DORA Metrics

The DevOps Research and Assessment (DORA) team spent six years to identify four key performance indicators for software development teams. Setting these performance metrics as the baseline for your organization can help improve efficiency and effectiveness of delivering work to customers.

DORA Metrics

Deployment Frequency

Elite performers aim for continuous deployment, releasing code to production multiple times per day, regardless of what this metric measures on a team-by-team basis. Deployment frequency refers to how frequently an organization or team releases or rolls out new features to customers.

Mean Lead Time for Changes

It gives a sense of how long it takes for code commits to be deployed to production. This metric can help DevOps and engineering leaders gauge how quick their team's cycle time is and whether they can handle more or less work. It provides a measure of software delivery speed—its velocity—in the same way, deployment frequency does.

Mean Time to Recover

How long does a team recover from an unplanned outage or incident? This data point refers to the team's mean time to recover. Restoring service as soon as possible (with a low mean time to recover) is essential. Elite performers improve this metric by using robust monitoring and progressive delivery practices.

Change Failure Rate

How often production errors occur after a team makes changes is referred to as the team's change failure rate. This metric helps ascertain stability that will drive infrastructure investment decisions. Rollbacks, failed deployments, and quick fixes count toward this metric regardless of their origins.