https://github.com/Arun-Kumar-Venugopal/Retail-Brand-Performance-PowerBI-AzureSQL
This project demonstrates an end-to-end Power BI solution using Microsoft Azure SQL Database as the cloud data source.
It covers:
- Azure SQL Database creation and configuration
- Data loading and SQL-based data cleaning
- Secure Power BI connectivity (SQL Authentication & Microsoft Account)
- Data transformation using Power Query
- Business calculations using DAX
- Interactive dashboard design
- Publishing and sharing reports using Power BI Apps
This project helps retail and merchandising teams evaluate brand performance. It analyzes discount strategies, profitability, pricing, and brand variety to optimize product mix, pricing decisions, and overall retail performance.
The retail business needs to evaluate brand-level performance to understand pricing strategies, discount effectiveness, and profitability.
- Average Discount %
- Average Profit %
- Sales Price vs Marked Price
- Top 5 Brands by Profit %
- Bottom 5 Brands by Profit %
- Brand Variety Count
- Profit % is simulated for analytical purposes
- Cost price is derived using calculated logic, not actual procurement data
- Analysis is limited to available brand and pricing records
Microsoft Azure SQL Database
→ SQL Server Management Studio (Data Prep & Validation)
→ Power BI Desktop (Transformations & DAX)
→ Power BI Service
→ Power BI App
- Platform: Microsoft Azure SQL Database
- Dataset: Men’s T-Shirt Retail Data
- Initial Columns:
- Brand
- Title
- Original_Price
- Sale_Price
- Created a free Microsoft Azure account
- Created Azure SQL Database and SQL Server
- Configured SQL Authentication and firewall rules
- Connected via SQL Server Management Studio (SSMS)
- Imported flat file data (
Men Tshirt) into Azure SQL - Updated column data types for price fields
- Removed invalid characters (
?) from price columns - Cleaned
Original_PriceandSale_Priceusing SQL queries - Prepared data for numeric conversion
- Connected Power BI Desktop using:
- SQL Server Authentication
- Microsoft Account authentication (via Microsoft Entra ID)
- Verified schema and refresh functionality
- Enabled column profiling for entire dataset
- Identified and handled missing (
NA) values - Filtered invalid rows
- Estimated missing original prices using conditional logic
Final cleaned columns:
- Brand
- Title
- Sales Price
- Marked Price
- Discount %
- Profit %
- Cost Price
Final data model:
- Brand
- Title
- Sales Price
- Marked Price
- Discount %
- Profit %
- Cost Price
- Landing page for the report
- Displays available brands using Multi-row Card
- Focused on branding and navigation
Contains all analytical visuals:
- Top 5 Brands by Average Discount % (Bar Chart)
- Top 5 Brands by Variety Count (Donut Chart)
- Top 5 Brands by Average Sales Price (Ribbon Chart)
- Top 5 Brands by Average Profit % (Area Chart)
- Bottom 5 Brands by Average Profit % (Circle Chart)
- Scrolling company banner using App Source visual
- Published report to Power BI Service
- Created a dedicated workspace
- Shared report via Power BI App
- Granted access to organization users
- Creating and managing Azure SQL Databases
- Performing SQL-based data cleaning
- Secure Power BI authentication methods
- Advanced Power Query transformations
- DAX calculations for pricing and profitability
- Designing executive-ready dashboards
- Sharing reports using Power BI Apps
- Replace simulated profit with actual cost data from ERP systems
- Add time-based analysis if transaction dates become available
- Implement row-level security for brand managers
- Enable incremental refresh for large retail datasets
🔗 View-only Power BI App (Microsoft login required)
Arun Kumar Venugopal
Power BI | SQL | Azure | Data Analytics