facebook Stories, where engagement keeps rising while ad revenue falls. It tests validating the data, decomposing ad revenue into impressions and price, checking how time moves between surfaces, separating internal from external causes, and recommending an experiment with guardrails.