This project analyses a real-world retail transaction dataset using MySQL to uncover customer behaviour, product performance, sales trends, and business insights.

The goal was not just querying data — but answering real business questions retailers care about:

• When do customers shop?
• Who are the most valuable customers?
• Which products drive revenue vs volume?
• How much revenue is lost due to returns?
• Can we identify VIP customers?

The dataset contains transactional retail data with the following attributes:

Before analysis, the dataset required cleaning:

• Split InvoiceDate → Date & Time
• Converted European date format → MySQL format
• Handled NULL & blank values during import
• Fixed datatype issues (InvoiceNo stored as VARCHAR due to cancellations)
• Managed negative quantities (returns)

• Sales peak during specific hours of the day
• Weekdays generate more revenue than weekends
• Evening shopping contributes high order counts

👉 Helps staffing & promotion scheduling

• Identified repeat purchase cycle using purchase gap analysis
• Segmented customers into High / Medium / Low value
• Found VIP customers (Top 5%) using percentile ranking
• Detected one-time buyers and churned customers

👉 Helps retention & loyalty programs

• Top products by revenue differ from top products by quantity
• Some items sell frequently but generate low revenue
• Some premium items sell rarely but drive profit

👉 Helps inventory & pricing strategy

• Returns detected using negative quantity & cancellation invoices
• Certain products have disproportionately high return rates
• Calculated revenue lost due to returns

👉 Helps quality control decisions

• Revenue contribution varies significantly by country
• Different regions prefer different product types

👉 Helps market expansion planning

• RFM Analysis (Recency, Frequency, Monetary)
• Customer Segmentation
• Repeat Purchase Cycle Detection
• Cohort Identification
• Pareto (80/20) Customer Rule
• Order Pattern Analysis
• Time-based Sales Analysis

• MySQL 8 → Data analysis & querying
• Pandas → Pre-processing & formatting
• SQL Window Functions → Advanced analytics

This project simulates how a retail company would use transaction data to:

• Optimise inventory
• Improve marketing targeting
• Identify high-value customers
• Reduce product returns
• Predict repeat purchases

IKEA-Retail-SQL-Analysis/
│── ikea.zip/
│── │── ikea/
│── │── ikea_updated/
│── Cleaning.ipynb
│── ikea.sql
│── README.md

• Handling messy real-world datasets
• Writing analytical SQL beyond basic SELECT queries
• Using window functions for behavioural analysis
• Translating data into business decisions

Abhishek Singh Aspiring Data Analyst | SQL • Python • Power BI