Growth Analytics · Spotify Application

Hi, I am
Darshan Senthil

Data Scientist with 4 years in growth analytics, experimentation, and predictive modeling. I want to help Spotify turn 751M users into more paying subscribers.

4 Years Experience SQL + BigQuery A/B Experimentation 3.96 MS GPA 50+ Enterprise Clients

See Feature Proposal View Experience Download Resume

01. Why Spotify

I want to work where data shapes
what 751M people hear.

The problem I want to solve

Spotify's free-to-paid conversion rate has shifted from 42% to 39% even as total MAUs hit 751M. Paid media is acquiring free users but not efficiently converting them to premium. The attribution question: which channels actually drive premium conversions, not just installs, is one of the most interesting unsolved problems in streaming growth analytics right now.

That is exactly the kind of problem the Growth Analytics team exists to solve. And the work I want to do.

751M+ Monthly active users as of Q4 2025

Why I am a strong match

A/B experiments across 15+ enterprise clients: 55% conversion lift, 22% ROI increase
Logistic regression forecasting model predicting campaign outcomes, improving prioritization by 20%
Subscriber churn risk analysis using clickstream and web analytics data
KPI frameworks for subscription products used by 50+ enterprise clients
MS Computer Science, Rutgers University, GPA 3.96/4.00

02. Technical Depth

Every requirement.
All of them real.

Spotify Needs	I Use This	Where
SQL (BigQuery), required	SQL advanced + BigQuery	Vue.ai, WPTI, Rutgers, Hirestar
Python / R, required	Python (Pandas, NumPy, scikit-learn) + R	Rutgers, Vue.ai modeling
A/B + Incrementality Testing	A/B test design and analysis	Vue.ai, 15+ enterprise clients
Attribution Modeling	Logistic regression, hypothesis testing	Vue.ai forecasting model
Scenario and Forecast Modeling	Regression forecasting, cohort analysis	Vue.ai, Rutgers
Statistical Modeling	Regression, churn and segment analysis	Vue.ai, Rutgers
Dashboards and Pipelines	Tableau, Looker, BigQuery, SQL pipelines	Vue.ai, WPTI
Performance Marketing Analytics	Campaign ROI, conversion lift analysis	Vue.ai, 22% ROI lift
Stakeholder Presentations	Executive and cross-functional reporting	Vue.ai, WPTI, Rutgers

03. The Work

Four roles.
Real outcomes.

Vue.ai

Data Analyst, Client and Product Analytics

February 2021 to August 2022 · Chennai, India

A/B experiments across 15+ enterprise clients: pricing, promotion, personalization
Subscriber churn risk analysis using clickstream and web analytics data
SQL metrics layer for Vue Mail: open rates, engagement, campaign conversion
Logistic regression model forecasting campaign outcomes, improving prioritization by 20%

55%Conversion lift through targeted experimentation

Workforce Professionals Training Institute

Data Infrastructure Engineer

July 2024 to Present · New York, United States

Tableau dashboards surfacing retention risk signals for program leadership
SQL data quality automation saving 120+ hours per month
Batch data pipelines centralizing workforce data for faster analysis

25%Improvement in participant retention

Rutgers School of Public Health

Data Scientist

October 2023 to May 2024 · New Brunswick, United States

Regression analysis identifying survey drop-off drivers
RAG classification workflow labeling 50K+ social media posts at scale
Snowflake dashboards reducing manual reporting effort by 40%

18%Survey completion rate improvement

Hirestar.io

Analytics Engineer

January 2020 to November 2020 · Hyderabad, India

NLP resume parsing pipeline: skills, experience, education extraction
End-to-end hiring funnel metrics identifying stage-level conversion drop-offs

15%Candidate conversion rate improvement

04. Feature Proposal

Predictive Spend Reallocation:
solving Spotify's conversion gap.

The Problem

Spotify's free-to-paid conversion rate has declined from 42% to 39% even as total MAUs grew to 751M. Paid media is driving free user acquisition but not efficiently converting them to premium. Current attribution likely measures cost per install, not cost per premium conversion. The result: marketing budget is allocated to channels that acquire free users, not paying subscribers.

The Model

Build a logistic regression and gradient boosting model that scores new free users on their 90-day premium conversion probability based on listening behavior, onboarding actions, content engagement depth, and acquisition channel.

The Attribution

Run incrementality tests across Meta, Google, and TikTok to measure true lift in premium conversions, not just installs, by channel, region, and user segment. Isolate which campaigns drive high-intent users who convert within 90 days.

The Outcome

Reallocate paid media budget toward channels with the highest predicted premium conversion rate, not lowest CPM. Same budget, more paying subscribers. Improve free-to-paid conversion efficiency across the entire paid media portfolio.

Current attribution vs proposed attribution: measuring what actually matters

Cost per install is lower but doesn't reflect premium conversion. The proposed approach measures cost per premium conversion to reallocate budget toward channels that drive paying subscribers.

This is not a generic feature idea. It is the exact problem the Growth Analytics team exists to solve: closing the gap between free user acquisition and premium conversion through better attribution, smarter experimentation, and predictive modeling. That is the work I want to do at Spotify.

05. Education and Certifications

The foundation.

Rutgers University

Master of Science in Computer Science

August 2022 to May 2024 · New Brunswick, NJ

3.96 / 4.00

Machine Learning · Business Intelligence and Visual Analytics · Natural Language Processing · Advanced Artificial Intelligence

Certifications

Databricks Certified Data Engineer Associate
AWS Cloud Practitioner
Google Data Analytics Certification
Tableau Desktop Certification

Hi, I amDarshan Senthil

I want to work where data shapeswhat 751M people hear.