Key takeaways from The State Of DevOps Report 2023

The DevOps Research and Assessment (DORA) research program has had a significant impact on the software industry over the last decade. It is the longest-running research program of its kind and every year produces a State of DevOps Report which offers a comprehensive view of the critical capabilities that drive software-delivery success while providing benchmarks that teams can use to assess their own performance. Here are our main takeaways from this year's report.

The latest DORA State of DevOps report draws on responses from over 36,000 professionals across a variety of industries and companies of every size. This year’s research focuses on three key outcomes and the capabilities (or ways of working) that contribute to achieving those outcomes:

  • Organisational performance - the generation of value for customers and the community beyond revenue.

  • Team performance - assessing the ability of teams to create value through innovation and collaboration.

  • Employee well-being - investigating the impact of organisational strategies on reducing employee burnout, and enhancing job satisfaction and productivity.

The report also examines two crucial performance measures:

  • Software delivery performance - the capability of teams to safely and efficiently implement changes in technology systems.

  • Operational performance - the reliability and quality of the experience delivered to users.

In past years, research into performance measures led to the identification of four key metrics which have been widely adopted across the software industry as a tool to drive continuous improvement in teams. One important point that this year’s report reiterates is that metrics and measures are a tool to help guide teams and should not be used to set targets (see Goodhart’s law).

Over the years, the insights from the DORA reports have been transformational across the software industry and in the following sections, we’ve tried to capture some of our key takeaways from the latest report.

Culture Is Key

Culture is foundational to building technical capabilities, igniting technical performance, reaching organisational performance goals and helping employees be successful.

Teams with generative cultures have on average 30% higher organisational performance. A generative culture is one which is performance driven (i.e. a high degree of trust and autonomy vs. command and control) where information flows freely, siloes are broken down and failure is used as a tool for learning rather than attributing blame. In every DORA report, a generative culture is cited as one of the key predictors of team and organisational performance whilst also improving employee well-being.

Cultivating a successful organisational culture hinges on both daily practices and leadership. Practitioners shape culture through their everyday actions and fostering collaboration between teams. This enhances knowledge sharing and trust. Transformational leadership plays a crucial role by emphasising the importance of culture and fostering an environment that supports experimentation and autonomy.

The most effective approach to developing a strong culture combines these elements, blending top-down leadership initiatives with bottom-up contributions from everyday practices. This dual approach ensures that culture is not only driven by strategic leadership decisions but also by the practical, everyday experiences of team members.

Focus On Users

One crucial finding in this year’s report is that a user-centric approach to building software is one of the strongest predictors of overall organisational performance. Focusing on users drives improvements across technical, process and cultural capabilities with teams that have a strong user focus exhibiting 40% higher organisational performance along with a 20% improvement in job satisfaction for employees.

Regardless of how quickly teams are shipping software and it’s perceived quality, if the software does not meet the needs of its users then it is not delivering value. Leaders can help create an environment where focus on users thrives by creating incentive structures that reward teams for delivering value to users rather than just churning features.

Use Of Generative AI Does Not (Yet) Predict Performance

One of the most interesting takeaways from this year’s report is that use of Generative AI (e.g. coding assistants) has not yet shown to predict an improvement in performance. The researchers themselves attribute this to larger enterprises typically being slower to adopt new and emerging technologies so this will likely change as more data becomes available in the next year or two.

Invest In Technical Capabilities

Investing time and resources into developing technical capabilities such as continuous integration and delivery, trunk based development and loosely coupled architectures is likely to lead to many beneficial outcomes such as improved organisational, software delivery and operational performance.

Teams that reduce their code review times achieve up to a 50% improvement in software delivery performance. Streamlining code reviews not only enhances the quality of the code but also facilitates knowledge sharing, promotes collective ownership, encourages team responsibility and increases transparency in the development process.

High quality documentation is also extremely important and amplifies the impact technical capabilities have on organisational performance. Most notably, trunk based development is estimated to have a 12.8x greater impact on organisational performance when compared against low quality documentation.

Whilst enhancing technical capabilities is important, it is also worth noting that a balance must be struck between software delivery and operational performance along with user focus to see the best results at an organisation level while still improving employee well-being.

Cloud Increases Flexibility

Perhaps one of the least surprising findings is that leveraging cloud platforms provides flexibility which in turn predicts performance. Utilizing a public cloud platform increases infrastructure flexibility by 22% compared to other environments. Although using multiple cloud platforms also enhances flexibility, the increase is less significant than using a single cloud platform. This is because managing multiple cloud platforms can add overhead and complexity in managing governance, compliance and risk.

Cloud computing's real value lies in its ability to offer flexible infrastructure, which is crucial for improving organisational, team, software delivery, and operational performance. Many organisations start by migrating their existing infrastructure to the cloud (aka lift and shift) but to fully benefit from cloud computing, it's essential to modernise and refactor these workloads to exploit the cloud's flexibility.

Success in cloud computing isn't just about transferring workloads to the cloud; it's about leveraging the cloud's core characteristics, such as on-demand self-service, broad network access, resource pooling, rapid elasticity, and measured service. This approach requires rethinking how applications are built, tested, deployed, and monitored to maximize the benefits of cloud computing.

Distribute Work Fairly

The report reveals that individuals from underrepresented groups, including women and those who self-describe their gender, experience higher levels of burnout. These individuals are also more likely to engage in repetitive work which contributes to increased burnout. Specifically, underrepresented respondents report 24% higher levels of burnout and perform 29% more repetitive tasks than their counterparts, while women and those who self-describe their gender do 40% more repetitive work than men.

To help reduce levels of burnout, organisations can implement formal processes to distribute work evenly and while the research here demonstrates some positive results, this does not extend to all underrepresented groups so this will likely be a focus of next year’s study.

Flexible Working

The Covid pandemic forced most of us into a period of fully remote working and since then the DORA research program has been studying the impact of working arrangements on employee well-being, organisational and team performance. Out of all respondents this year, nearly 33% worked exclusively from home with only 9% working primarily from an office and roughly 63% of respondents working from home more often than an office.

Since the start of the pandemic, the DORA research has found no conclusive evidence one way or another to indicate that where an individual works has an impact on team or organisational performance but the authors are also clear that there are a lot of factors at play here. However, what the evidence does show is that having flexibility over how, when and where an individual works does correlate with an increase in value delivered to users and improved employee well-being. This also extends to new hires where office working during onboarding does not demonstrate an observable increase in performance.

Wrap Up

This year’s report, as always, contains a wealth of information and we’d encourage anyone working in the software industry to study it for themselves but hopefully we’ve provided a brief overview of some of the key findings!

You can find the report here:

Article By
blog author

Chris van Es

Head of Engineering