Root-Cause a 25% Jump in Spotify App Installs: Hypotheses and How to Test Them
Company: Spotify
Role: Product Manager
Category: Analytics & Experimentation
Difficulty: medium
Interview Round: Other
Spotify's app installs have gone up by 25%. What could have caused the increase? Give at least four plausible reasons, and show how you would work out which of them is actually behind it.
```hint What is being counted
Before listing causes, pin down what an install counts and what the 25% is compared with; a change in counting can look like growth. Then check that your reasons do not all come from the same place.
```
### Clarifying Questions
- Over what period, and against which baseline: the previous weeks, the same period last year, or a forecast? Did it jump at once or build up gradually?
- Is the rise global, or concentrated in certain countries, platforms (iOS, Android, desktop) or acquisition channels?
- Do the installs come from people new to Spotify, or do they include existing users reinstalling or setting up a new device?
- Did new sign-ups and first-time listeners rise in step with installs?
### What a Strong Answer Covers
- Clarifying the metric and baseline, and ruling out tracking or definition changes before explaining real growth
- At least four distinct, plausible causes that span Spotify's own actions and outside events
- A plan that tells the causes apart, naming the evidence each would leave in the data by time, region, platform, channel and user type
- A judgment on whether the growth is healthy, based on activation and retention of the new users
### Follow-up Questions
- Installs are up 25% but new monthly active users are flat. What does that point to?
- The whole increase comes from Android in one country. Which causes move to the top of your list, and why?
- A campaign appears to be driving the rise. How would you check that its installs are incremental before spending more on it?
Overview: A product analytics interview question about a 25% rise in Spotify app installs. It asks for at least four plausible causes and tests whether the candidate checks the metric definition first, separates Spotify's own actions from outside events, and uses segmentation and new-user quality to find the real driver.