Technological Debt Kills Flexibility

S. Schuchart

Summary Bullets:

• Systems that have incurred excessive technological debt are brittle, especially when confronted with change.

• Long delays in updates have down the road costs that should not be underplayed.

The concept of technological debt is one that was originally for software development. But the reality is that technological debt can be had across a functional system as a whole. Old servers, old storage, old networking, old security, all of it can incur technological debt. Technological debt is a hard subject, with some dismissing the idea out of hand or downplaying the difficulties caused by excessive amounts of technological debt.

There is a good example of recent real-world technological debt causing and continuing to cause significant problems. In the United States, Congress authorized additional money, $600, to be added to unemployment checks at the beginning of the COVID-19 pandemic. States each run their own unemployment benefit systems, the U.S. federal government supplied the money to the states, and the states used their existing unemployment benefit systems to distribute it. Or at least that is what was supposed to happen.

The reality was that with the mass unemployment of the COVID-19 pandemic, many state unemployment systems quickly became overloaded. With relatively stable unemployment for the last decade, many state governments simply did not update their unemployment systems. As a result, the deployment of the extra federal money didn’t happen anywhere near as quickly as it should have. As a matter of fact, getting unemployment benefits at all became a frustrating and time consuming activity that saw people literally camping overnight outside unemployment offices. Worse yet, because of the sheer age of the systems, additional capacity was difficult to add, and updating to an improved system would require that all of that technological debt be paid. The technological debt means that no new system was forthcoming in any reasonable amount of time because the code base and hardware required considerable work, including possible re-architecting.

It’s easy to draw a parallel of what happened to some U.S. state unemployment systems to some applications that businesses are running today. Old, unmaintained systems become break points when business stress occurs. The agility that is one of the hallmarks of digitization simply cannot happen if critical systems are too far out of date. It’s important to understand that the older the system is, the more brittle it becomes. It’s popular, especially if a company is under economic duress, to say that if a system works there is no need to change it. That only works when there is no other change and as 2020 is teaching everyone, change comes whether you wish it or not.