Measure

Change failure rate

You might say: failed change percentage, change fail rate

The percentage of production deployments that cause a failure. Define the failure signal and remediation window before comparing periods. Most outages come from changes, so a factory that merges more also needs progressive rollouts and fast rollback after merge. That work sits outside the factory, which ends at merge.

What it does

Shows whether faster delivery is producing changes that need rollback, hotfix, or other recovery work.

Ask

Which deployments caused a production failure, and what counts as a failure for this service?

What to look for

1. A percentage of deployments marked failed
2. Links from a failed deployment to incident, rollback, or fix-forward work
3. A trend beside lead time and deployment frequency

Sources

1. DORA, A history of DORA’s software delivery metrics, From four keys to five (2024)Change fail rate: Percentage of deployments causing failures in production.
2. Betsy Beyer et al., Site Reliability Engineering, Chapter 1, IntroductionSRE has found that roughly 70% of outages are due to changes in a live system. … Implementing progressive rollouts … Rolling back changes safely when problems arise

Use it on your project

teardown-apply: hold "Change failure rate" (change-failure-rate) from the teardown at https://teardownatlas.com/software-factory/ against my project

Paste it into a coding agent that has the teardown skills.

An AI agent wrote this map with the teardown skill. It can contain errors: check the sources before you rely on a part. About and legal