Independent Project
Modelled 62,884 Transactions Into an Executive Sales Dashboard for a Global Electronics Retailer
0
Transactions Modelled
Executive Summary
As a self-directed project outside client work, I built an end-to-end Power BI solution for a global electronics retailer's transaction dataset, spanning 62,884 sales records across multiple product lines, stores and customer segments. I designed a star schema in Power Pivot to hold the model together, then built an executive-facing dashboard with KPI cards and slicers for self-service exploration. The project was a deliberate exercise in doing the data modelling work properly before touching a single chart, the same discipline I now bring to commercial engagements.
Context
This was a self-directed analytics project built on a public global electronics retailer training dataset, used to practise end-to-end business intelligence delivery on a large, realistic transactional dataset outside client work. It is not a client engagement.
Challenge
The raw dataset spanned 62,884 transactions across multiple product lines, store locations and customer segments, with no existing data model. Turning it into a dashboard an executive team could actually use meant designing a data model first, not just charting the raw tables.
Strategic Approach
Phase 1
Data Modelling
Built a star schema in Power Pivot connecting sales, products, stores and customers, so KPI calculations stayed consistent across every view of the report.
Phase 2
KPI Design
Defined revenue, margin and customer engagement KPIs relevant to a global electronics retail business, structured for interactive slicing by region, store and product category.
Phase 3
Dashboard Build
Delivered an executive-facing dashboard with KPI cards and slicers, designed for self-service exploration rather than a static monthly report.
Quantifiable Outcomes
- Modelled 62,884 transactions into a single, query-ready star schema.
- Delivered an interactive KPI dashboard with region, store and product-level slicers for self-service analysis.
- Applied the same data modelling discipline later used in commercial engagements at DashboardWorx.
Qualitative Achievements
- Practised end-to-end BI delivery, from raw transactional data to an executive-ready dashboard, independent of a client brief.
- Reinforced the star-schema and Power Pivot modelling discipline that underpins fast, reliable KPI reporting.
A dashboard is only as reliable as the data model underneath it. Getting the star schema right before building a single KPI card is what makes the rest of the build fast.
This project sits alongside the Woolworths dashboard as evidence of dashboard design discipline built and rehearsed independently, then carried directly into paid commercial engagements.