COBOL, the incessant number cruncher

Its 2024, so how come COBOL is still crunching the numbers for the world economy?
Let’s figure it out…

In March 2024 I gave a presentation to the UK MainframerZ group—a mainframe user forum—on the topic of COBOL. The organizers had asked for a quick-fire “15 facts in 15 minutes” format. I wanted to share some of those key stats about COBOL.

COBOL—by the numbers

So, what are this year’s headline numbers about COBOL? Let’s examine the list I compiled.

65. No prizes for guessing what this stat represents—65 is the years that the COBOL programming language has existed. In September 2024, it will be 65 years since the CODASYL committee scribbled the acronym COBOL on a piece of paper to propose the name of the language they were developing for business use. COBOL was finalised as a proposal, and as a technology, in 1960, but its birth was 1959, 65 years ago.  

3. COBOL materialised thanks to inspired thinking by 3 pioneering females. Grace Hopper gets a lot of the credit. We must also thank Mary K. Hawes, who lobbied heavily for a business computing language, and then invited Hopper to influence what was to become COBOL. The implementation of the technology was then the brainchild of Jean Sammet. Modern IT owes a huge debt of gratitude to what was one of the very first women in tech project.

250 / 800. We can’t to count every line of production COBOL code on the planet, we studies have estimated it, based on market samples. Recently, both the Open Mainframe Project and Micro Focus (later OpenText) did, separately. Their estimates were 250 billion and 800 billion respectively. Whichever estimate you choose—or an average—it affirms that COBOL persists. The previous estimate, made by Gartner in the 1990s, was 220 billion lines.

92. A glimpse as to why there’s so much of it might point to this statistic, which is from the same survey, where IT leaders talk about the strategic importance of these applications to their business. They are important, so they keep them, evolving and adding to them. Enduring business value is a sensible reason to keep hold of things.

90. If business applications have strategic importance, organizations should ensure they are running on highly reliable, high-performance hardware, so the value is constantly delivered. Interestingly, 9 out of 10 IBM mainframes are running COBOL, according to a variety of sources. If the application is strategic, you’d want the platform to be, too.

85. This is the percentage of all COBOL code that runs on the mainframe. The most proven, reliable, and powerful computer of all is the server of choice to deploy many of industry’s critical COBOL systems. The mainframe is a larger environment in terms of processing capacity and storage than many others, perhaps another reason why so much of the world’s COBOL runs there. 

58. The percentage of organizations that run COBOL on more than platform. In this digital era, no application is an island, nor any platform. The interconnectedness of all things requires many integrations and connectivity across the IT landscape. Interestingly, studies show that over half of all COBOL shops are using more than one platform for deployment. Many want business critical systems to run COBOL on a mainframe, as we have already mentioned. But some COBOL apps may be hard to justify, as they may be less valuable, especially when capacity is tight. COBOL offers this flexibility due to its portability. 

500. Speaking of portability, COBOL’s original vision and design was to be portable, so it could operate and deploy with the same results—no matter what platform you chose. Since that time, COBOL has been engineered to work on over 500 recognized unique platforms—a number which came from an internal Micro Focus study. COBOL’s portability made it the original write once, run anywhere technology. 

52. Over half of those surveyed expect their COBOL systems to last another 10 years. Quite a telling statistic, because user perspective of the value of core business applications is a real barometric indicator, as is how long they expect the applications to continue. So it looks like those responsible for the applications see COBOL continuing for the foreseeable future.

72. Of course, those COBOL applications are not static. They do need to change, and 72 percent of those surveyed think that as and when they need to change those core COBOL applications, they think that a ‘modernization’ approach (namely, evolving what’s already there, rather than throwing away) is their best option. 

43. And asked why those changes are necessary, the most popular answer was the trend towards embracing cloud computing. Hybrid models are creating need for COBOL to connect or execute on other platforms, and the cloud—at 43 percent—is the most popular driver for modernization, according to the surveys.

1. Making those changes needs a ready supply of professionals of course, which means ensuring the education and training is available. I was interested to read there’s a learn COBOL in one day free course available online. COBOL is famously very easy to understand, making learning the language comparatively easy. 

16. Of course 1 day isn’t enough time to become proficient in coding COBOL. It is just the start. So elsewhere, one example of an internal training course for COBOL, for professionals, is 16 weeks long, at a large banking organization in Europe. They, among many others have established a robust training program for COBOL because of the importance of the technology to their organization. Other companies contract training agencies to support them, while a small number of universities also teach it. A one-day course rather underscores that point.

140. To get a sense of the demand for online training, I visited Udemy and searched on COBOL courses. If there’s a need for skills, there will be an online course, I determined. There are 140 available today. It’s a solid indicator that the demand for skills is there.

20. To finish our list, a truly revealing, enduring fact is this: a measurement of the popularity of the language based on internet traffic (rather than anyone’s opinion), as compiled by the TIOBE index (here), COBOL has never left the charts. And in January 2024 was at number 20 of the most popular languages by internet traffic. Everlastingly important, it seems.

We’ve got your number. 

We’re at the end of our little tale of numbers. And the result is clear: COBOL still adds up. For so many who think it is dead or dying, the statistics are a revelation, in its 65^th anniversary year. 

The next question is where is it still in use, and how are those IT systems, of which there are clearly so many, evolving? That’s a wider debate—the question of application modernization—and the undoubted growth in that discipline goes a long way towards explaining COBOL’s ongoing importance. Next time someone questions the value of COBOL, tell them to look at the numbers.