The advent of the application economy has made software arguably one the most critical and valuable assets of organizations across a broad spectrum of industries—financial, telecommunications, entertainment, and government, just to name a few. In the era of “app store ratings” and an unprecedented ability of the consumer to easily switch suppliers of whatever they’re consuming, delivering application feature updates as quickly and as safely as possible while maintaining high quality is absolutely critical.
Most of us with a pulse in the IT world are aware that the concepts and practices we know as “DevOps” and “continuous delivery” continue to dominate the software delivery conversation in IT organizations everywhere. Almost every business my company works with has some sort of continuous delivery or DevOps initiative in flight (with, naturally, varying degrees of maturity and success). And yet, we frequently encounter organizations where these ideas are anathema to many within, are considered overhyped buzzwords from consultants and vendors or, most frequently, simply do not apply to their organization in meaningful way. Others maintain that they have tried and failed so it’s back to business as usual. Many are swayed by a number of DevOps “myths” that often stop initiatives before they even get started. So, what is the objection over and some myths about DevOps that keep some organizations from taking the steps to improved software delivery? We’ll take a look at a few that we have seen firsthand and, in subsequent posts, suggest some ideas and strategies for getting past them.
Objection: “DevOps may work for some greenfield apps or startup but that’s not us.”
Many of our most reluctant customers acknowledge the benefits of DevOps and continuous delivery but maintain that their systems are too old, too monolithic, mainframe-dependent, etc., to reap any tangible benefits from a transformation. In other words, “it doesn’t apply here.” This point of view is especially frustrating because the benefits to organizations like these can be just as dramatic and transformative to the culture and IT “psyche” as found among the “unicorns.” Maybe even more so. As the 2015 “State of DevOps” study stated, “It doesn’t matter if your apps are greenfield, brownfield or legacy—as long as they are architected with testability and deployability in mind, high performance is achievable.”
Sometimes these groups will look to hybrid strategies such as bimodal where greenfield projects, innovation labs and other internal IT incubators are chartered to use new technologies and methodologies such as DevOps while the rest continue on as always. Opinions differ here, but these strategies generally are fine as a “way station” within an overall transformation road map with the potential to impede progress if accepted as the status quo. If legacy applications still change with some frequency and if deployments are painful for those involved, then those, too, should be included in the DevOps road map. We’ve done it at my company, a 40-year-old software company, and our mainframe team led the way.
Objection: “We are doing fine as we are.”
Another common category we encounter is the sense that, after some measure of improvement, perhaps the successful adoption of a toolset, that we’ve done enough. Things are better than before, our deployment pain has diminished so we have accomplished what we set out to do. A corollary to that is “we’ve implemented [Jenkins/Chef/Puppet/Cloud/etc.]” implying a belief that implementing one or more of these tools is, in effect, “doing” DevOps.
Successful DevOps is not about adopting a tool or even making measurable progress in organizational goals. DevOps is actually about embracing a culture of continuous improvement, collaboration and, perhaps most importantly, focusing first on maximizing value for your customers. It’s about adopting patterns such as the Deming cycle—Plan, Do, Study, Act—as a core tenet of the way you approach how you think about what you do. No matter what we do, no matter how well we do what we do, there’s always an opportunity to do better.
Objection: “Implementing a major change like continuous delivery is simply too large and complex an undertaking.”
When large, well-established organizations seek to undertake a major change, it’s often a task so daunting, so frequently met with skepticism and cynicism across the spectrum of stakeholders, that the initiative, no matter what it is, appears doomed from the start. Couple that with the culture-shift generally required to affect a successful and genuine DevOps transformation, and it’s no wonder that some senior IT leaders either severely limit the scope of these initiatives or, in some rare instances, veto DevOps outright. Regardless, the belief that their teams are too entrenched in current methodologies to make meaningful progress and the seemingly overwhelming weight organizational inertia stops change cold.
Transforming organizations that have functioned in, more or less, the same fashion for years, even decades in some cases, is practically incomprehensible for many. Organizational cultures generally evolve over time and changing them is not something that can be accomplished by decree and certainly not overnight. The barrier to even starting is often the sense that “we have too much work to do right now and balancing that against some intangible future state is not something we can take on.” Of course, the problem with that line of thought is that there will always be too much work to do now.
And, of course, there are many other obstacles. Barriers to entry against meaningful transformation like DevOps frequently emerge when new ideas and thinking collide with the status quo, especially pre-existing processes governing the software development and deployment. These processes generally emerged from the best practices of the recent past coupled with reactions to major problems encountered years ago that linger on as faint echoes of bygone errors. Most of these rules were put in place with good intentions, applicable at the time of implementation but may have outlived their relevance. Others may be regulatory and not optional. Regardless, if there ever was an area where DevOps and continuous delivery can deliver benefits, this is it.
Many large “traditional” IT shops have, indeed, transformed their culture and fostered environments that continuously improve. So, what do you do? Perhaps the most important step is to decide to take the first step. First, understand why you are seeking to change in the first place. What is the ultimate driver and does everyone, top to bottom, the organization have a clear understanding of that “why”?
Recognize that transformation is an ongoing process where successes and failures are to be expected in equal measure. Learn from both. Don’t attempt to roll DevOps out as a big bang strategy. Work incrementally.
In upcoming posts we’ll explore some strategies for moving past objection and making real strides toward meaningful and lasting change. What are some of the challenges you have faced in your company’s transformational journey? How did your teams get past the obstacles that inevitably appear?
— Jeff Bradley
