In a world where the speed and reliability of software delivery are essential, DORA (DevOps Research and Assessment) has established itself as the benchmark for measuring and improving DevOps performance.
But how do these indicators actually transform the way tech teams are managed? What impact do they have on agility, resilience and the quality of delivery? In this article, we take a closer look at the DORA framework, from its foundations to its practical applications, to help you make the most of your DevOps strategy.
DORA, the benchmark for measuring DevOps performance
Created in 2014 by Nicole Forsgren, Jez Humble and Gene Kim, the DORA (DevOps Research and Assessment) programme has become the gold standard when it comes to assessing DevOps practices.
Backed by years (10 this year!) of research and integrated with Google Cloud, DORA analyses the success factors of top-performing teams.
Its impact? A scientific approach to measuring and improving software delivery through precise metrics.
With its annual “State of DevOps” reports, DORA directly influences the DevOps strategies of companies seeking to combine speed, stability and quality.
The 4 DORA metrics for monitoring DevOps performance
DORA metrics enable you to evaluate the effectiveness of DevOps teams on the basis of key criteria. They offer a balance between speed and reliability of delivery, guaranteeing optimal performance.
Lead Time for Changes
Measures the time needed for a change (code, feature) to go from commit to production. A low lead time encourages agility and continuous innovation.
Deployment Frequency
Indicates how often a team deploys changes in production. Successful organisations deliver on a regular basis, thereby minimising risk.
Change Failure Rate
Evaluates the percentage of production changes that lead to incidents. A low rate means better code quality and optimised risk management.
Mean Time to Recovery (MTTR)
Measures the time needed to restore a service after a failure. Reducing MTTR is essential if you want to limit the impact incidents have on users.
Debates and perspectives on DORA metrics
Each year, the DORA (DevOps Research and Assessment) report provides an in-depth analysis of the trends and best practices that influence technology teams’ performance. The 2024 edition highlights strategic and operational issues, including:
- the impact of AI,
- new performance indicators,
- the role of Platform Engineering,
- the importance of transformational leadership.
Let’s break down these key lessons and look at the implications they may have for tech organisations.
1. AI: productivity lever or factor creating instability?
In recent years, there has been an explosion in the extent to which artificial intelligence has been integrated into DevOps processes. According to the DORA 2024 report, 81% of respondents indicated that their organisation had redefined its priorities to better integrate AI. What’s more, 49.2% of respondents described this change as “moderate” or “significant” in scope.
What the figures say:
The following figures have been established in the context of one person increasing their use of AI by 25%.
Generally speaking, the respondents talk about their individual perception of AI, and it is clear that it is having a beneficial impact:
- Flow up 2.6%
- Job satisfaction up 2.2%
- Productivity up 2.1% (pp. 31-32)
While AI can optimise certain high value-added tasks, it can also accentuate structural problems linked to excessive automation and errors that have not been rectified.
Recommendations:
- Define a clear framework for the use of AI.
- Prioritise the optimisation of critical processes before integrating AI on a large-scale basis.
- Combine AI with rigorous human supervision.
- Give devs time to explore AI at work (through experimentation rather than the training imposed on them).
2. Indicators specific to software development
The DORA 2024 report introduces two indicators that can be used to better assess the performance of DevOps teams:
- Delivery rate: measures delivery frequency and efficiency.
- Delivery stability: measures the reliability of deployments and their impact on production.
The aim of these metrics is to avoid a bias towards speed alone, by reintroducing the notion of quality and reliability.
Surprisingly, in view of previous figures which indicated that using AI was beneficial, the findings of the DORA report indicate that the adoption of AI has a negative impact on software delivery:
- Stable deliveries down 7.2%
- Throughput of deliveries down 1.6%
Recommendations:
- Automate tests to limit failures in production.
- Simultaneous monitoring of flow and stability.
- Short, iterative delivery cycles.
3. Platform Engineering: catalyst or hidden brake?
Platform Engineering is emerging as a key approach when it comes to optimising the software lifecycle. Popularised by companies such as Spotify and Netflix, this model relies on the provision of in-house tools and services (APIs, frameworks, CI/CD) to streamline developers’ workflows.
What the DORA report says:
The overall impact of Platform Engineering is positive:
- Individual productivity up 8%,
- team performance up 10%,
- overall organisational performance up 6% (p. 49)
However, as surprising as it may seem, these beneficial effects are accompanied by some negative ones too:
- flow down 8%
- stability of modifications to applications developed and managed using Platform Engineering down 14%
While Platform Engineering brings gains in productivity, it can also generate risks of complexity and rigidity, hampering team agility.
Recommendations:
- Promote developer independence through self-service tools.
- Set up SLOs and error budgets to limit instability.
- Don’t try to shoehorn one-of-a-kind cases into specific categories
- Establish a culture of continuous improvement and inter-team collaboration.
4. User-centred development: a lever that’s under-used
Many organisations focus on the technical performance of their software. However, the DORA 2024 report stresses the importance of putting the user back at the heart of development, to improve product quality and team motivation.
Key figures:
✅organisational performance for user-centred teams up 40%.
✅ 93% of employees believe that meaningful work is essential.
Recommendations:
- Incorporate continuous user feedback into the development process.
- Keep priorities stable, so that changes are not constantly being made.
- Foster a culture of inter-team collaboration, to enrich perspectives.
5. Transformational leadership: a key asset for performance
The most successful organisations share a common trait: transformational leadership that is capable of inspiring and aligning teams, and guiding them towards excellence.
The impact of transformational leadership:
✅ a 25% rise in transformational leadership = a 9% rise in employee productivity.
✅ Reduce burn-out and improve team satisfaction.
Good leadership is about more than just giving people a vision. It must inspire and motivate people and create an environment that is conducive to innovation and commitment.
Recommendations:
- Define a clear vision that is shared by the entire organisation.
- Encourage team empowerment and autonomy.
- Use data to adjust strategies and drive transformation.
Conclusion: 360° transformation for high-performance DevOps
The DORA 2024 report reminds us that performance relies not only on technology, but also on organisation and corporate culture. AI, Platform Engineering and new metrics offer major opportunities, but they must be adopted strategically and in a controlled way.
By putting the user at the heart of development, stabilising processes and adopting transformational leadership, companies can truly make DevOps a lever for sustainable growth.
What are your organisation’s priorities when it comes to optimising your DevOps practices in 2024?
