PERSONAL PROJECT · NEXTLEAP AI PM FELLOWSHIP
Turning Spotify from a streaming app into a skill-building habit
A zero-engineering MVP concept for Goal-Based Audio Journeys: structured audio tracks that let 18 to 30 year olds learn a skill through the music and podcasts they already stream, with progress and streaks built in.
What success looks like · Projected Pilot Targets
Activation Target
60%
set a goal in week one
Core-Loop Target
40%
finish a full track
Retention Target
25-30%
come back on day 7
Perceived Learning Target
40%
say "I'm learning"
CONCEPT
MVP
Designed and prototyped in a 3-week sprint with no engineering. Targets are projections for a 1,000-user pilot, not shipped results.
ROLE
Product Manager (self-directed)
TIMELINE
~ 3 Weeks
PROJECT TYPE
0-to-1 concept / AI-assisted MVP
TOOLS
Lovable, Airtable, Zapier, GPT
CONTEXT
NextLeap AI PM Fellowship case
MY CONTRIBUTION
Problem framing, scope, architecture, prototype, metrics.
The brief I was handed
Spotify India's leadership wanted to test whether audio could become a personal-growth tool, not just entertainment.
The task: ship an MVP of a skill-building feature in 3 weeks, with no India dev bandwidth, a team of PM, designer, and analyst, using AI builders and no-code tools, piloting to 1,000 students and early-career users.
"If Spotify can help people learn with audio, not just listen, we'll have a unique edge."
Wide reach, shallow engagement
Every day, over 100 million Indians open Spotify and stream for hours. But the experience is passive. Nothing turns that daily listening into something they feel is progress, and nothing makes leaving cost them anything.
On price and catalog alone, Spotify is one tap away from JioSaavn, Gaana, or Wynk. Retention looks healthy. It is not defensible.
Who we're solving for?
18 to 30 year olds who already use audio to learn, streaming English music for vocabulary, then leaving for YouTube and Duolingo the moment they want structure.
People already learn through audio.
Nobody gives it structure.
"Music motivates me, but I want it tied to goals."
The problem
No structure on what people already stream
No sense of progress, sessions blur together
Switching cost to a rival is near zero
Learning intent leaks to YouTube and Duolingo
Retention is high but undifferentiated
Who are we solving for and where the loop closes
● TODAY - Listening Goes Nowhere
● WITH JOURNEYS - The loop closes
"The whole design goal is to close the loop before the user leaves for YT/Duolingo."
Worth doing, Worth doing now
Why now?
Behavior proven — Audio-first learning has mainstream adoption.
Platform ready — Spotify already owns daily audio consumption.
Window closing — Learning intent is moving to competing platforms.
Why it wins?
Users
Build consistent 10–15 minute learning habits.
Learn without leaving an app they already use daily.
Feel tangible progress through streaks and structured goals.
Business
Differentiate beyond music catalog and pricing.
Increase retention through accumulated progress.
Build a proprietary intent graph for future personalization, advertising and premium learning experiences.
"High retention isn't the same as defensible retention. This makes leaving feel like abandoning progress."
What went in, and what I deliberately cut
With 3 weeks and no engineering, I scored every feature on value versus effort and applied one rule: ship only what tests the core hypothesis that a structured audio track feels like skill-building and that users perceive progress.
● Ships
Structured daily track — 4 songs + 1 podcast clip + a daily focus line.
Goal selection — 3 goals, English vocabulary as hero.
Progress tracker — streak + completion %.
Light nudges
● Deffered
Quizzes — proxy learning with a survey first.
Certificates / badges — retention amplifier, premature.
AI-personalized paths — 3 goals + 3 weeks, curation beats a model.
Creator marketplace — separate monetization product.
The daily loop
Four screens carry the loop: goal picker, daily track, mark complete, progress.
Open the live prototype →
Measuring learning, not listening
The trap: minutes streamed is the entertainment metric this feature is trying to transcend. It cannot be the success measure. The metric tree below is deliberately narrow.
Projected pilot targets targets
NORTH STARt
Weekly structured-track completions per active user.
ACTIVATION
% who set a goal — target 60%.
CORE LOOP
% completing a full track — target 40%.
HABBIT
% reaching a 7-day streak.
RETENTIONt
D7 return to the journey, not just app open — target 25–30%.
PERCEIVED LEARNING
In-app survey "do you feel you're learning?" plus willingness-to-pay — target 40% yes.
COUNTER METRIC
Does core listening time drop, or do users get confused?
"The switching-cost edge holds only if the habit and perceived-learning numbers hold."
What I'd build next
Quizzes to validate learning beyond self-report.
Real personalization, once there's completion data to train on.
Badges to reinforce streaks and social proof.
Then the creator marketplace, the supply-side bet.
SNIGDHA NAGPAL
Personal Project · 2026