> ## Documentation Index
> Fetch the complete documentation index at: https://docs.growthbook.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Setting Up a Custom Data Source

> Write assignment and metric SQL for a custom event schema so GrowthBook can analyze experiments in your warehouse.

GrowthBook just needs you to write a couple SQL queries in order to query your data. Writing this SQL is a (mostly) one-time setup task. After building out this library of queries, they can easily be reused across many experiments.

Don’t worry about the potentially huge number of rows returned by these raw queries. They are never run directly as-is and are instead combined, filtered, and aggregated as part of larger queries. Most of the final queries run by GrowthBook result in fewer than 10 rows returned.

In the spirit of transparency, any time a query is run, you should see a `View Queries` link in the app to view the raw SQL sent to the data warehouse. This can help with debugging or let you move a query into a tool like Mode Analytics for more advanced analysis.

See [data source configuration settings](/app/datasources#configuration-settings) for what to configure.