Step 1 / 13

Project Overview

Steps
  1. 01Project Overview
  2. 02Dataset & Downloads
  3. 03Project Roadmap
  4. 04Step-by-Step Implementation
  5. 05Data Cleaning Phase Summary
  6. 06Exploratory Data Analysis
  7. 07Business Questions
  8. 08Tool-Specific Implementation (Python / Seaborn)
  9. 09Expected Output After Every Major Step Recap
  10. 10Visualization & Dashboard
  11. 11Insights & Recommendations
  12. 12Final Project Deliverables
  13. 13Final Project Checklist

Project Overview

Project Title: Northwind Commerce — Customer & Product Analytics (Seaborn)

Business Scenario: Northwind Commerce is a multi-category retailer (Furniture, Office Supplies, Technology) selling to Consumer, Corporate, and Home Office customers across the US. Leadership has sales numbers, but no statistical view of which customers are actually valuable and which products are quietly losing money.

Business Problem: The Head of Customer Insights wants to know: are our biggest-spending customers actually our most profitable ones? Are certain product sub-categories dragging down margins regardless of category? Is there a real statistical relationship between discounting and profit, or does it just look that way anecdotally?

Project Objective: Use Python and Seaborn — a statistical visualization library built on Matplotlib — to explore distributions, relationships, and categorical comparisons across customers and products, and turn that into a customer & product analytics report.

Key Stakeholders: Head of Customer Insights, Product/Category Managers, Marketing (customer segmentation), Finance (margin risk by product).

Final Business Goal: Identify which customers are high-value vs. high-risk, which product sub-categories should be reviewed for pricing or discontinuation, and quantify how strongly discounting erodes profit — backed by statistical charts, not gut feeling.

What You Will Finally Build

  • A cleaned transactional dataset with a derived customer-level summary table

  • A Python (Jupyter Notebook) analysis pipeline using pandas + Seaborn

  • 8 statistical charts: distribution plot, violin plot, bar plot with confidence intervals, correlation heatmap, regression plot, pairplot, faceted categorical plot, and a top-customers plot

  • A written customer & product insights report

  • A short final presentation summary

Why Seaborn instead of plain Matplotlib? Seaborn is built for exactly this kind of question — it natively understands DataFrames, computes confidence intervals, draws regression lines, and facets charts by category with a fraction of the code Matplotlib alone would need. This project is intentionally scoped as Python/Seaborn-only (no SQL, Excel, or BI tool) to build deep statistical-visualization skill.