SAP Supply Chain Analytics | MRP MD04/ME5x, MM/PP | Power BI/DAX, SQL | Orders,, Inventory & Procurement | OTIF↑, Lead Time↓, Backlog↓
Why separating “Dim” and “Fact” tables are important in Power BI Many people start with a single table. It works... until it doesn’t. As your model grows, you realise data lives in two worlds: 1️⃣ Fact tables They’re the events: every transaction or movement that happens over time. Examples: sales, orders, production, consumption. They contain dates, quantities, values, and keys that connect to dimensions. 2️⃣ Dimension tables They’re the context: the labels that give meaning to those events. Examples: materials, customers, sales orgs, plants, calendar. They’re what you use to filter, group, and explain. The secret is in how you connect them: • Dimension tables stay on the one side, with no duplicates • Fact tables stay on the many side • Filter direction flows from Dim to Fact Real example: I had to compare monthly forecasts with daily customer orders. So I built a Dimension “Date × SalesOrg × Material” and connected both fact tables (Forecast and Orders) to it. Without that layer, Power BI was summing everything as if all materials and orgs were the same. In the end, a solid model is less about formulas and more about structure. When your relationships are right, DAX almost writes itself.
