🚀 PROJECT: E-commerce Data Analysis

Scenario A data analyst at an online store.

The company says:

“Revenue has been unstable. We want to understand what's happening and how to improve it.”

---

🚀 Overview

This project analyzes e-commerce data to identify key drivers of revenue, conversion performance, and business growth opportunities.

The dataset initially contained messy and inconsistent data, which was cleaned and transformed into an analysis-ready format.

---

🎯 Objectives

• Understand revenue trends over time
• Identify key drivers of revenue changes
• Analyze conversion rate and user behavior
• Detect data quality issues
• Provide actionable business recommendations

---

🧹 Data Cleaning

The raw dataset contained several issues:

• Missing values (country, device, sessions, product)
• Inconsistent categories (e.g., "Germany" vs "germany")
• Invalid date formats
• Non-numeric revenue values

Cleaning Steps:

• Standardized text fields (lowercase)

• Replaced missing values with:

  • unknown (categorical fields)
  • median (sessions)
  • 0 (revenue/conversions where necessary)

• Removed invalid dates

• Converted all numeric fields properly

---

📊 Key Metrics

• Conversion Rate = conversions / sessions
• Average Order Value (AOV) = revenue / conversions

---

📈 Analysis

1. Revenue Trend

• Revenue analyzed over time to detect patterns and anomalies

2. Conversion Rate Analysis

• Compared across devices and countries
• Identified underperforming segments

3. Product Performance

• Identified top products by revenue
• Analyzed AOV for pricing insights

4. Data Quality Insights

• Measured impact of unknown categories
• Highlighted tracking/data issues

---

🔍 Key Insights

• Revenue fluctuations driven primarily by changes in conversion rate
• Mobile users show lower conversion compared to desktop
• Some products generate high revenue but low AOV
• A portion of revenue is linked to unknown categories, indicating tracking issues

---

💡 Business Recommendations

• Improve mobile user experience to increase conversions
• Focus marketing on high-performing segments
• Promote high AOV products
• Fix tracking issues for missing product and country data

---

🛠 Tools Used

• Excel (data cleaning, pivot tables, charts)
• Python (data generation and preprocessing)

---

📁 Files

• data/ → raw and cleaned datasets
• images/ → charts and visualizations
• analysis_summary.md → detailed insights

---

🧠 Key Learning

This project demonstrates how messy data can be transformed into actionable insights that drive business decisions.

---

📌 Author

Reshma Chougule