We asked 10 tech leaders what DevOps teams need to know heading into the new year.
Delivering better software, faster remains the primary driver for engineering leaders everywhere, regardless of the macroeconomic environment, tech layoffs, or the framework you use to get there. In 2023 the platform engineering hype train kept rolling, as organizations searched for new ways to unlock their engineering teams’ full potential.
As we move into 2024, what should DevOps teams keep in mind as they look to push themselves in a challenging environment? Hint: It’s not a new language or tool. And there’s no panacea or magic pill. Engineering remains not just a technical but a sociotechnical endeavor which can’t be 100% automated.
We spoke to ten DevOps experts to find out what they predict for the year ahead, and how you can prepare to make the most impact possible in 2024.
DevOps is a team game
Fourteen years in, a lot of organizations have failed in their DevOps transformation because they’ve been too focused on the technology. Yes, continuous integration and delivery and even cloud migrations are often involved. But you could have CI/CD or even be working in the cloud without DevOps. As the godfather of DevOps, Patrick Debois said: “DevOps is about removing the friction between silos. All the rest is engineering.”
Moving into 2024, DevOps experts are united on the fact that software development is about team effort, not technology.
That team is also getting bigger. “DevOps culture will take even more cross-team collaboration in 2024,” said Anna Daugherty, director of product marketing at continuous orchestration platform vendor Opsera. “Now, DevOps is reaching further into the business outside of developer teams. Business leaders are becoming more exposed to the needs and benefits of these processes.”
Crucially, that cross-functional expansion requires a common language, one where teams “bring their best data and conversations to leadership meetings to show how DevOps positively impacts the business,” Daughtery said.
Align authority and responsibility
One of the biggest hurdles to DevOps transformation is and always has been leadership. In a time when developer autonomy is highly desired, everyone must also claim responsibility.
“It takes stern, consistent, unyielding leadership to rewire the organizations while the worker bees learn the technical skills they have been thus far able to avoid,” Kent Beck, the founder of Extreme Programming, said.
“If you decide to take a shortcut, you are the one who wakes up in the middle of the night. Organizations – dysfunctional ones – tend to separate consequences from actions. The boss makes a promise, I don't keep it, I get punished. I have an idea, some super confident white guy takes the credit. I make a programming mistake, some poor operations person gets awakened in the middle of the night.”
A good place to start is by aligning authority with responsibility, where “consequences flow towards those who make decisions,” Beck said.
There is never an end state
Another core value of DevOps is kaizen or continuous experimentation, learning and improvement.
“For teams that develop, deliver, and operate software, progress is made from ongoing improvements. Some of these improvements may come from large initiatives, like setting up a CI/CD pipeline, but there’s never an end state,” said Michelle Irvine, technical writer at Google, and part of the core DevOps Research and Assessment (DORA) team.
“A healthy culture has a positive impact on all of the key outcomes we study, including team, software delivery, and organizational performance. Culture is the foundation that you need to foster, including generative culture practices, even work distribution, and increased knowledge transfer.”
The tricky part is identifying your own particular bottlenecks, and “where to invest in change, including how you incorporate new technology like AI, assess both the capabilities that drive improvement and your own context,” she said.
Remember your roots
While AI may look like a tempting skeleton key to unlocking developer productivity, DevOps experts see 2024 as a good year to get back to the basics.
“Staying close to the roots of DevOps methodology will give these teams a step up,” said Abby Bangser, principal architect at Syntasso. “Because implementing faster feedback loops and tighter collaboration across the organization is the key to leveraging any new technologies that become available.”
This focus on the basics will also last beyond whatever the latest technology hype cycle is. “As long as you keep in mind that your core mission is to remove the friction, there will always be a place for you in the organization, even when platform engineering becomes the norm, and some other concept comes along. In the end, those are only tools in your toolbox,” said Ivan Krnic, director of engineering at CROZ and host of the 0800-DEVOPS podcast.
This year’s DORA report found that it didn’t really matter which specific methodology you’re following, but having a continuous focus on users increases organizational performance.
“I heard a great phrase by Amy Willard and Matt Ring from John Deere,” Krnic said. “They call such skills the ‘no regret skills.’ It's such a good way to describe them, because no matter what we do in an organization, we have our users and need to understand how we can create value for them.”
For Andrea Goulet, cofounder and CTO at Heartware, the current climate of smaller teams and fewer resources is even more reason to focus on effective collaboration. “DevOps teams should get back to the fundamentals by investing in organizational cultures that support its practices,” she said. “This requires a holistic approach, such as the SPACE framework, to actively counteract the reductionist temptation to pin problems on individual contributors or tie success to a single magic metric. DevOps has always been about developers and operations working together – and it’s more important than ever not to lose sight of that.”
Developer experience is vital
If you want to move fast and not break things then it’s important to continue to focus on developer experience in 2024.
The popularity of platform engineering in 2023 may have been spurred on by the spate of tech layoffs across the industry, but the demand for developer autonomy hasn’t changed. Internal developer platforms and portals are only adopted once they are proven to reduce friction, frustration and interruptions, so workers can enter that coveted flow state.
“Looking in my crystal ball of things that need to be in every DevOps toolkit in 2024, the concept of software templates instantly comes to mind,” said Helen Gruel, head of engineering for the open source Backstage project at Spotify. These templates are designed to combat the complexity of things like service creation and look to instill standards and clarity across an organization. “They are truly force multiplier capabilities, making them a powerful augment for any organization looking to scale efficiently,” she said.
Efforts to automate or pave that “golden path” not only reduce time to deliver value, but consistently reduce developer friction.
“Instead of being inbound request-driven, you can build a menu of capabilities that users can select from in a self-serve manner,” Gruel continued. This helps improve both developer and ops experiences.
“It won’t be a one-and-done situation, but rather a long-term investment – these platforms will need to be tweaked and refined on an ongoing basis, depending on developers’ needs,” Heather Joslyn, editor in chief at The New Stack said. “Adopting platform engineering successfully will require a cultural shift on DevOps teams.”
Generative AI will drive productivity
According to this year’s GitHub Octoverse report, 92% of U.S.-based developers are already using AI-based coding tools. But has your organization created and communicated a GenAI policy yet? Have you considered and talked about the security and privacy implications as an organization?
“DevOps is all about enabling an ever-increasing number of silos to work together, which just keeps growing, including now generative AI and data science,” said Gene Kim, author of the DevOps Handbook and Wiring the Winning Organization. “This makes the field more exciting than ever, and continues to help push the frontiers of performance.”
DevOps teams need to be aware that generative AI not only lacks provenance, it also often lacks accuracy. Indeed it could be introducing errors and security vulnerabilities at scale. Now, more than ever, critical thinking is a valued skill on any team.
“In 2024, the evolving dynamics of AI within the software supply chain are compelling a reevaluation of how security and transparency are fundamentally interlinked, particularly in open-source environments,” Sal Kimmich, CEO at GadflyAI, said.
It’s the role of a good DevOps team to temper this excitement with a well-communicated generative AI policy so that whole organizations can safely get the most out of this new, potentially world-changing tool. As with all things DevOps past, present and future, it will come down to cross-organization communication and collaboration.
What would you add to our list of what DevOps teams need to know for 2024? Post on LinkedIn with #DevOps2024 so we can continue this discussion into the next year.