🏬 Exploratory Data Analysis — Super Store Retail

Perspective: Business Manager / Data Analyst Goal: Identify weak areas, profit leaks, and growth opportunities through data exploration and modeling.

This project transforms the classic Superstore dataset into an analytics-ready star schema and delivers a Power BI dashboard focused on actionable business insights.

---

🎯 Business Questions

• Which categories and products generate profit vs loss?
• How do discounts affect profitability?
• Which states and cities drive or destroy margin?
• Do shipment modes or customer segments actually matter?

The goal is not just visualization — but decision support.

---

🧱 Project Structure

This project is split into two clear phases:

1. Data Preprocessing & Modeling Using Power Query & M-language, the raw dataset is transformed into a star schema for performance and clarity.

2. Power BI Analytics Dashboard Interactive dashboards surface insights for sales, profit, geography, products, and discounts.

📂 Dataset: Superstore 📊 Dashboard file: SuperStore.pbix

---

⚙️ Part 1 — Data Preprocessing & Modeling

The raw dataset contained:

• ~10,000 records
• 13 wide columns
• no explicit date dimension

Key Transformations

• Split the flat table into fact & dimension tables

• Modeled measures around Sales, Profit, Quantity

• Created dimensions for:

  • Shipment Mode
  • Customer Segment
  • Geography (Region / State / City)
  • Product Category & Sub-Category

Final Star Schema

This model enables:

• faster queries
• cleaner measures
• scalable analytics

---

📊 Part 2 — Dashboard Analysis & Insights

📈 General Business Overview

• $2.3M total sales across 9,994 orders
• $288K total profit
• Average order value: $230
• Average products per order: 3.79

---

🗂️ Category Insights

• Office Supplies dominate volume:

  • 22,981 products sold (60.5% of quantity)

• Technology is the most profitable category:

  • $839K sales
  • $146K profit

• Furniture is the weakest category:

  • $720K sales
  • only $19K profit

🔻 Biggest loss makers:

• Tables → $207K sales, $17K loss
• Bookcases → $115K sales, $4K loss

---

📦 Product & Geography Insights

• Most profitable products:

  • Copiers → $56K profit
  • Phones → $45K profit

• Most losing products:

  • Tables
  • Bookcases

🌎 Geographic findings:

• Top states: California ($76K profit), New York ($71K)
• Worst states: Texas ($25K loss), Ohio ($18K loss)
• Top cities: NYC ($65K), Los Angeles ($34K)
• Worst cities: Philadelphia ($15K loss), Houston ($14K loss)

---

🚚 Shipment Mode Analysis

• Standard Class accounts for ~59% of orders
• First Class is the most profitable ($31K profit)

📌 Insight:

Shipment mode has minimal impact on sales or profit.

---

👥 Customer Segment Analysis

• Home Office and Corporate segments are most profitable (~$60K each)
• Consumer segment shows net loss (~$11K)

📌 Insight:

Customer segment has little influence on overall sales volume.

---

💸 Discount Impact (Critical Insight)

• Discounts above 30% consistently lose money

• Worst discount: 80% → $2K loss

• Best performing discounts:

  • 10% → $7K profit
  • 20% → $6K profit

Recommendations

• Limit discounts to ≤ 20%

• Avoid deep discounting (>30%)

• Explore alternatives:

  • free shipping
  • bundled gifts
  • loyalty incentives

---

📊 Dashboard Overview

Due to Power BI sharing limitations, screenshots are provided below.

📂 Interactive file: SuperStore.pbix

Home Page

Sales & Trends

Category & Product Analysis

Findings Summary

---

🧠 Key Takeaways

• Volume ≠ Profit
• Technology drives margin, Furniture erodes it
• Discounts are the largest controllable profit lever
• Geography matters more than customer segmentation

---

⭐ Why This Project Matters

This project demonstrates:

• analytical thinking beyond charts
• proper data modeling for BI
• translating insights into business actions
• real-world decision support

---

📌 Developed as part of The Spark Foundation Internship Program. This dashboard is a showcase case study, not a production system.