Data modelling

Data modelling is the process of structuring, describing, and visualizing how data is organized, stored, and related within an information system. Its purpose is to create a consistent and logical representation of real-world entities such as customers, products, or transactions, enabling efficient use of data across databases and analytics platforms.

• Conceptual model: High-level description of business information requirements, independent of technology.

• Logical model: Structured design with entities, attributes, and relationships, often using ER diagrams (Entity-Relationship).

• Physical model: Technical implementation adapted to a specific database or platform.

Data modelling supports database design, data warehouse development, and business intelligence. It ensures data consistency, quality, and reusability — foundational elements of sound data governance.

The discipline emerged in the 1960s alongside early database systems. In the 1970s, the relational model introduced by E.F. Codd revolutionized the field, forming the basis for SQL databases. In recent decades, object-oriented, semantic, and graph-based models have expanded their reach.

Within Microsoft’s ecosystem, data modelling is central to SQL Server, Azure Synapse Analytics, Power BI, and Dataverse. Power BI’s tabular model (based on DAX) enables efficient relationship management and calculations across large datasets. Tools like Visual Studio and Azure Data Studio provide support for logical and physical modelling workflows.

Data modelling is a cornerstone of modern data management. Well-designed models transform raw data into structured knowledge, improving analytics, governance, and decision-making.