Customer-Segmentation Study for a DTC Subscription Box
Overview
What this challenge is about.
Use 18 months of anonymized data: order history, churn events, NPS responses, box-rating data, referral activity, marketing-channel attribution. Engineer features (RFM-style + behavioral cohort signals). Run unsupervised clustering: k-means with elbow + silhouette, plus HDBSCAN as a comparison. Validate clusters: silhouette score, segment stability over time, qualitative inspection with 5 marketing-team members. Identify 4-6 actionable segments. For each: behavioral profile, retention/upsell action, expected reach + impact estimate. Deliver: notebook (Jupyter) with full analysis, 6-page segmentation playbook, segment-membership export for the marketing-ops team, and a 1-page CMO summary.
The Brief
What you'll do, and what you'll demonstrate.
Find 4-6 behaviorally coherent customer segments via unsupervised learning and produce a playbook with actions per segment.
Earning criteria — what you'll demonstrate
- Apply k-means and HDBSCAN to real behavioral data with proper feature engineering
- Validate cluster quality with statistical + qualitative checks
- Translate unsupervised-learning outputs into marketing-actionable segments
- Communicate ML findings to non-technical CMO audiences
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
Careers
Roles this prepares you for.
Real titles. Real skill bridges. Pick the one closest to your trajectory.
Product Manager
Product managers who can read segmentation studies design lifecycle features that resonate per-segment instead of one-size-fits-all.
This challenge sharpens
- business-analytics
- ml-applications
- model-evaluation