Fictional practice, not an employer interview question.
Imagine two fictional pilot groups, each with 100 distinct listeners and the same seven-day observation window. Every total below includes only that group and window. A “returning listener” means someone who came back on a later day.
| Measure | Existing experience | Proposed experience |
|---|---|---|
| Listeners | 100 | 100 |
| Total plays | 1,000 | 1,200 |
| Saved tracks | 200 | 190 |
| Listeners returning on a later day | 60 | 60 |
Work through the case
The proposed experience has 20% more plays. Saves per listener fall from 2.0 to 1.9, a 5% reduction. The return rate stays at 60%. Saves per play also fall, from 20% to about 15.8%, but that ratio uses an activity count rather than people and may change when autoplay increases.
These results do not support declaring discovery better solely because plays rose. They also do not prove the proposal made discovery worse: a listener may enjoy an item without saving it. You need to define the intended experience and investigate why the measures diverged.
Prepare your application
A useful Spotify preparation project explains what a listener was trying to do and how you would tell whether the experience improved. A recommendation that generates more plays might create more useful discovery, or just more automatic activity. The numbers need a denominator and a user question.
Keep the application status separate from the practice case
Checked October 5, 2026: Spotify’s official student page says information about 2027 internship opportunities is coming soon. It does not publish an application opening or deadline in that announcement. Check an active posting’s own dates and requirements before applying. The exercise below is original practice for product, data or engineering work; it is not a reported Spotify assessment.
Choose the next question before another feature
Check whether plays were deliberately started or automatic, whether listeners wanted novelty or familiar background audio, and whether the groups had comparable opportunities to return. Without those definitions, “engagement” can blend different activities into one favorable headline.
For a student project, build a small analysis using explicitly synthetic data and include an event dictionary. Define a play, a save, a listener and the return window. Recalculate each result from the records and show what happens when duplicate play events are included. A duplicate event can increase total plays without representing another listening action.
Separate observation from a rollout recommendation
If this were a real experiment, you would also check how groups were assigned, whether events were logged consistently and how uncertain the differences are. The table alone does not establish causality or a long-term effect. A defensible immediate recommendation is to resolve the event and listening-intent questions before expanding a feature based on the play count.
A product candidate could write a one-page decision memo. An engineering candidate could build the event reconciliation; a research candidate could plan a neutral study of what listeners wanted from the session. Use only authorized data. Private listening histories are unnecessary for the practice case.
Describe the project without inventing a business result
After completing it, a clear bullet is: “Analyzed a synthetic discovery experiment, separating play counts, saves per listener and return rates; documented why one rising metric did not settle the product decision.” In the interview, explain which definition changed your interpretation and what evidence would make you recommend the proposal.
Sources and related guides
Official sources
Official sources and current instructions
The source collection was revised September 5, 2026. These links help you find the employer’s current guidance. This guide does not establish a current opening, eligibility, deadline, or interview process.
About this guide
By InternshipDeadlines · AI-assisted editorial adaptation · Updated October 5, 2026
Original practice material, not a reported employer interview question. Scenarios are hypothetical. Original fictional listening experiment; not Spotify data, interview questions or product policy. Official student announcement rechecked October 5, 2026; application dates remain unannounced.
Published by InternshipDeadlines. Guide updated . How we use sources.
Cite this guide
InternshipDeadlines. Spotify internship application: evaluate discovery beyond immediate plays. Updated . https://internshipdeadlines.com/career-guides/spotify/
Use the dates attached to the sources behind your claim. An editorial update does not reverify every source or confirm that an internship is open.