Global COBOL Growth Spurs Reevaluation of ‘Legacy’ Assets

February 16, 2022 Alex Woodie

COBOL is supposed to be a dead language, tied (mostly) as it is to the IBM mainframe, which is also dead. But according to a recent survey from Micro Focus, the amount of COBOL code in the world actually is expanding. What in COBOL is going on?

Micro Focus commissioned the research group Vanson Bourne to evaluate the state of COBOL use in the world, and the results of that research were released on February 4. Perhaps the most eye-popping figure references the title of the report, “How much COBOL is really out there?”

The answer: Anywhere from 775 to 850 billion lines of COBOL. According to Micro Focus, that far exceeds previous estimates, which have come in around 200 to 300 billion lines.

“800 billion lines of code reinforces the importance, and continued investment, in this most trusted of core business system technologies,” Ed Airey, Micro Focus’s director of COBOL product marketing, said in a press release.

While the general perception of the market is that the amount of COBOL is decreasing as mainframe applications are retired, that doesn’t appear to jibe with the reality that Vanson Bourne detected. The research group found 48 percent of survey respondents expect their COBOL volume to increase over the next 12 months.

What’s more, 52 percent said they expect COBOL applications to be running in their organizations for the next 10 years at least, while more than 80 percent said COBOL will still be in use in their shops when they retire.

The changes in COBOL use are driven by several factors, with “customer requirements” (44 percent) leading the way, followed by “future IT strategy” at 41 percent and “application portfolio alignment with new technologies” at 35 percent.

More than 90 percent of people surveyed in Vanson Bourne’s survey say COBOL applications are strategic to their organizations, which somewhat contradicts the prevailing notion that organizations can’t wait to turn off these systems because they are old and don’t meet needs anymore.

Modernization of existing COBOL code (versus “rip and replace”) is the preferred path forward for 64 percent of survey-takers, while 72 percent say modernization is a legitimate business strategy.

The 800 billion lines of COBOL figure is an estimate, of course. For that reason, Vanson Bourne let survey respondents qualify their responses regarding how much COBOL code was lurking in their shops.

The results of that analysis shows 40 percent were very confident and another 46 percent were somewhat confident in their estimates. Even those who expressed some skepticism in their estimates indicated that they believe their response was within 10 percent (for what that’s worth).

Vanson Bourne — which didn’t disclose the number of people who took the survey but did state that they came from 49 countries and trended toward working in larger shops with at least 1,000 employees — took the responses and weighted them against their relative level of certainty. It then extrapolated the estimated volume of COBOL code using market size estimates of the various installed bases for the most popular COBOL platforms.

What platform companies run COBOL on is another interesting figure from the report. It’s no surprise that the mainframe plays heavily in the COBOL world, and Vanson Bourne’s data shows that 43 percent of COBOL code runs on z/OS. The number two platform? Microsoft Windows, surprisingly, with a 31 percent share.

The long tail of COBOL users dwindles down from there, with IBM z/VSE-VM and IBM’s z Cloud occupying the third and fourth slots, followed by several non-mainframe environments, including Red Hat on Intel, Microsoft Azure, AWS, and IBM i, which came in 8th with a 5 percent share.

So, what does this have to do with IBM i? Well, one side effect of the pandemic appears to be that it has shook up IT departments, particularly in industries that saw a significant rise in demand, such as government, retail, and manufacturing. As the core underlying business systems have been strained, organizations have accelerated their modernization timetables.

Because of similarities in architectures, many organizations that run COBOL on IBM Z and mainframe systems often look to the IBM i server as the next-best place to run. “It’s supported the same on iSeries as it is on mainframe,” says Jerry Thomas, the vice president of migration services at Cothern Computer Systems, which specializes in COBOL migrations to IBM i.

As mainframe shops look to find new homes for their COBOL applications, the IBM i stands to inherit some of them, at least temporarily. Eradani CEO Dan Magid recently noted what seems to be a renewed interest in COBOL activity on IBM i, which seems to be somewhat borne out of the marketplace data collected by HelpSystems.

According to Walter Camp, Cothern’s vice president of business development, many mainframe customers who are looking to rehome COBOL applications will move those apps to IBM i for a few years before deciding what to do next. That buys them some time to decide whether modernization, a full-write, or ditching the homegrown system for a packaged ERP system makes the most sense.

“If you’re converting one of these COBOL programs it will take about a year,” Camp says. “If you’re rewriting, that could take three or four years to develop the same applications and get them running, because there’s a time and a cost. Or if you want to change languages, there’s tooling that will change languages.

“So you just have these multipliers,” he adds. “And cost is time. So instead of being a million dollar project, you end up with $2 million. If you can change languages, you can be up to $4 million before you load everything. It’s a function of what do you want? What’s your business objective? How much can you invest in this? How much time do you have? And most companies don’t have four years to develop new code. They’re better off just getting a package honestly.”

The folks at Micro Focus, which develops some of the most widely used COBOL compilers and tools in the market, clearly see a lot of value still balled up in these aging COBOL applications. The Vanson Borne survey indicates that 70 percent of COBOL users favor modernizing the COBOL applications as opposed to rewriting them or ditching them.

That means COBOL’s will have a strong position for at least the next 10 years, says Chris Livesey, senior vice president of application modernization and connectivity at Micro Focus.

“As we see the attitudes around COBOL modernization with changes to where and how it needs to be delivered and how its usage continues to grow, COBOL’s credentials as a strong digital technology appear to be set for another decade,” Livesey states in a press release. “With 60 years of experience supporting mission-critical applications and business systems, COBOL continues to evolve as a flexible and resilient computer language that will remain relevant and important for businesses around the world.”

RELATED STORIES

Is COBOL on IBM i Experiencing a Renaissance?

As I See It: COBOL In The Time Of COVID

COBOL Is Not The Problem; The Data Is