Core Principle

DORA (DevOps Research and Assessment) provides the most empirically grounded framework for measuring and improving software delivery performance. It consists of five delivery metrics and 35+ capabilities that predict performance. The central finding: speed and stability are not a tradeoff; they reinforce each other.

Why This Matters

DORA gives concrete, measurable language for arguments about engineering practices. Instead of “we should use trunk-based development because it’s better,” you can say “DORA’s research across thousands of organizations shows trunk-based development predicts superior delivery speed, stability, and availability.” The framework also situates itself within a broader measurement landscape alongside SPACE and DevEx.

Evidence/Examples

The Five Metrics

MetricMeasuresCategory
Change Lead TimeCommit to production durationThroughput
Deployment FrequencyDeploys per time periodThroughput
Failed Deployment Recovery TimeTime to restore after failureStability
Change Fail Rate% deployments needing remediationStability
Deployment Rework Rate% unplanned deploys from incidentsStability

Key Capabilities (35+)

Technical: trunk-based development, continuous integration, continuous delivery, test automation, deployment automation, infrastructure-as-code, monitoring and observability, database change management, code maintainability, version control

Process: working in small batches, streamlining change approval, team experimentation, customer feedback, WIP limits

Cultural: generative organizational culture, learning culture, transformational leadership, job satisfaction, well-being

Relationship to SPACE and DevEx

Nicole Forsgren: “DORA is an instance of SPACE, as a measure of the performance of software.”

  • DORA (2014-): Team/org delivery outcomes via five metrics
  • SPACE (Forsgren et al., 2021): Five dimensions: Satisfaction, Performance, Activity, Communication, Efficiency
  • DevEx (Greiler, Storey & Noda, 2022): Developer-centric perception data alongside system metrics

Implications

  • Use DORA metrics as leading indicators for engineering practice decisions
  • No single metric captures productivity; use SPACE dimensions to avoid Goodhart’s Law
  • DORA’s 2025 finding: “AI acts as an amplifier, but the greatest returns come from focusing on the underlying sociotechnical systems”

Questions

  • How do DORA metrics apply to solo developers or very small teams where “deployment frequency” may not be meaningful?
  • The 2025 AI amplifier finding: does this mean infrastructure maturity is prerequisite to AI productivity gains?

Sources

  • DORA Team. dora.dev
  • Forsgren, N., et al. “The SPACE of Developer Productivity” (2021). ACM Queue.
  • Greiler, M., et al. “An Actionable Framework for Understanding and Improving Developer Experience” (2022). arXiv:2205.06352.
  • Forsgren, N., Humble, J., & Kim, G. Accelerate (2018). IT Revolution Press.