Top takeaways from the 2024 DORA Report – sponsored by Gearset!

I’m excited to share that Google's 2024 DORA Report — the definitive guide to understanding how teams around the world are adopting DevOps practices — has been released! As one of Gearset’s DevOps Advocates, I’m thrilled that we’ve sponsored this invaluable resource that continues to shape the future of software delivery.

For teams striving to improve their performance and adopt industry best practices, the DORA report is an essential tool packed with data-driven insights into what high-performing teams are doing differently.

This year’s report reveals several key findings that shed light on how development teams are operating. It covers the continued rise of elite performers, the impact of artificial intelligence on delivery teams, the growing importance of security in the software delivery lifecycle, and the critical role of cultural practices in driving team success.

Gearset believes that empowering teams with the right tools and processes is crucial for delivering better software, faster. The DORA report provides the insights teams need to make informed decisions on their DevOps journey. Let’s dive into some of the highlights!

What are the DORA metrics?

The DORA report continues to focus on four key metrics that indicate the performance of software delivery teams, categorized into throughput and stability metrics.

Throughput:

• Change Lead Time: Time taken from committing code to deploying it
• Deployment Frequency: How often code is deployed

Stability:

• Change Failure Rate: Percentage of changes that fail and need remediation
• Failed Deployment Recovery Time: Time to restore service after a failed deployment

Elite performers can deploy multiple times a day, recover from failures in less than an hour, and have change failure rates as low as 5%.

High-performing teams excel across all four metrics, while low performers struggle in these areas.

The adoption of artificial intelligence

It’s clear that the hype around artificial intelligence (AI) isn’t going away anytime soon. That’s reflected in this year’s report in terms of adoption numbers, but it also highlights some intriguing insights as to how much AI technology is useful and trusted.

In terms of adoption, 75% of respondents said that they rely on AI for at least one of their daily activities. Code generation and summarizing information were the activities which AI was most widely used for here, which resonates largely with what we’re hearing from across the Salesforce ecosystem.

Coinciding with adoption, 70% of respondents said they were more productive thanks to AI, which is also a positive indicator of how AI might be adopted for many teams.

Trust in AI is fragile

While adoption of artificial intelligence in day-to-day work is high, there are still obvious trust issues. 40% of respondents said that they had little to no trust in AI generated code.

There seems to be a long way to go in building this trust. I’d hazard a guess that it won’t be until we see more real-world use cases and robust success stories using AI that trust will increase. I also predict that we’ll see a rise in adoption of code scanning and analysis tooling so that teams are reassured by the quality of AI code, should they choose to deploy it.

Deploying AI into applications is also hampering throughput and stability, two key areas measured by the DORA metrics. In the survey, a 25% increase in AI adoption showed a 1.5% decrease in throughput and a 7.2% decrease in stability. Concrete reasons for this are unclear. However, it could be attributed to bigger batches of changes being deployed for new AI applications, rather than more granular deployments that lend themselves to increased deployment speed and stability.

Despite the delivery performance drawbacks, AI contributed to a 2.3% increase in organizational performance and a 1.4% increase in team performance overall. Though small, it does demonstrate the potential here.

Platform engineering could be on the rise

Platform engineering is an emerging discipline which is proving to be a successful approach for teams looking to improve their software delivery performance, albeit with some tradeoffs. My fellow DevOps advocate, Rob Cowell, wrote a great piece recently on what platform engineering is and how it’s applicable to Salesforce, which I highly recommend reading.

Implementing internal developer platforms increased individual developer productivity by 8% and team productivity by 10%. However, there was a trade-off with a decrease in change throughput (8%) and stability (14%). There are hypotheses outlined in the report as to why this tradeoff exists, but signs show that this emerging discipline is likely to evolve further.

Teams that enabled developer independence (self-service platforms which form the basis of platform engineering) saw a 5% productivity increase. It’s important to note that feedback collection from developers using platforms was essential for the ongoing improvement and success, however. Teams should be mindful of good engineering practices, such as robust testing, before adopting a platform engineering approach.

Developer experience is the beating heart of success

This section of the report made me extremely happy to read. Long have I advocated for people-focused processes and tooling, and this year's DORA report validates that.

Users are at the heart of what we do as technologists. The report found teams that focus on user needs not only see improvements in productivity, but in product quality, too. Additionally, a user-centric approach was shown to reduce burnout for developers and ensure that products meet real user needs, leading to higher job satisfaction.

Unfortunately, burnout is a topic that’s all too familiar in engineering circles. The report also found that unstable organizational priorities — where organizations shift their focus frequently — lead to increased burnout. Perhaps a cautionary tale for any organization which has one foot in and one out of the AI camp.

In organizations where shifting priorities are frequent, 90% of teams experience drops in productivity. Strong leadership and documentation practices help, but cannot fully mitigate the burnout risks generated from unstable priorities. There’s a fantastic section in the report on transformational leadership, which I encourage you to take a close look at.

Read the full report

DORA’s tagline is “Get Better at Getting Better”. This report provides a mechanism to help you do just that.

Take a step back, evaluate your team's performance, and see what you can do to improve in 2025. Have a read and let me know what resonates with you!

Plus, don’t miss our free DevOps Launchpad certification, Assessing and improving DevOps performance, which provides practical steps on how to apply the DORA metrics in your own organization and get the most out of your DevOps process.