The Global Superstore Analysis

"Retail is no longer local, it’s a global battlefield."

This project transforms the Global Superstore dataset into clear, data-driven business insights, to uncover patterns in sales, profit, discounts, and shipping efficiency. It combines data cleaning and ER modeling, descriptive analytics, statistical testing, and interactive dashboarding to provide actionable business insights. , the goal is to answer real-world retail questions and highlight where growth, risks, and opportunities lie.

---

Preview

Business Questions

Every project starts with the right questions. Here we defined what matters most: sales, profit, customers, and markets.

Data Cleaning Process

Raw data isn’t ready for insight. We cleaned and prepared the Global Superstore dataset so analysis could be accurate and trustworthy.

Entity Relationship Models

Before analysis, we designed the structure. The ERM shows relationships at a high level, and the ERD details how tables connect.

ERM

ERD

Power BI Dashboards

We translated the analysis into dashboards that tell the story visually.

1. Market Insights – Which regions and markets drive the most growth?

2. Efficiency & Operations – How well do shipping modes and delivery times perform?

3. Product Performance – Which categories and subcategories are most profitable?

Hypothesis Testing

To go beyond dashboards, we tested assumptions with statistical methods, adding depth to the business insights.

Tools & Technologies

• Power BI – KPI dashboards & visual analytics
• Python – Data cleaning, EDA, hypothesis testing
• SQLite – Database creation & SQL queries
• Excel – Initial exploration & data checks

Key Dashboards

1. Market Performance: Top/Bottom countries, regional profitability
2. Efficiency & Trends: Shipping, discounts, seasonality
3. Product & Customer Insights: Categories, sub-categories, segments

Hypothesis Testing

• Test: Profit margin difference between orders with and without discounts
• Result: Large negative impact of discounts on profitability
• Effect Size: Hedges’ g = -0.531 (large), Cliff’s delta = -0.590 (large negative)

Key Insights

• Large regional variability in profitability
• Discounts significantly reduce profit margins
• High-profit categories: Technology & Office Supplies; Loss-making: Tables
• Standard shipping dominates, but delivery times vary widely
• Corporate segment generates highest profits despite fewer orders

Recommendations

• Reassess pricing & discount strategy
• Optimize product mix in loss-making categories
• Improve logistics efficiency in slower shipping modes
• Focus marketing on high-margin regions & segments
• Phase out or reprice unprofitable sub-categories

Folder Structure

/GlobalSuperstore-Analysis │── data/ # Dataset files │── notebooks/ # Python EDA & analysis │── dashboards/ # Power BI files │── images/ # Screenshots & charts │── README.md # Project documentation

Project Files

• Notebook → Open analysis.ipynb
• Power BI Dashboard → Download Final_Project_Dashboard.pbix
• Presentation → Download PowerPoint (.pptx)
  (https://docs.google.com/presentation/d/1DsdYUjyjQjNTQVWSHFW5C4xvO2wGCCL1/edit?slide=id.p1#slide=id.p1 )

Author

Egbe – Data Analyst
GitHub: [https://github.com/Egbe34]

Trello (https://trello.com/b/Eh8xoOw9/final-project-global-superstore-analytics) Presentation (https://docs.google.com/presentation/d/1DsdYUjyjQjNTQVWSHFW5C4xvO2wGCCL1/edit?usp=sharing&ouid=101322982193861291493&rtpof=true&sd=true )