dbt's Neutral Control Plane Advantage
Tristan Handy, CEO of dbt Labs, on dbt’s multi-cloud tailwinds
The real advantage is neutrality at the workflow layer. Snowflake, Databricks, and the hyperscalers can each bundle storage, compute, governance, and transformation inside their own stack, but they are still trying to keep data work inside their cloud. dbt sits one layer above that. It manages the SQL models, tests, metadata, and job orchestration that teams want to run across multiple warehouses and clouds without rewriting business logic each time.
-
In practice, cross cloud means a team can keep the same dbt project and operating model while one business unit runs Snowflake, another runs Databricks, and another runs BigQuery. That matters in large enterprises, where multiple warehouse choices often already exist and expansion happens account by account, not through one big standardization event.
-
Warehouse vendors can support open formats, but that is not the same as being neutral. Snowflake added Iceberg tables and Open Catalog, and Google added Iceberg support in BigLake and BigQuery, but each still anchors customers in its own management layer, pricing model, and sales motion. A separate control plane benefits when data is shared across them.
-
Apache Iceberg is the technical unlock that makes this more than a pitch. Before open table standards, multi cloud usually meant copying data twice, paying twice, and creating sync errors. Once major platforms committed to Iceberg in 2024, dbt and pipeline tools could write once in a standard format and let multiple engines read the same underlying tables.
This is heading toward a split market. Data clouds will keep rebundling more native tooling inside their own walls, while control plane companies that stay portable across clouds will own the shared workflow, metadata, and governance layer above them. If open table standards keep spreading, that neutral layer becomes more valuable, not less.