Case Study : For a Supermarket Retail Chain Stores

---

Problem Statement

A supermarket chain faced critical challenges that impeded their ability to make timely and informed business decisions:

1. Slowness of Reports:
   • Reports took hours to generate, delaying operational and strategic insights.
   • High query complexity on large datasets led to performance bottlenecks.
2. Fragmented Data Sources:
   • Data was spread across multiple systems: SQL Server, Teradata, Salesforce, and CSV files.
   • Manual consolidation resulted in inconsistencies and errors.
3. Outdated Reporting Structure:
   • Reports were poorly designed, lacking actionable insights and a unified structure.
   • Key performance indicators (KPIs) were scattered across multiple, unrelated reports.

Proposed Solution

To address these challenges, the supermarket chain implemented a robust ETL pipeline using SQL Server Integration Services (SSIS), consolidated data into a SQL Server data warehouse, and restructured reporting with Power BI.

Scenario Details

Data Sources

1. SQL Server:
   • Stores transactional sales data from POS systems.
   • Example Data: Transaction ID, store ID, product ID, quantity sold, sales amount.
   • Daily Volume: 5 million rows.
2. Teradata:
   • Contains historical and aggregated sales data for trend analysis.
   • Example Data: Monthly sales summaries, customer loyalty data, and product trends.
   • Daily Volume: More than 3 million rows.
3. Salesforce:
   • Maintains customer data, including loyalty profiles and purchase history.
   • Example Data: Customer ID, loyalty tier, lifetime value, and recent interactions.
   • Daily Volume: 1 million rows.
4. CSV Files:
   • Regional campaign data and inventory restocking schedules provided by regional managers.
   • Example Data: Campaign ID, region, forecasted sales, and restocking dates.
   • Daily Volume: 1 million rows.

Solution Architecture

1. ETL Tool: SQL Server Integration Services (SSIS)
   • Data Extraction:
     • SSIS connects to SQL Server, Teradata, Salesforce, and ingests CSV files via FTP or direct upload.
     • Incremental loading ensures only new and updated records are processed.
   • Data Transformation:
     • Data Cleansing:
       • Remove duplicates, resolve null values, and validate data types.
     • Standardization:
       • Harmonize date formats, product codes, and region identifiers across all sources.
     • Enrichment:
       • Join Salesforce loyalty data with POS sales for enhanced customer insights.
       • Integrate campaign data with historical sales from Teradata.
     • Aggregation:
       • Calculate metrics like revenue, profit margins, customer lifetime value, and sales per region.
   • Data Loading:
     • Transformed data is loaded into SQL Server as the centralized data warehouse.
     • Partitioned tables by date, region, and product category optimize query performance.
     
   • Target Data Warehouse: SQL Server
     • Schema Design:
       • A star schema with fact tables for sales, inventory, and campaign performance.
       • Dimension tables for products, customers, regions, and time.
     • Performance Enhancements:
       • Clustered indexing on transaction dates and product IDs.
       • Compression and partitioning minimize storage overhead and improve query speed.
   • Power BI for Reporting
     • Restructured Dashboards:
       • Unified dashboards for sales, inventory, and marketing campaigns.
       • KPIs such as daily revenue, stock levels, and campaign ROI are visually highlighted.
     • Optimizations:
       • Power BI leverages SQL Server's aggregated data for faster query execution.
       • DirectQuery mode ensures real-time insights for frequently accessed data.
       • Scheduled refresh every 6 hours ensures dashboards remain current.

Reporting Capabilities

1. Sales Performance Dashboard:
   • Track daily and monthly revenue trends by store, region, and product category.
   • Identify top-selling products and customer segments driving sales growth.
2. Customer Insights Dashboard:
   • Analyze customer loyalty tiers, purchase patterns, and lifetime value.
   • Segment customers by region and preferences to tailor promotions.
3. Inventory and Campaign Dashboard:
   • Monitor stock availability and flag potential stockouts or overstocks.
   • Evaluate the ROI and sales uplift from marketing campaigns.

Scalability and Performance Enhancements

1. Data Volume Management:
   • SQL Server handles 10 million daily rows with partitioning and indexing.
   • SSIS pipelines are designed for parallel execution, reducing ETL runtime.
2. Improved Reporting Speed:
   • Pre-aggregated data in SQL Server minimizes query complexity in Power BI.
   • Reports load within 3-5 seconds for 90% of queries, compared to hours before.
3. Resilience:
   • SSIS error-handling mechanisms include retries and logging for failed ETL jobs.
   • SQL Server backups ensure data integrity and disaster recovery capabilities.

KPIs Monitored Post-Implementation

1. Sales Performance: Achieved 15% monthly revenue growth.
2. Customer Engagement: Improved customer retention by 8% through loyalty analysis.
3. Inventory Efficiency: Reduced stockouts to less than 1.5% of SKUs.
4. Campaign ROI: Increased ROI to 5x by targeting high-value customer segments.
5. Report Performance: Reduced report loading time to under 5 seconds for 90% of queries.

Outcomes and Benefits

1. Streamlined ETL Process: SSIS efficiently handles 10 million daily rows across SQL Server, Teradata, Salesforce, and CSV files.
2. Centralized Data Warehouse: SQL Server consolidates data from disparate sources, ensuring consistency and accessibility.
3. Optimized Reports: Power BI provides faster, actionable, and visually appealing insights.
4. Enhanced Decision-Making: Unified reporting empowers teams to identify trends, optimize inventory, and improve customer engagement.

This solution resolved the supermarket chain's data integration and reporting issues, delivering a scalable and high-performance analytics platform.