Enterprise IT organizations are struggling to plug their DevOps skills gaps. Kevin Davis explains why standardization is the solution.
I remember the pressures on IT that led to the emergence of DevOps services. Enterprise product and project management had fully adopted Agile/SCRUM methodologies and siloed IT Ops and middleware teams had become the neck of the bottle.
This was back in the early 2010s and enterprises were all buzzing about Lean Startup and DevOps. This was due, in part, to Jez Humble, Eric Ries and other writers at the time leading the change toward Agile methodologies that product management had been using as a framework of operation since 2007.
Enterprises that were experiencing headwinds, rapid growth needs, or disruptions in the market were realizing that their survival and future success depended on being more agile end-to-end. Many started to understand that it was better to be proactive, rather than reactive, in the face of digital disruption
What we were hearing from consultants and advisors was a clear message: if IT Infrastructure teams were going to be as responsive as their businesses needed them to be, they would need to adopt DevOps. Furthermore, if they wanted to adopt DevOps, then they would need to adopt technologies that allowed them to develop infrastructure rather than rack infrastructure.
With the right leadership sponsors (and a willingness to try), some infrastructure teams adopted cloud technologies. This allowed them to provision servers as needed and turn development into a parallel activity – freeing their business from the pain of infrastructure-as-a-ticket and four-week delivery cycles.
Cloud and DevOps complemented each other perfectly, and as the adoption of both has accelerated, the value proposition for traditional infrastructure engineering has declined rapidly.
The rise of the DevOps skills gap
Over the last decade, DevOps adoption has steadily increased. Gartner research suggests that 70% of enterprises will have adoption goals for Infrastructure as Code. A figure that is up from the 20% reported in 2019. This would prompt me to guess that every enterprise IT organization is practicing some form of DevOps in at least a portion of their portfolio.
In my own observations, from both leading organizations and advising them over the last 10 years, I’ve witnessed a significant rise in the demand for DevOps skill sets. This is predominantly from organizations that are invested heavily in delivering maximum value from cloud adoption.
The demand for DevOps skills is now at an all-time high and growing. Even as far back as 2019 ZDNet published an article stating that DevOps are the most in demand skill sets in IT. It also points to the challenge of finding resources that have the skill set.
We have seen a scarcity of resources in the market and that has had the natural impact of driving up the cost of talent and experience. Since 2019 the scarcity of these resources has become more apparent with accelerated digital initiatives triggered by COVID-19 further increasing demand in 2020 and 2021. This pattern looks set to continue.
The long-term solution to this supply/demand problem will not simply be more training, upskilling, and scale of human capital. The solutions will come, increasingly rapidly, in the form of standardization, reducing the value of the skill set much as the standardization of core infrastructure devalued the infrastructure engineer.
Is now the time for DevOps standardization?
There have been many attempts in the past to standardize the technology stacks that make up the majority of the DevOps world today. All have failed to this point.
There are lots of opinions, preferences, and requirements that account for this lack of standardization. Typically IT organizations have run individualized core infrastructure platforms and technology stacks. Even in cloud-born application stacks, there are variations in tools and methods based on teams and organizational decisions/processes, some of which are held onto with an almost religious fervor.
But this time it’s different. This time we have a real shot at standardization, industry-wide, for a large range of our IT value chain. We have a chance because:
- IT operating models and business operating models are becoming aligned in digital adoption
- Standardized core infrastructure platforms in the cloud
- There’s a higher bar of business expectation and lower barriers to adoption
- It is a necessity driven by lack of resources
We’ve seen the Cloud Service Providers standardizing the supporting structures in the form of Database-as-a-Service and Platforms-as-a-Service for things like logging, search, IoT, Active Directory, LDAP, etc. Basically, for anything that causes us enough pain, they will eventually feed us a pain reliever in the form of a standardized platform.
However, now we are seeing full stacks being standardized and offered as either solutions or outright managed services. As more technology stacks become prevalent as fully managed options from the CSPs, more businesses will alter their requirements to enable their businesses to utilize those standardized solutions.
The next phase of standardization will reduce the demand for DevOps skill sets, much in the same way cloud adoption has reduced the need for traditional infrastructure skill sets. I believe the scarcity of DevOps resources will accelerate the adoption of standardized technology stacks provided by hyperscalers. While this may ease resourcing pressures on IT organizations over the next few years, it also means that DevOps engineers will need to continuously evolve to keep up with the changing demands of digital business strategies.