What is a data model?
Definition
A data model describes entities, their attributes and their relationships. For example, customers can place orders, and each order can contain several line items.
Models can describe business concepts, logical structures or physical database tables. A semantic layer adds business definitions that analytics tools can use on top of those structures.
Why it matters
A model helps people understand which records belong together before they start calculating. For example, an order can have several line items and several payment events. Treating those as though they were all one-to-one relationships can produce a query that runs successfully but multiplies the values being added.
Document what one row represents in each important table, how identifiers are used and which relationships are optional. Include the meaning of dates and statuses where those affect a calculation. This gives analysts a more dependable starting point and provides useful context for AI-generated queries. Keep the model aligned with source changes, so documentation does not quietly drift away from the structures people are actually querying.