My colleague Paul Krill does a great job rounding up the highest-paid cloud computing certifications you can obtain these days. In other words, by passing these courses, you’ll likely have no problem finding a well-paying job.

While there are a few non-brand-related certifications on his list, most are specific to a cloud provider, focusing on particular security services, database services, devops services, etc. I get why this is a thing. Those seeking cloud talent are given marching orders from IT leaders with very narrow requirements, usually certain cloud computing capabilities, so those are the certifications they seek.

The problem with staying too cloud-specific

First, this is a lazy way to look for people. Suppose the value of somebody is defined by the certification exams they can pass, representing a subset of cloud computing technology skills. Candidates get pretty good at aligning to those demands, meaning they obtain the certificates on Krill’s list.

You’ll get a group of people who may know much about a specific cloud brand and technology subsystem but cannot expand from that role. Versatility is the one fundamental skill that people need to have as we continue to rethink cloud solutions. We need an evolving array of skills, not one-trick ponies.

It drives me nuts when we’re looking at better ways to do something, such as using another cloud brand to find better technology, but then the company won’t move in that direction because they’ve spent millions of dollars hiring people with very narrow certifications. They become limited by that investment. And if the people hired with specific credentials are unwilling to be retrained? Rock, meet hard place.

For instance, we need to move some processes from Cloud A to Cloud B, which is better at building and deploying cloud-native solutions. Although this is a better-optimized architecture and will save millions of dollars a month, the fact that the company is entrenched with Cloud A-certified staffers means that moving to Cloud B, no matter how compelling, won’t happen.

What is more, you’re criticized for even suggesting Cloud B as an option—in some cases, causing a morale problem with those who have Cloud A certifications and believe they are being left behind. The reality is that so much has been invested into Cloud A that the company is unwilling to look at other solutions. That will lead to grossly underoptimized solutions that drain value from the business. I am not sure we’ve learned our lesson yet on that one.

We’re walking into a cloud trap

My complaining about this won’t change much. Billions of dollars drive the marketing that the hyperscalers are producing on brand awareness and certification programs. They have convinced cloud users that their certifications are required to make their stuff work, and for the most part, that’s true.

Moreover, those taking the courses are aware of the demand created by the major cloud providers, and thus, they take the courses to meet the demands in the job marketplace. We’ve seen this formula play out for the past 25 years, such as network brand certifications, database brand certifications, and now cloud and cloud service certifications. It’s an old problem, but the adverse outcomes will likely be more impactful now.

The focus needs to be on general skills, such as architecture, database design, devops, security, governance, etc., with a secondary focus on specific skills. Specific skills will likely change over time; thus, staffers must also adjust. An excellent general foundation makes it easy to map those foundations to different technologies. Today, only some people can do that.

For change to occur, it needs to come from the consumers of the cloud technologies. IT needs to refocus on the fundamentals regarding training and hiring. This does not mean ignoring candidates with specific certifications; you need to ensure they have a more rounded skill set. In their interview, ask what they would think if you moved some processes and data from Cloud A to Cloud B. The response should not be confusion.

I can’t impart strongly enough how important this is in changing how cloud systems are selected, configured, built, and deployed. Right now, we’re heading for walled gardens that are going to suck value directly from the business. Worse, most companies won’t understand it’s happening until it’s too late.