This repository provides an E-commerce Customer Behavior Dataset (v2) designed for data analysis, machine learning, and educational purposes. The dataset captures detailed customer demographics, purchasing behavior, order characteristics, device usage, and satisfaction metrics.
Version v2 extends the earlier release by supporting multi-order customers, making it suitable for more realistic behavioral analysis such as customer lifetime value, repeat purchasing patterns, and cohort analysis.
Key Statistics (v2):
- Total records: 17,049
- Total columns: 18
- Date range: 2023-01-01 to 2024-03-25
- Average Total Amount: ~1,277.44
- Median Total Amount: ~1,250 (approx.)
- Average Customer Rating: ~3.9 / 5
- Returning Customer Rate: ~88.21%
- Mobile Transactions: ~55.97%
- Missing values: None detected
Note: Statistics are computed directly from the provided CSV file and may differ from earlier versions of this dataset.
Dataset Versions:
- Single-order per customer
- ~5,000 records
- Intended mainly for basic exploratory data analysis (EDA)
- Multi-order per customer (each customer may place multiple orders)
- 17,049 records
- More realistic transactional behavior
- Suitable for advanced analytics and machine learning tasks
Order identifiers may appear multiple times (e.g., ORD_1023-1, ORD_1023-2) to represent multiple purchases by the same customer.
Column Description:
| Column Name | Description |
|---|---|
| Order_ID | Unique identifier for each order |
| Customer_ID | Unique identifier for each customer |
| Date | Date of the transaction |
| Age | Customer age |
| Gender | Customer gender |
| City | Customer city |
| Product_Category | Category of the purchased product |
| Quantity | Number of items purchased |
| Unit_Price | Price per unit |
| Discount_Amount | Discount applied to the order |
| Total_Amount | Final total amount paid for the order |
| Payment_Method | Payment method used |
| Device_Type | Device used for purchase (Mobile, Desktop, etc.) |
| Is_Returning_Customer | Indicates whether the customer is a returning customer |
| Customer_Rating | Rating given by the customer (1–5) |
| Shipping_Method | Shipping option selected |
| Delivery_Time | Delivery time in days |
| Order_Status | Final order status |
Intended Use Cases:
- Exploratory Data Analysis (EDA)
- Customer segmentation and profiling
- Predicting repeat purchases
- Recommendation systems
- Customer lifetime value (CLV) estimation
- Educational projects and demonstrations
Data Quality & Assumptions:
- The dataset contains no missing values.
- Values are internally consistent and cleaned.
- Monetary values represent the final transaction amount, after discounts.
- Customer and order identifiers are anonymized.
- The dataset is either synthetic or anonymized and does not contain personally identifiable information (PII).
Repository Structure (Recommended):
├── README.md
├── LICENSE
├── requirements.txt
├── data/
│ ├── ecommerce_customer_behavior_dataset_v2.csv
├── notebooks/
│ ├── ecommerce_customer_eda.ipynb
│ └── ecommerce_customer_eda.html
---
Installation:
```bash
git clone https://github.com/your-username/your-repository.git
cd your-repository
pip install -r requirements.txt
Data Source:
The dataset used in this project was obtained from Kaggle:
E-Commerce Customer Behavior and Sales Analysis (TR) Author: Umut Tuygur Source: https://www.kaggle.com/datasets/umuttuygurr/e-commerce-customer-behavior-and-sales-analysis-tr
The dataset has been reused for educational and analytical purposes in accordance with Kaggle’s dataset usage policies.
License:
- Dataset: CC0 1.0 Universal (Public Domain)
- Source code: MIT License (recommended)
You are free to use, modify, and distribute this dataset and code without restriction.
Citation:
If you use this dataset in academic or professional work, please cite it as:
E-commerce Customer Behavior Dataset (v2), 2024.
Contribution:
Contributions are welcome. Please submit pull requests or open issues for suggestions, improvements, or bug reports.
Disclaimer:
This dataset is provided for research and educational purposes only and does not represent real customer transactions.