How to Reach Elite DevOps Efficiency: Insights from the 2019 State of DevOps Report

Your R&D efficiency compared to industry?

The report classifies organizations into low, medium, high, and elite efficiency based on four key metrics: deployment frequency, lead time for changes, mean time to recovery (MTTR), and change failure rate. Elite organizations deploy on demand, achieve hour‑level lead times, keep MTTR under an hour, and maintain a failure rate below 15%.

How to become an elite‑efficiency organization?

There is no one‑size‑fits‑all DevOps solution; each company must address its unique context, constraints, and bottlenecks. The report recommends a five‑step improvement loop:

Identify the most pressing problem causing delays.

Use an efficiency‑improvement model to select targeted practices that remove constraints.

Allocate dedicated experts to solve the identified problems.

Focus exclusively on the biggest issue before moving on.

Repeat the cycle once the constraint is resolved.

Two key efficiency‑improvement models

The 2019 report introduces the long‑standing Software Development & Operations Efficiency Model and a new Engineering Productivity Model . Both models map causal relationships that drive higher delivery performance, such as better technical practices and improved change‑management processes.

Software Development & Operations Efficiency Model

This model emphasizes capabilities like automated testing, continuous integration, continuous delivery, monitoring, loosely coupled architecture, and cloud usage. The 2019 edition adds change‑management processes, fault‑recovery testing, code maintainability, and psychological‑safety culture.

Heavy change‑management processes often hinder delivery. The report shows that formal change‑approval boards (CAB) correlate with lower efficiency, while lightweight peer‑review combined with automated checks improves outcomes.

Engineering Productivity Model

Productivity is defined as the ability to complete complex, time‑consuming tasks with minimal interruption. The model highlights four practice areas: usable tools, effective search, technical‑debt reduction, and a culture of psychological safety.

Reducing technical debt

Technical debt appears in code, scripts, configurations, and infrastructure. Common debt includes unresolved defects, insufficient test coverage, poor code quality, abandoned artifacts, lack of domain knowledge, incomplete migrations, obsolete technologies, and missing documentation.

Addressing debt requires continuous refactoring, supported by tools such as Facebook’s fastmod and Google’s ClangMR.

Conclusion

The 2019 State of DevOps report (82 pages) demonstrates that adopting DevOps practices and the two efficiency models can dramatically improve software delivery performance and engineering productivity. Organizations should continuously identify bottlenecks, apply the models, and aim to become elite‑efficiency teams.