DORA Metrics: Complete guide to DevOps performance measurement (2025)
Measure and improve your software delivery performance with DORA metrics—deployment frequency, lead time, change failure rate, MTTR, and reliability.
Taylor Bruneaux
Analyst
DORA metrics are four key performance indicators that measure software delivery performance: deployment frequency, lead time for changes, change failure rate, and mean time to recover (MTTR). A fifth metric, reliability, has been added as an additional measurement. These metrics help engineering teams identify bottlenecks, improve processes, and deliver software faster and more reliably.
What are DORA metrics? (Quick Answer)
DORA metrics are performance measurements that help teams deliver software more efficiently and quickly. Created by Google’s DevOps Research and Assessment team, the team identifies what makes high-performing development teams successful.
The four core DORA metrics:
- Deployment Frequency - How often you deploy code (daily vs. monthly)
- Lead Time for Changes - How long from code to production (hours vs. months)
- Change Failure Rate - What percentage of deployments fail (zero to 60%)
- Mean Time to Recovery - How quickly you fix failures (minutes vs. weeks)
Plus one additional metric: 5. Reliability - How consistently your service meets performance goals
Research shows that elite performers who excel in these metrics are twice as likely to meet organizational performance targets. The DORA framework uses these measurements to classify teams as Elite, High, Medium, or Low performers, as outlined in the book “Accelerate” by Nicole Forsgren, Jez Humble, and Gene Kim.
The four DORA metrics in detail
1. Deployment frequency
How often your development team releases code to production.
Benchmarks:
- Elite: Multiple deployments per day
- High: Daily to weekly
- Medium: Weekly to monthly
- Low: Monthly to every six months
Key factors: CI/CD pipeline maturity, automated testing, code review processes, and release workflows.
2. Lead time for changes
How long code takes to go from development to production deployment.
Benchmarks:
- Elite: Less than one day
- High: One day to one week
- Medium: One week to one month
- Low: One month to six months
Improvements: Agile methodologies,automated code reviews, small batch deployments, and streamlined processes.
3. Change failure rate
Percentage of deployments that cause production failures requiring immediate attention.
Benchmarks:
- Elite: Zero to 15% failures
- High/Medium: 16% to 30% failures
- Low: 46% to 60% failures
Improvements: Comprehensive automated testing, quality assurance, test-driven development, staging environments, and feature flags.
4. Mean time to recover (MTTR)
Average time to restore service after a deployment failure.
Benchmarks:
- Elite: Less than one hour
- High: Less than one day
- Medium: One day to one week
- Low: One week to one month
Improvements: Incident response automation, monitoring and alerting, automated rollbacks, and detailed runbooks.
5. Reliability (Additional metric)
Consistency in meeting availability, performance, and accuracy goals.
Reliable services maintain 99.99% uptime, respond quickly to requests, and provide accurate results. Teams measure reliability using uptime percentages, response times, and Mean Time Between Failures (MTBF).
How to measure DORA metrics
Measurement approaches
- Deployment frequency: Count deployments using Git, CI/CD tools like GitLab CI, or Jenkins
- Lead time: Track timestamps from commit to deployment using Jira or Git
- Change failure rate: Calculate failure percentage using CI/CD data and tools like Grafana
- MTTR: Measure incident duration using PagerDuty or OpsGenie
Popular platforms
- GitHub - Built-in insights
- GitLab - Integrated DevOps platform
- Jira - Project management with metrics
- DX platform - Comprehensive measurement tools
Beyond quantitative DORA tracking
DORA metrics provide valuable system data but lack context about why performance changes occur. The most effective approach combines these quantitative measurements with qualitative developer feedback through the DX Core Four framework.
The Core 4 is a balanced measurement strategy that covers DORA metrics within four dimensions: speed (DORA delivery metrics plus perceived productivity), effectiveness (Developer Experience Index), quality (DORA stability metrics plus code quality perceptions), and business impact (ROI and value creation).
By combining system data with developer experience snapshots, organizations gain the context needed to understand whether DORA improvements reflect genuine productivity gains or potential developer burnout.
Understanding DORA performance levels
DORA research has identified four distinct performance levels based on how teams score across the four core metrics. These classifications help organizations benchmark their capabilities and identify areas for improvement.
Elite performers consistently deliver software faster and more reliably, while low performers struggle with lengthy development cycles and frequent failures.
The key insight is that high-performing teams excel across all four metrics simultaneously – they don’t sacrifice stability for speed or vice versa.
Performance | Deployment Frequency | Lead Time | Change Failure Rate | MTTR |
Elite | Multiple/day | Less than one day | Zero to 15% | Less than one hour |
High | Daily-weekly | One day to one week | 16% to 30% | Less than one day |
Medium | Weekly-monthly | One week to one month | 16% to 30% | One day to one week |
Low | Monthly-biannually | One to six months | 46% to 60% | One week to one month |
How AI impacts DORA metrics
The latest DORA research on generative AI reveals a paradox that’s reshaping how we think about software delivery performance. While AI tools are helping individual developers work more effectively, they’re simultaneously creating unexpected challenges for team-level delivery metrics.
The good news: AI is demonstrably improving the development process itself. Teams using AI report a 7.5% increase in documentation quality, 3.4% better code quality, 3.1% faster code reviews, and 1.8% reduction in code complexity. These improvements suggest AI is helping developers produce better work more efficiently.
The concerning trend: Despite these process improvements, DORA metrics show a 7.2% reduction in delivery stability and a 1.5% reduction in delivery throughput when teams increase AI adoption. This counterintuitive finding suggests that while AI makes individual tasks easier, it may be disrupting fundamental delivery practices.
The likely culprit: DORA researchers hypothesize that AI’s ability to generate large amounts of code quickly is tempting teams to abandon small batch principles – a core tenet of high-performing delivery. When developers can produce more code faster, they may create larger, riskier changes that take longer to review and are more prone to failures.
Measuring AI’s actual impact on DORA performance
To navigate this complexity, organizations need more than traditional DORA metrics. The DX AI Measurement Framework provides a comprehensive approach that tracks AI alongside delivery performance:
- Utilization - How extensively teams use AI tools
- Impact - AI-driven time savings and effects on DORA metrics
- Cost - ROI optimization and spend efficiency
Organizations applying this balanced measurement approach are seeing remarkable results. Booking.com deployed AI tools to 3,500-plus engineers and achieved a 16% throughput increase while maintaining delivery quality. Intercom nearly doubled AI adoption rates and realized 41% AI-driven time savings. The key difference: these companies measured both AI benefits and DORA impacts, allowing them to optimize for sustainable performance gains rather than short-term productivity spikes.
DORA best practices
Use DORA to improve teams
Focus on team performance over individuals. Compare current performance to historical trends, not other teams. Use metrics to identify bottlenecks and set incremental goals.
Don’t use DORA for individual evaluation
DORA metrics reflect team capabilities, not individual productivity. Avoid gaming behaviors by focusing on learning rather than targets. For complete productivity measurement, combine with SPACE metrics and developer experience data.
Integrate with comprehensive frameworks
The most effective approach uses the DX Core Four, which balances DORA metrics with effectiveness, quality, and business impact across four dimensions. This prevents speed-only optimization while providing actionable insights. See Google’s measurement principles for additional guidance.
Getting started with DORA metrics
Step 1: Establish baselines
Use the DX Core Four approach - start with self-reported data while building automated collection. This prevents months of dashboard building without value.
Step 2: Implement measurement tools
Automate data collection using existing CI/CD platforms. Most modern tools include built-in DORA tracking capabilities.
Step 3: Set balanced goals
Focus on one metric at a time while monitoring others. If using AI tools, implement the DX AI Measurement Framework to track comprehensive impact.
Step 4: Create feedback loops
Review metrics regularly with teams. Use data to drive improvement discussions and celebrate progress.
DORA metrics FAQ
What does DORA stand for?
DevOps Research and Assessment - Google’s team that developed these metrics through research on high-performing organizations. See the science behind DORA and recent findings.
What are the four DORA metrics?
- Deployment frequency
- Lead time for changes
- Change failure rate
- Mean time to recover
- Reliability (additional metric)
How often should you measure DORA metrics?
Continuously through automated tools. Review trends weekly or monthly. See how top companies measure for best practices.
Can small teams use DORA metrics?
Yes. DORA metrics work for all team sizes, though measurement approaches may vary in complexity.
How do AI tools affect DORA metrics?
AI can improve code quality and review speed but may hurt delivery stability. Use the DX AI Measurement Framework to track utilization, impact, and cost while maintaining quality standards.
Why DORA metrics matter
DORA metrics have evolved from simple delivery measurements to essential strategic tools for navigating the modern software development landscape. As organizations grapple with AI transformation, remote work challenges, and increasing competitive pressure, these four core metrics provide the foundation for sustainable, high-performance software delivery.
The evidence is clear
Elite performers using DORA metrics are twice as likely to meet organizational goals, deliver faster customer value, and maintain higher developer satisfaction. But in 2025, simply tracking deployment frequency and lead time isn’t enough. The most successful organizations combine DORA metrics with comprehensive developer experience measurement to unlock their full potential.
DX transforms DORA measurement
While traditional DORA tracking provides valuable system data, it lacks the qualitative context needed to drive real improvement. DX’s approach combines quantitative DORA metrics with developer experience insights through the Core Four framework – speed, effectiveness, quality, and business impact. This balanced measurement strategy has helped more than 360 organizations achieve measurable results: three to 12% efficiency gains, 14% increases in R&D focus, and 15% improvements in developer engagement.
AI makes this even more critical
As our research shows, AI tools can improve individual productivity while paradoxically hurting team delivery performance. Organizations using DX’s AI Measurement Framework alongside DORA metrics are successfully navigating this complexity, like Booking.com’s 16% throughput increase across 3,500-plus engineers and Intercom’s 41% AI-driven time savings.
DORA metrics are your starting point, but comprehensive measurement is your path to sustainable high performance. DX provides the only platform that combines DORA tracking with developer experience insights and AI impact measurement, giving you the complete picture needed to drive real improvement.
Establish your DORA baselines, implement the Core Four framework, and begin measuring what truly drives software delivery excellence. The data shows that organizations taking this comprehensive approach don’t just improve their metrics – they fundamentally transform how their teams deliver value.