Exploratory-data-analysis-with-SQL

This shows the use of SQL to perform EDA on a dataset for a retail company. It starts off from exploring the entire database to the products offered by the company and eventually answering important business questions.

Project Overview

This project explores sales performance and product trends for a cycling retail store using SQL. The goal of the analysis was to uncover key insights that support data-driven decision-making around product demand, inventory planning, and revenue growth opportunities. The project focuses on identifying what drives sales, which products perform best, and potential areas for business optimization.

Objectives

This SQL-based EDA aimed to answer the following:

• Best and worst performing products by sales?
• Distribution of sold items across countries and gender
• Revenue generated by each category
• What opportunities exist to improve revenue and stock movement?

Dataset Summary

• Industry: Cycling Retail
• Data Type: Customers, orders, and product information (3 tables)
• Contents: Bikes, frames, accessories, cycling gear, and spare parts
• Size: 60,000+ rows across 3 tables covering approximately 4 years

Tools and Skills Used

• Language: SQL
• Concepts Applied: Joins, aggregations, CTEs

Key Findings

• Bikes are the top-performing product category, generating a significant share of revenue (28,316,272), indicating concentrated customer demand in this category.
• The United States generated the highest revenue and has the largest customer base, making it the strongest market but also showing high reliance on a single region.
• Europe (United Kingdom, France, Germany) is underperforming relative to market potential for cycling products.
• Australia is a strong emerging market with high revenue and customer count relative to population size.
• Customer purchases are almost evenly distributed by gender (Male: 50.5%, Female: 49.5%), indicating balanced market appeal.

Revenue by Category

Customer distribution by countries

Customer distribution by gender

Number of items sold by country

Lowest selling products

Best performing products

Business Recommendations

• The lowest-selling products are low-value accessories, but they are frequently needed items. Bundle them with bike purchases (for example as: "Maintenance Kit Add-On") to encourage cross-selling and increase order value.
• Develop a tailored go-to-market strategy for Europe and Australia (pricing, localized marketing, and product customization) to improve performance in these markets.
• Prioritize inventory for best-selling products such as the Mountain-200 series to prevent stockouts and maximize sales.
• 871 records in the revenue data and 337 in the customer table are tagged as "n/a" for country. This indicates incomplete data collection, which limits customer segmentation and targeted marketing. Improving data capture at the point of sale would provide stronger customer insights and support personalization.
• Gender distribution is nearly equal, showing that the brand appeals to both males and females. Marketing campaigns can remain gender-neutral without losing effectiveness.

What This Project Demonstrates

This project highlights my ability to:

• Apply SQL to extract, clean, and analyze data for business insights
• Translate raw data into clear and actionable commercial recommendations
• Approach analytics from a commercial and customer-centric perspective

Next Steps

Future improvements:

• Implement better handling of null values at the point of data collection, for example by using data validation rules so fields such as gender and country must be selected before a record is saved.
• Add a dashboard layer (Power BI or Excel) to visualize key insights and trends.
• Incorporate customer segmentation to understand shopping patterns across gender, customer lifetime value, customer feedback on products, and return rates, as some best-selling products may also have high return rates.