The legacy IT skills shortage in mainframe systems
All Blog Posts
Thought Leadership

6 min read

The Legacy IT Skills Issue Is Not Just About People

As valuable as core IT business systems are, they are only as viable as the available skilled experts to maintain and run them. The IT skills challenge remains a nagging question in the industry, especially in more traditional technologies, and the jury is out on whether the talent supply is meeting demand.

The impact of enduring success

Is there a legacy mainframe skills problem? One commentator thought so. “Seven in 10 businesses are now seriously concerned about the financial and operational impact of the worsening mainframe skills shortage.” Yet the perceived skills gap, skills challenge, or skills crisis—call it what you will—is not a new phenomenon. The quote is, in fact, from a 2011 article.

And we’ve been worrying about IT skills challenges ever since.

As organizations continue to rely on trusted core IT systems—many of them mainframe-based—to run their business, those concerns continue to grow, for two good reasons.

First, those systems were already long-standing "heritage" systems over a decade ago. And the skilled technical staff employed to build and support them were, in many cases, very experienced, older workers. A dozen or more years later, the toning of the retirement bell may have called time on many illustrious careers. Decades of deep, unique technical expertise has left the building.

Second, as many of us may be feeling, to remain worried about an IT skills deficit over a dozen years after it was first flagged, suggests an industrial-scale failure in IT succession planning.

Evidence of a fresh supply of skilled industry talent

But is it really as bad as that? Not everywhere, it would seem. The corporate world hasn’t stood still in many cases. SEB, a major bank in Sweden, is justly proud of its internal four-month-long mainframe COBOL training bootcamp, which produces a regular supply of skilled practitioners to work on their core IT systems.

Outside of corporate programs, third-party training organizations aim to pick up a lot of the slack. Market-leading mainframe training specialists Interskill Learning proudly announced that, in 2023, “Over 1,000,000 hours of IBM Z mainframe training [had been] delivered globally” by them. Technology vendors including Broadcom, IBM, Rocket, and others, as well as the Open Mainframe Project, also offer a range of training options.

Meanwhile, in academia, there are terrific examples of mainframe and COBOL educational initiatives. Marist College, New York, enjoys a fabled partnership with the nearby IBM Poughkeepsie facility, while elsewhere, pioneering HBCU professor, Cameron Seay, recently confirmed that his mainframe COBOL training classes are “full each term."

Demanding times

But is it enough? IBM’s buoyant mainframe revenues and BMC’s latest mainframe market survey results both claim “significant momentum," claiming increasing mainframe workload, which implies a continued increase in demand for skilled technicians to run those systems into the future.

IBM’s newly unveiled skills program, the Mainframe Skills Council, is an initiative to help boost the numbers entering the space. The program somewhat resembles similar announcements from the past, suggesting that there is unfinished business in resolving the skills challenge. Commentators agree. "IBM has had formal programs to develop the next generation of mainframe talent going back to at least 2003," remarked Gartner managing VP Mike Chuba. "This Skills Council looks to add one more arrow to the quiver of initiatives IBM has taken … [but] I don't know that we are at the point where we can say there is an adequate amount of talent coming in to completely offset those folks exiting the workforce."

It is hard to ignore a statistic from a Forrester Research study, which states, “79% cited their top mainframe-related challenge is acquiring the right resources and skills to get work done.” There’s little to suggest that the supply has demonstrably caught up with the demand.

Thinking smarter

IT leaders are turning more and more to technology itself to help accelerate skills acquisition and core modernization activities. But as we’ve already said, that’s a lot of expertise in some highly sophisticated, technical areas. It’s not just a case of teaching a bit of COBOL or knowing your way around TSO and ISPF; this is about inferring the systemic, institutional business rules buried deep and encapsulated in decades-old applications and data structures.

How do you accelerate understanding of that? Even a good technical course in one tool or another won’t make you an expert in a corporation’s unique application set. It requires a forensic appreciation of how applications work to support the business. The fix to a people question is, therefore, not just about people—it involves innovative technology.

The question becomes: How can technology supplement and offer important capabilities to help organizations plug their modernization skills gaps—and replace lost subject matter expertise?

Gap analysis

Where are the gaps? Let’s look at some typical requirements of large-scale modernization programs.

Application Technical Onboarding. Big change programs are so often the catalyst for hiring new staff or contractors, who might have skills but no institutional knowledge. Accelerating their understanding of incumbent architectures, naming conventions, coding practices, application composition and behavior and key technologies could take months if not years. Using technology to automatically visualize, document, and decipher complex technological relationships and dependencies is a huge part of the onramp they face.

Modernization Assessment – Code and Data. Another vital preliminary modernization task is to "take stock" of what system(s) are being considered for modernization. What they comprise will have a huge bearing on how best to achieve the aims. Gone are the days of manually reviewing PDS or SCM repository contents, paging down through miles of COBOL or JCL; the assessments must be rapid, repeatable, complete, and verifiable. This requires technology that understands the current environment fully, suggests actions, and can report on it automatically and succinctly. In other words, we need to ask more of our tech.

Application Understanding Reams of Code. Organizations looking to modernize often preside over tens of millions of lines of code. Assessing just one million lines is equivalent to reading the classic novel War and Peace 14 times over. Imagine trying to determine how it all fits together, let alone pinpointing areas of risk or concern for future changes. Even the seasoned experts don’t know it all. Understanding applications at the code or data layer requires insight that only technology can assist with. Sophisticated technology helps accelerate the learning required to truly understand core applications in detail.

Given this, it would be folly to consider a manual or semi-manual process to undertake more complex modernization tasks that rely on robust application understanding. In each of the following cases, without complex technological support, the tasks become all but impossible, however many people you assign to it.

  • Business Rule Identification and Extraction – Business rules are never simply defined and structured in a convenient, modular way. They must be derived through a nuanced understanding both of the application and the business it supports.  
  • Augmented Application Change – Any large-scale change program for an application, for example—and what is sometimes called "Reengineering or Accelerated Rewrite"—or any similarly major plumbing works, require a clear and full understanding of the "current" situation before a future state can be planned.
  • Technical Debt and Redundancy Eradication – In larger organizations, code redundancy can comprise a worryingly high percentage of application code. Technical debt is usually a mountain of as-yet-unplanned work. Finding efficiencies in identifying and planning such activities is crucial to start to catch up on what are often vital modernization-related tasks.  

Of course, we’re scratching the surface here, and there are other use cases like mainframe core tooling and administration—which is where Interskill excels—where the gap doesn’t exist in the same way. A lot depends on the situation, the organization, and its technical roadmap.

Tool up to skill up

Even the most experienced surgeon wouldn’t attempt a complex operation without adequate preparation, examination, diagnosis, planning, and the right equipment for the procedure. Intricate surgeries involving vital organs is a fair analogy for the complex process of modernizing business-critical applications. To provide the next generation of skilled ‘IT surgeons’, we must train them well, and give them access to the most efficient and effective tools for the job. How AI may support that aim, we will cover in another article soon.

🎥 Interested in learning how the AveriSource Platform™️ addresses the legacy mainframe skills gap? Check out this video interview with The Futurum Group's Steven Dickens and AveriSource's Ed Airey.

🗓️ Curious to see the AveriSource Platform's Accelerated Rewrite™️ capabilities live in action? Schedule a demo with the AveriSource team of mainframe modernization experts.

Learn more aboutDerek Britton

Industry Thought Leader

With over 30 years in the enterprise software industry—all of it in the application modernization arena—Derek is an accomplished technology marketing leader, writer, and presenter. With software development, marketing, sales enablement, and services experience, Derek regularly commentates across the IT press, and at events such as Gartner, Open Mainframe Project, SHARE, and GSE. Derek holds a Bachelor of Science degree in Computer Science from De Montfort University.

Learn More about author