Data model

What is a data model?

A data model is the structure that defines how data is organized, how pieces relate to each other, and what rules they follow. It's the blueprint a database or analytics system is built from. Before a single table exists, the model decides what an "order" is, how it connects to a "customer," and which fields are required.

Data modeling usually moves through three layers. The conceptual model names the entities and their relationships in plain business terms. The logical model adds attributes, keys, and structure without committing to a specific technology. The physical model maps all of that onto real tables, columns, types, and indexes in a chosen database. In analytics work, dimensional models like the star schema arrange data into facts (the measurable events, such as a sale) and dimensions (the context, such as product, date, or store) so reporting stays fast and intuitive.

A good data model makes the right queries easy and the wrong data impossible. A poor one leaks through every dashboard and report built on top of it, which is why teams that skip modeling early tend to pay for it later in confusing numbers and slow queries.

Data models at Dallonses

We start data projects by modeling, not by writing queries. Sitting with a client to agree on what each entity means and how it connects is what stops two dashboards from reporting different revenue figures six months in.

From there we build the warehouse and the transformations that shape raw data into something analysts can trust, then layer data analytics and visualization on top. The model is the contract underneath all of it. Get it right and reporting becomes a matter of asking questions. Get it wrong and every answer needs a footnote. We treat the model as shared work, refined with the people who know the business best.

Numbers that don't agree across reports? It usually starts with the model.

Talk to us about data modeling