E-Commerce Data Analysis

About the Project

This is my first data analysis portfolio project. The goal was to take a real-world, messy dataset, clean it, and extract basic business insights about sales, products, and customers using Python and Pandas.

Dataset

• Source: E-Commerce Data on Kaggle
• Description: Contains ~540,000 transactions from a UK-based online retail store (Dec 2010 – Dec 2011).
• Key Columns: InvoiceNo, StockCode, Description, Quantity, InvoiceDate, UnitPrice, CustomerID, Country.

What I Did

1. Data Cleaning & Preprocessing

• Loaded the data and checked for missing values and duplicates.
• Dropped rows with missing CustomerID (as customer analysis is a key goal).
• Converted InvoiceDate to datetime format and extracted Month and DayOfWeek for easier analysis.
• Created a new TotalPrice column (Quantity * UnitPrice).
• Filtered out canceled orders (negative quantities) to focus on actual sales.

2. Exploratory Data Analysis (EDA)

• Time Analysis: Calculated total revenue and number of orders by month and day of the week.
• Product Analysis: Found the top 10 most popular products by quantity sold and by total revenue.
• Customer Analysis: Identified the top 10 customers by total spending and calculated the average order value.
• Geographic Analysis: Checked which countries generate the most revenue.

3. Data Visualization

Created clear and simple charts using Seaborn to visualize:

• Monthly revenue trends.
• Top selling products (bar chart).
• Sales distribution by country.

Key Findings

1. Best Month: The highest revenue was generated in November, likely due to holiday shopping.
2. Top Product: The most frequently purchased item was POPCORN HOLDER.
3. Customer Base: The majority of transactions come from the United Kingdom, which is expected for a UK-based store.
4. Average Order: The average customer spends approximately £534 per transaction.

Technologies Used

• Python
• Pandas
• Matplotlib & Seaborn
• Jupyter Notebook

How to Run This Project

1. Download e-com.ipynb and data(from repository or Kaggle) OR clone repository
2. Install required libraries(check requirements.txt)
3. open and run e-com.ipynb notebook