As COBOL reaches 65 years of loyal service to the IT world, our four-part homage attempts to answer how this veteran IT technology has been successful for so long.

The COBOL programming language – which stands for Common Business Oriented Language – began life in 1959, with its first release a year later.

In the 65 years since, the world of IT has largely overlooked or downplayed its continued success as a technology instrument, and COBOL has—with a couple of notable exceptions—stayed out of the limelight. That lack of attention might have caused other problems, as we will explore, but the fact remains that vast swathes of the global economy still run on COBOL.

By sharing the opinions of those who have built, used, and supported successful COBOL-based IT systems, this four-part article attempts to answer the question – how has this veteran IT technology been so successful for so long?

Cutting-edge COBOL?

Let’s start with the disclaimer. If you came here expecting a fresh announcement about a seismic change, with COBOL now doing all kinds of bleeding-edge innovations, I’m here to disappoint you. That’s not what’s in store.

In fact, if you investigate the Gartner hype cycle for new and innovative technology, you won’t find any mention of COBOL, nor indeed would you if you looked in all the previous reports.

“Fundamentally, COBOL isn’t newsworthy.”

But let’s say that you want a balanced perspective on how COBOL stacks up against other programming languages. First, finding a genuinely dispassionate, impartial point of view is harder than it should be. 

Using the TIOBE index, a monthly indicator of programming language popularity, as a reference point, you would scroll down well into the 20s or even 30s to find a mention of COBOL in the programming language popularity chart. In terms of internet traffic, COBOL sits a distance behind chart-toppers C, Java, and C++.

Why? Fundamentally, COBOL isn’t newsworthy. When it comes to technology choices, COBOL is a venerable, plain, reputationally clunky language. 

Many argue that COBOL belongs to an era past. Move aside, COBOL. This week, we are all about hybrid, containerized large language models (LLMs), and then on to the forefront of generative AI next week. This trend-chasing attitude toward technology affects the talent pool, too; the COBOL programmer community continues declining in number, replaced by newer technologists with fresher skills. All we need is AI prompt experts now, right?

Chasing Dreams

Or do we? Consider that only 10 percent of the Fortune 500 companies from 1955 remained on the same list in 2019. We can all remember when Palm owned the handheld device market, when Apple only made desktops, or when Amazon was just a bookstore.

Examining the latest Gartner Hype Cycle puts Generative AI near the very front, right on the apex of what it calls the Peak of Inflated Expectations – just ahead of the near-vertical drop into the Trough of Disillusionment. 

Such are the expectations of innovative technology that the path to widespread adoption is typically not a straight line and always something of an educated guess for decision-makers. Snazzy new technology, however alluring, never equates to widespread adoption.

“Broken promises litter the pathways of IT history, where expectations often exceed the reality of exciting modern technology.”

Broken promises litter the pathways of IT history, where expectations often exceed the reality of exciting modern technology. Smart CIOs know that you cannot bet the farm on a new trend until you are confident that it will improve upon what you already have.

Back to the TIOBE index again. The team at TIBOE confirmed that COBOL has never left the charts since records began, a feat only matched by C, C++, and, interestingly, LISP. This underscores the point that any incumbent system has measurable business value. As such, any decision to replace it must be considered low risk and higher value for the effort to be worthwhile. The TIOBE index implies an enviable ‘staying power’ for COBOL.

A Modern Mixture

IT landscapes are forever evolving as requirements, strategies, and personnel change. CIOs are the proverbial bartenders, endlessly looking to blend ingredients for the best possible result—but doing so takes real skill. With these tech cocktails, get it wrong, and the headaches last ages.  

For many established institutions, mixology started out several decades ago. It involved mainframe or mid-range computers and languages such as Assembler, COBOL, PL/I, Fortran, ADA, and Algol. Fast forward, and the platform choices extended to more mid-range offerings. Then, a new interloper emerged on the scene called UNIX, which became a standard — until it fragmented into several options, each one driven by a separate manufacturer. The C language cemented its place on the TIOBE index by being the language of choice for UNIX -and elsewhere.

The IT world ambitiously attempted to decentralize across networks in the form of client server, just as the platform, operating system, and language choices expanded until it was hard to keep track of best practices. And this was all still in the twentieth century.

IT leaders had yet to ride the crest of e-commerce, the coming of age of the Internet, the dawn of the mobile device, and the emergence of cloud computing. Sprinkle on Big Data, containerization, and the noisy arrival of AI into the mainstream, and it has been quite a technological evolution.

Multiple Choice Carries Costs

These milestones illustrate the bewildering options available to senior IT leaders. Most organizations approach IT buying based on the need and best available tooling at the time. This ensures that decisions zigzag up the fairway, each one carrying its own value and cost and arriving at different times.

Worse still, some IT leaders made technology investment decisions in isolation from each other. (The horror!) The outcome is a smorgasbord of IT investments (tools, platforms, data, procedures), which seldom integrate cleanly and sometimes are mutually exclusive and incompatible.

A long supplier list and low integration mean higher cost of service, slower outcomes, and greater risk. It also does nothing to help the fight against security risks, staffing costs, or operational simplification. One jaded CIO confided that he felt like he had. ‘One of everything, and a whole lot of nothing.’

Against this backdrop of complexity, COBOL’s chances of retaining a strategic position in IT have been severely tested. Tune in next week to explore how COBOL fared then and now, and why.

Read more on COBOL by Derek Britton