🛒 E-commerce Sales Data Pipeline & Analysis

📌 Project Overview

This project demonstrates a professional end-to-end data engineering and analysis pipeline. Using a dataset of 10,000 retail records, it automates the process of extracting raw data, cleaning inconsistencies, and generating business-critical insights.

📊 Visual examples of the analysis results

Click the images to explore the full interactive analysis in the Jupyter Notebook.

🛠️ Tech Stack

• Language: Python 3.11
• Primary Library: Pandas (Advanced Data Wrangling)
• Numerical Logic: NumPy (Vectorized Operations)
• Visualization: Seaborn & Matplotlib
• Environment: Jupyter Notebooks / VS Code

⚡ Key Engineering Features

• Median Imputation: Automated handling of missing unit_price values by grouping by product type to ensure statistical accuracy.
• Categorical Normalization: Cleaned inconsistent naming conventions using vectorized string operations.
• Feature Engineering: Creation of performance KPIs like total_revenue and time-based metrics.
• Outlier Detection: Statistical analysis to identify anomalies in sales quantities and prices.

📊 Business Insights

• Sales Seasonality: Revenue peaks during the second quarter (April and May), with earnings exceeding €1.6M. Conversely, February is identified as the month with the lowest commercial activity, suggesting an opportunity for pre-emptive marketing campaigns.
• Weekly Sales Trends: There is a stronger purchasing trend during weekends, with Saturday having the highest average revenue per transaction. Mondays show the lowest sales volume of the week.
• Product Performance Ranking: Keyboards and Smartphones lead revenue generation, totaling over €7.6M combined. Keyboards slightly outperform smartphones in total sales volume.
• Revenue Distribution by City: Seville stands as the most profitable market with €3.87M in total sales, outperforming major hubs like Madrid and Barcelona.
• Category Dominance: The Electronics sector is the primary business driver, accounting for over 50% of total revenue (€9.33M), followed by Accessories and Office products.
• Customer Segmentation: There is a clear predominance of Gmail users, who account for more than 50% of total orders, which simplifies targeted segmentation for future email marketing campaigns.

📈 Key Results (2025 Analysis)

After processing 10,000 records, the pipeline yielded the following insights:

• Total Revenue: $18,339,362.04
• Average Order Value (AOV): $1,833.94
• Top Segment: The type of customer 'User' segment represents the highest revenue share.
• Regional Insight: Identified a significant market gap in the 'Office' category within the Seville region.

🚀 How to Run

1. Clone the repository.
2. Install dependencies: pip install -r requirements.txt.
3. Open notebooks/E-commerce_Data_Analysis.ipynb and run all cells.