pricing-discount-governance-system

Pricing Discipline & Discount Governance

A governed pricing analytics system that identifies where growth is price-led versus structurally discount-dependent. Built as decision-support for commercial leadership, finance, and RevOps.

Live dashboard: https://mfidalgomartins.github.io/pricing-discount-governance-system/

Why this exists

Discount-led growth can look healthy until it damages margin performance and commercial discipline. This project separates sustainable pricing from discount dependency, making exposure explicit across customer, segment, product, region, and channel views.

What it delivers

It builds a governed pricing layer from transactional data, quantifies realized discount and price realization, and prioritizes governance risk with interpretable scoring and recommended actions. Outputs are reproducible, auditable, and built for decisions, not just reporting.

Decisions it supports

• Which customers and segments need immediate pricing governance intervention.
• Which products and channels are structurally discount-dependent.
• Where margin-at-risk concentration is rising.
• Which governance actions should be prioritized first.

Architecture

Ingestion -> Processing -> Feature engineering -> Risk scoring -> Validation -> Analysis outputs -> Executive dashboard.

Repository map

• src/ pipeline logic
• scripts/ runnable entrypoints
• config/ thresholds and metric contracts
• docs/ methods and governance notes
• outputs/ analysis artifacts and dashboard source build
• tests/ regression checks

Core outputs

• GitHub Pages entrypoint: docs/index.html (published at repository root URL)
• Final dashboard file: outputs/dashboard/pricing-discipline-command-center.html
• Published dashboard artifact: docs/pricing-discipline-command-center.html
• Visualization pack: outputs/visualizations/*.png
• Validation evidence: outputs/*validation*
• SQL warehouse evidence: outputs/warehouse/*

Why this is stronger than a typical portfolio build

• Governance scoring is tied to operational actions, not vanity metrics.
• Metric contracts and validation checks reduce silent analytical drift.
• Release readiness is explicitly gated before publication.
• Business framing, outputs, and dashboard are aligned to pricing decisions.

Run

python3 -m venv .venv
. .venv/bin/activate
pip install -r requirements.txt
python scripts/run_pipeline.py
python scripts/publish_pages_dashboard.py

Limitations

• Data is synthetic; production use requires real commercial inputs.
• Margin is a proxy, not audited accounting gross margin.
• Scoring supports prioritization, not autonomous pricing decisions or committee-grade attribution.

Tools: Python, SQL, DuckDB, pandas, NumPy, Matplotlib, Seaborn, Chart.js, pytest.