DevOps has evolved in response to a growing demand for greater business agility. It is, as the late Robert Stroud described it, an imperative for survival. But changing institutionalized behavior, he said, is the "toughest of all management challenges." And, indeed, as organizations adopt DevOps ways of working, many are failing because of culture issues.
In fact, the three top reasons DevOps initiatives fail—according to feedback from more than 400 teams over the last year that participated in our Phoenix Project DevOps simulation—are all related to culture. Here are those causes, along with tips on how you can avoid a similar fate.
1. Failure to instill a DevOps culture
DevOps requires new ways of working and behaving, especially across dev and ops teams that traditionally have operated in silos—and don't trust each other. Suddenly, they all have to start collaborating, and they're probably not great at it.
Simply telling people they need to collaborate is like telling a five-year-old to tidy up her room and expecting it to happen. One way to address this is to get your teams together to identify which behaviors demonstrate effective collaboration, and commit to practicing those behaviors.
But collaboration also requires feedback and coaching. Managers are often either poorly equipped to do so or give too little attention to help these behaviors become "the way we do things here." It’s a significant cultural shift, and changing your company's culture takes time and effort.
2. Failure to apply continual learning and improvement
DevOps is not something you can simply install and implement overnight. You don't just follow DevOps practices or install a DevOps tool set and consider it done.
[ Special Coverage: Focus on IT Service Management at Pink19 conference ]
Organizations that have derived sustainable value from DevOps have been on that journey for many years. Key factors behind their success lie in experimentation and continual learning and improvement. Unfortunately, our global surveys reveal that fewer than 20% of organizations have effective end-to-end continual service improvement. This needs to be a core capability.
Even if you're not doing DevOps, you can practice retrospectives by getting your team together to explore what went well and what needs improvement. Have them make a visual improvement or impediments backlog, then prioritize that list and reserve time to work on it.
That sounds simple enough, but unfortunately many managers fill up their teams' time with other types of work. And in some cases they don't even see making improvements as work. They see it instead as an "improvement project," and such projects always take a back seat to new business or innovation projects.
As Mike Orzen, author of the book Lean IT: Enabling and Sustaining Your Lean Transformation, said, "Improving your work is just as important as doing your work." That's especially true these days, when companies are facing an ever-increasing pace of change.
3. Failure to move the focus from features to outcomes
Unfortunately, many managers see getting features out as the prime objective. They don't understand what the true value is, and they don't know about the concept of value creation versus value leakage.
Value creation is what organizations want to achieve, such as growth in revenue, or increased consumer loyalty and satisfaction. Every business wants to optimize value creation.
Value leakage occurs when solutions don't meet needs, opportunities are lost, or you face excessive rework because solutions were not properly tested. Or perhaps you experience unacceptable downtime as a result of defects and incidents in delivered solutions.
One of the key behavior changes that business stakeholders and developers need to achieve is to shift their focus from getting new features out as quickly as possible to considering other factors. This includes prioritizing technical debt (such as maintenance work or known errors), defects (such as incidents), and risks (such as security, privacy, and compliance issues).
Features focus primarily on value creation, but other aspects—such as debt, defects, and risks—usually end up being very visible only when the new features go live. This is where value leakage occurs, and that value leakage can damage strategic goals.
And who gets the blame in that case? IT operations. The IT Ops manager gets service level agreements (SLAs) and targets to ensure stability, availability, and reliability, while the developers get key performance indicators (KPIs) for the fast deployment of new features. Meanwhile, business stakeholders are saying, "I want my features, and I want them now, and I don't care about technical debt."
Changing the mindset
This is one of the reasons the DevOps Agile Skills Association (DASA) developed its DASA DevOps Product Owner courseware: to help shift mindset and behaviors to ensure that the product owner is the sole person responsible for managing the product backlog.
IT has had a difficult time convincing the business to prioritize these other types of work. Why? Because it traditionally has done a poor job when it comes to understanding and demonstrating the business impact of that work.
To address this, consider sending your IT employees into the business for one day so they can see for themselves how stakeholders are using IT products and services and the impact—positive or negative—that they have on the business. Your priorities must be aligned with the business value of IT if you are to effectively allocate scarce resources.
It's about changing the behavior of the business
The key reasons for DevOps failure are all people-related; there’s nothing new about that. But it isn't just the behavior of dev and ops that's at issue here; it's the behavior of the business as well. To ensure DevOps success you must focus not just on process, such as DevOps practices, or a product, such as DevOps continuous integration and deployment software, but on the attitudes, behavior, and culture of your people.
Ultimately, it's all about end-to-end collaboration, continual learning and improvement, and credibility. You must demonstrate value, and this can happen only when the entire business is committed to applying the recommendations above.
Want to know more about how to change mindsets and behaviors and create truly collaborative, agile capabilities that deliver business value and protect the business from damage? Come to my session, "DevOps is going to fail … unless," at the Pink19 conference, which runs February 17-20. My session takes place on February 18 at 10 am.
Keep learning
Take a deep dive into the state of quality with TechBeacon's Guide. Plus: Download the free World Quality Report 2022-23.
Put performance engineering into practice with these top 10 performance engineering techniques that work.
Find to tools you need with TechBeacon's Buyer's Guide for Selecting Software Test Automation Tools.
Discover best practices for reducing software defects with TechBeacon's Guide.
- Take your testing career to the next level. TechBeacon's Careers Topic Center provides expert advice to prepare you for your next move.
