Reviewing DORA’s 2019 Accelerate State of DevOps Report

Diffblue HQ
3 min readJan 21, 2020

--

The 2019 Accelerate State of DevOps Report from DevOps Research and Assessment (DORA) was released in August, and the results revealed that a large number of companies have advanced in their DevOps maturity since the previous report in 2018.

The content of the study is primarily of interest to tech leaders and decision-makers, including the development of a new measure called Software Delivery and Operational Performance (SDO) and analysis of the relationships between performance metrics (lead time, deployment frequency, time to restore, etc.). But many of the points made in this report also have immediate applications for developers who want to enact positive change in their organization’s software delivery efforts.

What are the most important takeaways for developers?

Here are the main points we’ve identified:

1. Developers should advocate for the right productivity tools

Two research models are developed in the report, which are each meant to be used to achieve one of two goals: ‘SDO and Organizational Performance’ and ‘Productivity.’ For developers, productivity has more of an immediate impact, as it can “drive improvements in work/life balance and reductions in burnout.” Starting on page 56 of the report, the model for improving productivity is broken down into further goals, such as reducing technical debt and burnout.

The main finding related to the model is that useful, easy-to-use tools make all the difference to productivity. The highest performing engineers are 1.5 times more likely to have easy-to-use tools and use a mix of proprietary tools, open source, and commercial off-the-shelf software. Specifically, the attributes that drive productivity are how easy it is to use the toolchain, and how useful the toolchain is in accomplishing job-related goals. Automation and integration are key.

The barrier to getting the right tools for many developers is often a disconnect between what developers need and what the people who purchase tools for developers assume they need (or whether they instead prioritize other factors, like cost). Empowered development teams can advocate for access to the tools that will actually be useful to them, both for their own benefit and the benefit of the company.

2. Becoming an “elite” performer requires continuous improvement

The report this year identified growth in the group of “elite” DevOps performers, characterized by far more frequent code deployments, much faster lead time from commit to deploy and time to recover from incidents, and a lower change failure rate than the low performing groups.

The simple but effective advice offered by the report for how to improve performance: “Start with foundations, and then adopt a continuous improvement mindset by identifying your unique constraint (or set of constraints). Once those constraints no longer hold you back, repeat the process.” This mindset can be especially effective when all members of the team adopt it.

3. Wider industry trends to be aware of: DevOps adoption is still on the rise

DevOps adoption has been growing for years, and the “medium performers” group has grown while the “low performers” group has shrunk. With the velocity of DevOps expertise increasing as well, most developers can expect to see this reflected in their organizations in the coming years.

A Common Theme

The report ultimately suggests that organizations can support productivity by fostering “a culture of psychological safety” and making “smart investments in tooling, information search, and reducing technical debt through flexible, extensible, and viewable systems.”

These results align with what we learned in our own 2019 developer survey, where 300 respondents indicated that the achievability of software quality at low cost and high speed requires access to the right automation tools.

--

--

Diffblue HQ
Diffblue HQ

Written by Diffblue HQ

Diffblue Cover autonomous AI-powered Java unit test suite generation & maintenance at scale. | We’re making developers wildly more productive with AI for Code

No responses yet