DigiPlus Subscription Value & P&L Case Study

This project turns subscription activity, plan revenue, and direct cost data into an operating view of member value.

It is designed around four business questions:

• Which member segments generate the strongest order and margin contribution?
• How do membership fees and direct benefit costs reconcile at portfolio level?
• What share of eligible subscriptions renew within the agreed grace period?
• How do active-member economics change month by month?

What the pipeline produces

Output	Decision supported
portfolio_pnl.csv	Reconciles fee revenue, commerce margin, shipping contribution, and direct costs
segment_performance.csv	Profiles active, lapsed, and never-member customers by RFM segment
renewal_summary.csv	Tracks eligible plans, renewed plans, renewal rate, and renewal timing
monthly_member_trend.csv	Shows active members, orders, NMV, and contribution over time
customer_period.csv	Provides the customer-level audit table behind the aggregates
summary.md	Collects the main period results in a short business readout

The P&L is built from observed member activity and directly recorded costs. It is a portfolio performance view, not an estimate of incremental impact.

Run the case study

python -m pip install -e ".[dev]"
digiplus-case demo

The demo creates a reproducible synthetic dataset, runs the full analysis, and writes the outputs to artifacts/.

To run the pipeline on another dataset:

digiplus-case analyze \
  --data-dir path/to/input \
  --output-dir path/to/output \
  --start 2025-07-01 \
  --end 2025-10-01 \
  --as-of 2026-01-01

Project structure

src/digiplus_case/     analysis pipeline and synthetic data generator
tests/                 metric, reconciliation, and end-to-end tests
docs/                  case-study design and metric definitions
sql/                   warehouse extraction contract

Validation

pytest -q
ruff check .

The public version contains no company data. See Data Provenance for the synthetic-data design and Case Study Design for the analytical decisions.