Warehouses Building Native Transformation Capabilities
dbt Labs
This is the core platform squeeze on dbt. Snowflake and Databricks are turning transformation from a separate tool into a built in feature of the warehouse itself, which lets them bundle storage, compute, refresh, and governance into one contract and one interface. That does not erase dbt, but it does shift dbt from being the default transformation layer inside one warehouse to being the neutral layer for companies that run across multiple clouds and want one place for business logic, testing, and metadata.
-
Snowflake now offers Dynamic Tables, where a team writes a SQL query for the target table and Snowflake handles refresh timing and pipeline upkeep. Databricks offers materialized views, streaming tables, and Lakeflow Declarative Pipelines, so the warehouse can increasingly do transformation work natively instead of handing it off to a separate product.
-
The economic gap is massive. dbt Labs was at about $100M estimated revenue at the end of 2024, while Databricks was at about $5.4B estimated revenue as of February 28, 2026, and Snowflake reported $2.8B product revenue for fiscal 2025. That scale lets the platforms subsidize new transformation features to protect and expand their core compute businesses.
-
dbt still has a concrete advantage where customers use more than one warehouse or want the same transformation logic, tests, and metrics definitions to travel across Snowflake, Databricks, and other clouds. That is why dbt has expanded from model building into orchestration, observability, cataloging, and a broader data control plane above the warehouses.
The market is moving back from a stack of separate point tools toward warehouse centered suites. Going forward, dbt wins by owning the layer above the warehouse, where teams define business logic once and run it everywhere, while Snowflake and Databricks keep pushing native features deeper into the daily transformation workflow inside their own platforms.