Mainframers Are Getting Younger, and Other Takeaways From BMC’s Annual Survey
Eric Odell, director of solutions marketing at BMC, interprets the results, which tell the story of a generational handoff and all-time high positive views of the platform
That the IBM mainframe is handling a big job shouldn’t come as any revelation. And perhaps it shouldn’t be shocking to see young IT professionals, people in their 20s, embrace the platform. These folks are Generation Z, after all.
Even if these happenings are ultimately unsurprising, it is nonetheless striking to witness the transition that continues apace in the mainframe world. Within the past decade, we were warned about the skills gap, COBOL disappearing from college curricula and the graying of the mainframe workforce, among other seemingly intractable issues. Of course these challenges persist, but a bigger, brighter picture is taking shape: progress continues, steadily and smoothly.
This outlook is reflected in BMC Software’s annual mainframe survey, which the Houston-based provider of enterprise software and IT and consulting services released last fall. Survey results are discussed in this TechChannel webinar (registration required).
Positive Mainframe Sentiments Reach New Heights
The 2025 survey stands as a milestone for BMC. It’s the 20th the company has conducted, and the largest to date in terms of response—more than 1,100 clients shared their views and insights. However, for Eric Odell, director of solutions marketing, a few numbers stand out. Among survey respondents, 97% held a positive view of the mainframe, with 93% saying that their enterprises would continue to invest in the platform.
“That’s very incredible,” he tells TechChannel. “What we’ve seen here is a change where only 3% of respondents to our survey say, ‘We’re considering an exit over the next period of time.’ A few years back, 10% said they were looking for an exit strategy.”
Even with only scant room below the approval ceiling, there’s no sign of the mainframe slowing its roll. This wave of positivity reached yet another level over the past year, considering that the mainframe approval rate stood at an already robust 94% in the 2024 survey.
Odell concurs with the commonly held opinion that IBM innovation, starting with the unveiling of the z16 in 2022, is behind the prevailing enthusiasm among mainframers. He describes the z16 and z17 as “pivotal upgrades because of their capabilities with AI and also what’s now coming out with the Spyre cards, so many different things.”
The Workforce Handover
For so long, there existed a common image of the typical mainframer. This individual would be male, in his late 50s or early 60s, working in systems operations or poring through decades-old COBOL code. The job remained rewarding, but retirement was certainly on his mind.
Again, it’s not simply that things have changed. It’s the suddenness of the changes, coupled with the relative smoothness of the transition. “I think that one of the most striking shifts with this year’s survey is just how quickly the mainframe workforce is getting younger,” says Odell. “A lot of people didn’t think this would happen. … It’s a complete handover.”
The BMC survey compares the composition of the mainframe workforce between 2018 and 2025. In 2018, there was a near split among mainframe pros. While a thin majority—53%—were younger (aged 18-49), still nearly half—47%—were age 50 or up. So the generational transformation was well underway. However, by 2025, this evolution was nearing completion, with 18-49-year-olds comprising about 80% of the mainframe workforce.
Odell breaks it down further by citing the 18-29 age group, those considered Gen Z. “These numbers for Gen Z may sound low, but think about the jump. In 2018, 1% of respondents were Gen Z. Today, it’s 15%. That’s a big jump,” he says. “Millennials and Gen Z have surged in the talent pool while baby boomers have nearly disappeared. This is the amazing thing.”
Another thing about the youngsters who have arrived on the mainframe: They’re happy to be here, and they’re excited for the future. The survey showed that, as a group, 73% of Gen Z’ers expect the mainframe to continue to grow and attract new workloads. Nearly as many Millennials (69%) share that perception, while the figures are somewhat lower (59% and 51%, respectively) among Gen X and baby boomer (roughly age 50-64) respondents.
The Language Transition and Evolution of AI
“Think about the Millennials or Gen Z’ers,” Odell says. “They’re willing to use the mainframe and they see it as a platform for the future. If you’re taking over the mainframe world, what are the kinds of things that you want to do?”
One thing they’re doing is relying upon the code languages they know. This explains how Java has emerged as the most popular programming language on the mainframe. C++ and Python have also surpassed COBOL usage on the platform. However, none of this is to suggest that COBOL is disappearing, as Odell explains.
“You’ve got people making assessments: ‘Hey, we’re going to have to make changes to this application with this code on a regular basis. So this portion, we need to do this in Java, whereas other portions of code, like your core code, won’t change as often,” he says. “So you can leave that portion in COBOL and the things that need to change quickly for the business, you can do that in Java or Python or whatever is the best for your organization. It’s some very, very interesting twists and turns.”
The Shift Toward Agentic AI
However, for all that’s changed, the mainframe world still proceeds deliberately. This is evident as organizations move forward with AI. Clearly, survey results point to AI development figuring prominently in near-term planning. Nearly three-quarters of respondents identified generative AI as a strategic initiative.
“Mainframe priorities have changed,” Odell says. “Think about big banks like Lloyd’s and Bank of America, some of the big financial services companies like an AmEx. Compliance and security are huge. With those bigger organizations, we’re seeing AIOps and application modernization suddenly become major, major priorities.”
In assessing the evolution of AI use among mainframe enterprises, Odell describes an environment where people are gradually becoming more comfortable with and confident in its capabilities. “AI sounds easy. When you use ChatGPT, doesn’t it sound easy to you? But there’s no easy button with the mainframe. You cannot just jump to where AI does everything and I can just go and sip margaritas out on the beach. You have to put in the work to get there.”
Through a few survey questions, BMC gauged respondents’ willingness to allow AI to handle tasks ranging from issuing alerts to recommending actions to, ultimately, completing actions on its own. Taking a broad view, more than 60% of respondents said they are looking to AI issue alerts and recommend actions. Around 30% would be willing to have AI complete certain tasks on its own.
“I would say right now we have the capability to be able to alert for the need for the action and the ability to recommend for action. I can tell you that we have those capabilities because I know, because I promote those kind of technologies today,” Odell says. “Now, have people rolled that out? A few. But the important thing here is the willingness.”