Member-only story
Data Modelling Basics
What is Data Modelling?
Don’t have a Medium subscription? You can read this article for free using my friend link.
A data model is a structured representation of data, its relationships, and the rules governing it. Think of it like an architectural blueprint—you must design before you build. Without a data model, engineers build in isolation; data becomes inconsistent, siloed and untrustworthy.
Good data models make reporting reliable, pipelines maintainable and teams aligned.
Effective data models ensure reliable reporting, maintainable pipelines nd aligned teams. In analytics engineering, models bridge raw source data and business-facing dashboards.
The Three Phases of Data Modelling
- Conceptual: This phase focuses on identifying the key entities within a domain, without getting into technical details. The audience for this phase is typically business stakeholders.
- Logical: This phase defines attributes and relationships but doesn't tie to any specific database system. Its audience consists of data architects.
- Physical: This phase covers tables, columns, data types, indexes, and database-specific SQL. This phase targets data engineers.
Now, let’s take a closer look at each phase.
Phase 1: Conceptual Model
In the conceptual model, the primary objective is to identify the main domain entities. No data types, columns, or SQL are involved —just…
