dbt Becomes System of Record
Julia Schottenstein, Product Manager at dbt Labs, on the business model of open source
This is why dbt became a system of record for business logic, not just a helper for moving data around. Fivetran solves the mechanical job of copying rows from Salesforce or Stripe into Snowflake. dbt solves the harder organizational job after that, deciding how revenue, active users, churn, or pipeline should actually be calculated once, documented, tested, and reused across dashboards, notebooks, and teams.
-
The split maps to different users and workflows. EL tools are usually set up by data engineers who care about connectors, sync frequency, and broken APIs. dbt is used by analytics engineers and analysts who spend their day editing SQL models, reviewing pull requests, testing definitions, and keeping metrics consistent across BI tools.
-
Vendor neutrality is the product advantage. When a company has multiple warehouses, multiple BI tools, or a multi cloud setup, hard coding business definitions into Snowflake, Looker, or one ingestion vendor creates duplication. dbt keeps metadata and transformation logic one layer up, so the same definitions travel across the stack.
-
This position is strategically valuable, which is why the category keeps rebundling around it. dbt reached about $100M ARR in 2024, while Fivetran reached about $325M, and both expanded beyond their starting point into adjacent workflow tools. The fight is over who owns the daily interface where teams define and govern trusted data.
The next step is for transformation to become the control plane above warehouses and BI, with orchestration, catalog, observability, and semantic definitions bundled together. As multi cloud use rises and open table formats make data more portable, the company that owns business logic above the infrastructure layer will own the highest leverage point in the data stack.