Retail Supplier Performance Analysis

A data analysis project to optimize supplier performance in retail, inspired by Musgrave. Built with Python, SQL, and Power BI.

Project Overview

This project analyzes supplier performance and engagement for a retail scenario (simulating a supermarket like Musgrave), using the Online Retail dataset. The purpose is to evaluate key metrics such as delivery reliability, profitability, and return rates, providing actionable insights for business optimization.

Data Source

• Dataset: Online Retail.xlsx (541,909 rows) with columns: InvoiceNo, StockCode, Description, Quantity, InvoiceDate, UnitPrice, CustomerID, and Country.
• Location: Stored in the notebooks/ folder for initial exploration.
• Source: Downloaded from the UCI Machine Learning Repository.

Data Exploration and Organization

• Initial Assessment:
  • Total rows: 541,909.
  • Identified issues: 1,454 missing Description values (0.27%), 135,080 missing CustomerID values (24.93%), 10,624 rows with negative Quantity (1.96%) indicating returns, 2 rows with negative UnitPrice (0.0004%) labeled "adjust bad debt", and 2,517 rows with UnitPrice = 0 (0.46%) often with non-zero Quantity.
• Cleaning Decisions:
  • Imputed missing Description with "Unknown" to retain rows for item-level analysis.
  • Imputed missing CustomerID with "Unknown" where Quantity and UnitPrice were available, preserving financial and volume metrics.
  • Added IsReturn flag for negative Quantity rows to analyze returns separately.
  • Added IsBadDebt flag for negative UnitPrice rows for financial impact analysis.
  • Added IsZeroPrice flag for UnitPrice = 0 with non-zero Quantity for volume analysis.
• Outcome: The dataset is now organized in notebooks/01_Data_Exploration_and_Cleaning.ipynb, ready for further analysis with cleaned and flagged data.

Usage

1. Clone the repository: git clone https://github.com/seu-usuario/Retail-Supplier-Performance-Analysis.git
2. Navigate to notebooks/ and open 01_Data_Exploration_and_Cleaning.ipynb with Jupyter Notebook or VS Code.
3. Ensure Online Retail.xlsx is in the notebooks/ folder.

Contact

For questions, contact [https://www.linkedin.com/in/victor-garlet/].

Last Updated: August 20, 2025