A slowly changing dimension pattern that preserves history: when an attribute changes, you close the existing row (valid-to date) and insert a new row with the new value, so every historical state is queryable.

How does dbt handle slowly changing dimensions?

dbt snapshots implement Type 2 for you — they detect changes in a source and maintain the valid-from/valid-to history automatically.

GlossaryData modeling

Slowly changing dimension (SCD)

Also: SCD · SCD type 2

A pattern for handling dimension attributes that change over time. Type 1 overwrites the old value; Type 2 keeps history by adding a new row with valid-from/valid-to dates.

When a customer changes address or a product moves category, you have to decide what happens to your dimension. Slowly changing dimension (SCD) patterns name the options.

Type 1 simply overwrites the old value — you keep only the current state and lose history. It's simple and fine when history doesn't matter. Type 2 preserves history: instead of updating, you close the old row (set its valid-to date) and insert a new row with the new value and an open valid-to. Facts then join to the dimension row that was current at the event's timestamp.

Type 2 is the one that comes up in interviews and on the job, because “what did this look like at the time?” is a question stakeholders ask constantly. dbt has snapshots specifically to build Type 2 dimensions.

Why it matters

Point-in-time correctness is a real business requirement: finance needs to know a customer's tier at the moment of a sale, not today. SCD Type 2 is how you answer 'what was true then?' without rebuilding history by hand.

Choosing the wrong SCD type is expensive to fix later — once you've overwritten history (Type 1), it's gone. Deciding history requirements up front is part of modeling.

Common mistakes

Using Type 1 (overwrite) when the business later needs historical values you've already destroyed.
Forgetting to join facts to the dimension version that was current at the event time, so historical reports use today's attributes.
Building Type 2 by hand instead of using dbt snapshots, leading to subtle valid-from/valid-to bugs.

FAQ

What is SCD Type 2?: A slowly changing dimension pattern that preserves history: when an attribute changes, you close the existing row (valid-to date) and insert a new row with the new value, so every historical state is queryable.
How does dbt handle slowly changing dimensions?: dbt snapshots implement Type 2 for you — they detect changes in a source and maintain the valid-from/valid-to history automatically.

Go deeper

Related terms

Go deeper in the Data Modeling hub.

Learn this by building, not memorizing.

Definitions get you the vocabulary. The platform gets you the skill — graded exercises, real projects, and a portfolio capstone.

Try free practice See the course →