Change Failure Rate
Percentage of production deployments causing a degradation or rollback (DORA metric)
What is Change Failure Rate?
Change failure rate is the percentage of deployments to production that cause an incident or need a fix. It shows how reliable an engineering team's releases are.
It is one of the four key measures of software delivery performance, alongside deployment frequency, lead time and time to restore service.
How this KPI is measured
- Function
- Information Technology
- Sub-function
- Software Delivery
- Industry
- All Industries
- Type
- Lagging
- Impact
- Quality
- Unit
- %
- Measurement frequency
- Weekly / Monthly
- Formula
- (Deployments causing incidents / Total deployments) × 100
How to calculate it
Divide deployments causing incidents by total deployments, and multiply by 100.
For example, if 12 of 200 deployments caused incidents, the rate is 6%.
Define what counts as a failure, such as a rollback, hotfix or incident, and apply it consistently. Improving quality should not come at the cost of slowing releases.
Where it is used
Measuring release quality
See how often changes cause problems.
Guiding testing investment
Focus on weak areas.
Balancing speed and stability
Read with deployment frequency.
How to improve it
Tips and common mistakes
- Define a failure clearly.
- Do not slow releases dramatically to cut the rate.
- Read with the other delivery measures.
More Information Technology kpi library
Build better plans, faster
Join the waitlist for early access to curated insights, flexible templates, and rapid iteration across strategy, planning, and project management.