dbt exports materialize saved MetricFlow queries as warehouse tables

Exports extend saved queries by materializing their output as a table or view. The job scheduler writes those results into the warehouse so tools without native Semantic Layer support can read metrics and dimensions over SQL. Prerequisites include Starter or Enterprise-tier plans, supported platforms (Snowflake, BigQuery, Databricks, Redshift, Postgres), dbt 1.7+, a configured Semantic Layer, a scheduler-enabled environment, and a configured saved query plus export. Caching is recommended for common queries. A stated benefit is DRY representation versus many denormalized metric mart tables.

Data engineering teams can publish governed metric outputs to consumers that only speak SQL, without duplicating metric logic in each mart.

Materialized exports carry shared metric meaning into the warehouse as ordinary relations, bridging the Semantic Layer and systems that never call MetricFlow directly.