What are Dataverse environments and how many do I need?

By Emil Björk · Microsoft business apps consultant, Gothenburg

An environment is an isolated Dataverse database plus the security, apps, and flows around it. The standard split is three: Development (unmanaged solutions, iteration), Test/UAT (managed solutions from Dev, business testing), Production (managed solutions from Test, end users). Larger programmes add a per-developer sandbox and a dedicated release environment; nobody serious ships from a single environment.

Environments are the unit of ALM in Dataverse. Unmanaged solutions live in Dev; every downstream environment receives managed solutions moved through a pipeline (GitHub Actions, Azure DevOps, or Power Platform Pipelines). Skipping the split and editing Production directly is the single most common cause of unrecoverable damage in this stack — there is no reliable 'undo' once you have.

For very small teams a two-environment setup (Dev → Prod) is defensible if the release cadence is slow and users tolerate the occasional Dev-caused outage. Every organisation past a handful of makers should run three.

Go deeper

Other questions