Supabase as Default Startup Database
Charles Chretien, co-founder of Prequel, on the modern data stack’s ROI problem
The real threat is not Supabase taking big warehouse accounts away overnight, it is Supabase becoming the default starting point for new software companies before Snowflake or Databricks ever enters the picture. Supabase gives developers a cheap Postgres database, auth, storage, functions, local tooling, and branching in one place, which is exactly what fast moving AI app teams need when they are spinning up many small projects and agent workflows. That shifts control upstream to the system where data is first created and stored.
-
Snowflake and Databricks have already reacted as if this upstream layer matters. Databricks acquired Neon in 2025, and Snowflake acquired Crunchy Data and launched Snowflake Postgres, both moves aimed at owning operational Postgres before new developers standardize elsewhere.
-
This follows the same pattern Databricks used against classic warehouses. It first won data teams with managed Spark for raw data engineering, then expanded into SQL warehousing. Supabase could run the playbook one layer earlier, starting with the app database and later adding analytics, replication, or warehouse style workloads.
-
The wedge is especially strong in AI coding. Supabase makes it easy to start local projects with a full stack CLI, run browser based Postgres sandboxes, and pay only when apps get real usage. For small teams, that can remove the need to stand up a separate warehouse during the company’s earliest years.
Going forward, the data stack is likely to split less cleanly between app database and analytics warehouse. If Supabase keeps moving down into analytics while Snowflake and Databricks keep moving up into Postgres and developer tooling, the next battle will be over who owns a startup’s first database, because that choice can shape the entire stack that follows.