The DORA 4 key metrics become 5

Metrics are better than blank walls

Blank wall retrospectives are often used to drive the improvement process. The team will undoubtedly come up with ideas that improve their experience. Any motivated team of people will find ways to enhance their software delivery. Often, though, these improvements don’t translate into improved outcomes.

The DORA metrics can help with this. Using measures for throughput (the flow of features) and stability (the ability to deliver without disruption), they identify specific areas that need improvement, and the statistical model of practices and capabilities produced by the research offers many ideas on what specific changes you could experiment with to improve things.

This creates a complete feedback loop. For example, the metrics may highlight that you have a high change failure rate. If you frequently have to scramble to fix problems each time you deploy, you’ll disrupt the flow of valuable work, upset the software’s users, and damage your team’s reputation within your organization.

When an organization doesn’t trust a development team, it starts adding heavyweight change control practices and irritating procedures. Nobody wants this because the research found that this leads to even worse outcomes than those that made people want to assert more control.

So, the best way to use the metrics is within the team as part of the continuous improvement process. The numbers tell you where to look, the statistical model has some suggestions on how to improve (and you’ll have your own ideas, too), and you run an experiment to see if making a change to how you’re working results in an improvement.