The view materialization in dbt does not store the query's resulting data in the database. Instead, it creates a database view, which is essentially a stored SQL query definition. When this view is queried by a downstream model or a BI tool, the database executes the underlying query at that moment, pulling the latest data from the source tables. This means no data is physically stored for the view itself.

The statement describes the behavior of the table materialization, not the view materialization. A model materialized as a table will execute its SQL and store the results in a new physical table within the data warehouse.

1. dbt Labs Official Documentation

"About materializations": "Views don't store data

but rather run the query behind them each time they are referenced." This section explicitly states that data is not stored for views.

Source: dbt Labs Documentation

Core > docs > dbt-versions > v1.8 (latest) > "About materializations".

Link: https://docs.getdbt.com/docs/build/materializations#view

2. dbt Labs Official Documentation

"table materialization": "When using the table materialization

your model is rebuilt as a table on each run

via a create table as statement... Tables are useful for models that are computationally expensive to build

as they can be referenced by downstream models without being re-calculated." This provides the direct contrast to the view materialization

clarifying which type stores data.

Source: dbt Labs Documentation

Core > docs > dbt-versions > v1.8 (latest) > "About materializations".

Link: https://docs.getdbt.com/docs/build/materializations#table