Famed risk analyst and systems thinker Robert N. Charette has been trying to address the chronic problem of software failures for two decades now. His ideas have garnered tremendous national attention to this key issue. In his extensive career spanning 50 years, Charette has observed a troubling trend: IT professionals, government officials, and corporate executives often fall prey to delusional thinking that leads to repeated, costly mistakes in software development.
Charette’s insights could not be more relevant today. Below, he provides additional reflections on the themes from his original 2005 article, “Why Software Fails,” that continue to resonate today. He contends that no matter how far we’ve come in technology and management practices, the same mistakes still haunt the industry. Yet the biggest tragedy is that this software failure for the most part was predictable and avoidable,” he said. So few organizations take advantage of the urgency of preventing failure. This shortsighted mindset not only threatens their own success, it may eventually result in their demise.
The recent Canadian government Phoenix payroll system fiasco makes this perennial problem clear. For nine years, it has been side-tracked by sadly recurring issues. Charette uses this example to illustrate the costly impacts of software failures. These failures can expose thousands of your employees to dire consequences associated with inaccurate payroll systems.
Beyond discussing the federal government’s own projects, Charette addresses how common software problems are in medical devices. At the same time, he warns, the U.S. Food and Drug Administration (FDA) recalls an average of 20 medical devices a month. All three of these recalls occur due to software malfunctions. If you’re developing software associated with medical devices, there are dozens of other standards you need to follow. Second, there is a greater fear of the backlash if you fail,” he continued. It poses serious questions about the public safety ramifications and extreme lack of dedication to software standards in important industries.
Up second is Charette’s broader concern about society’s willingness to accept the deplorable state of software failure—akin to a zero tolerance for losing access to electricity. “Software is as significant as electricity. We would never accept the power going out every two days. We don’t appear to mind when AWS, or the telcos, or the banks go down. This analogy deserves to shine a light on the desperate need for a PR campaign to make people understand how software reliability has an impact on real life.
Throughout his career, Charette has seen just huge software failures. In his view, these flops were preventable through improved foresight and risk mitigation. Today he calls on organizations to take a more preemptive stance toward determining where their software development practices might freaking suck. These failures continue to recur. They reveal a systemic, industry-wide problem in which regurgitating lessons learned from past failures stops making any significant changes to practices.
Though little time has passed, the technology is vastly different today. Charette’s lessons are a timely reminder that accountability and foresight are essential in the world of software development. He challenges IT practitioners and leaders to change their perspective. To fix this, they need to do more to prevent software failures rather than overreacting to them after the fact.
