Robert N. Charette, a veteran risk analyst and systems architect who has spent the last twenty years studying software failure. Charette has dedicated the past 50 years of his career to this work. Throughout that time, he’s recognized the profound effect of delusional thinking on the minds of IT professionals, government officials and corporate executives. His observations shine a light on how preventable and downright predictable software failures can be. This importance is only amplified as organizations face unprecedented challenges from all sides in every sector.
Charette’s critique of software failures is far from academic. Still, he has written widely on these failures, most notably in his seminal 2005 book, Why Software Fails. Photo by Ward Villa on Unsplash He argued that software failure is pretty much foreseeable and preventable. Yet surprisingly, most nonprofits do not put prevention first. This lack of oversight can pose a significant danger. The Phoenix paycheck fiasco of the Canadian federal government is a case in point. It crashed out of the gate and continues to crash nine years later.
The consequences of software failure reach far beyond government systems. According to Charette, the U.S. Food and Drug Administration (FDA) recalls about 20 medical devices per month. These recalls occur because the software is faulty. Beyond the immediate management failures of these recalls, they underscore the existential threat of software complexity where it matters most, like in healthcare. Charette states, “When you’re building software for medical devices, there are a lot more standards that have to be met and a lot more concern about the consequences of failure.”
Downtime incidents also plague major industry players. Amazon Web Services (AWS), banks, and telecommunications companies have all faced significant outages that affect their operations and customer trust. Charette draws a striking parallel between software and electricity, noting, “Software is as significant as electricity. We would never put up with electricity going out every other day, but we sure as hell have no problem accepting AWS going down or telcos or [banks] going out.”
Charette is Special Projects Editor for IEEE Spectrum. He’s spent his career doing that same work for other urgent priorities. His work exposes just how desperately we need to understand software failures. These failures are not just an inconvenience, they have the potential to cause catastrophic impacts to organizations and the public at large.
“The biggest tragedy is that [software failure] is for the most part predictable and avoidable,” Charette emphasizes. “Unfortunately, most organizations don’t see preventing failure as an urgent matter, even though that view risks harming the organization and maybe even destroying it.”
Charette’s observations are more than just noteworthy quotes, though they’re a clarion call for any organization, no matter the sector. The ongoing damages caused by nonfunctioning software requires both an interrogation of incentives and an acknowledgment of the need to institute more reliable safeguards. By recognizing the patterns that lead to failure and addressing them proactively, organizations can mitigate risks and enhance their operational resilience.