At a NATO conference in 1968, the term “software crisis” was first articulated. This phrase really resonated with all the growing pains that have been experienced in software development. In the intervening years, technology has leaped forward and global IT spending has skyrocketed. Software project failures continue to plague organizations such as governments and private enterprises across the globe. High-profile disasters such as Canada’s Phoenix payroll system serve as reminders of the perils that still plague the industry. These cases pose important questions regarding just how effective our current strategies are at managing complexity and risk.
The Phoenix project, which went live in April 2016, was supposed to do just that—improve the payroll system for all Canadian federal employees. In practice, it completely blew up. These problems led to more than 349,000 open pay errors — affecting about 70% of the 430,000 current and former employees in the system. This failure was not a unique episode. It was the second time they had tried to replace an antiquated payroll system, after a previous unsuccessful effort to do so in 1995. The project’s ambitious goal involved customizing PeopleSoft’s off-the-shelf payroll package to accommodate 80,000 pay rules across 105 collective agreements with federal public-service unions.
The Cost of IT Failures
A report by NTT DATA released in 2024 revealed that 80% of organizations believe inadequate or outdated technology hinders their innovation efforts. For the vast majority, the financial stakes of unsuccessful IT projects are mind-numbing. Bent Flyvbjerg is professor emeritus at the Saïd Business School at the University of Oxford. He stressed that IT projects have astounding levels of baseline risk. He argued that these projects typically have the greatest monetary upside. Even so, they continue to draw enormous amounts of funding.
Global IT spending has skyrocketed from $1.7 trillion to $5.6 trillion in constant 2025 dollars since 2005. Success rate for software projects have hardly improved in the last 20 years. As author and engineer Henry Petroski lamented, the IT community has certainly learned how to compute software disasters due to risk. Even so, it fails to break through what he dubs “the failure of the mind.” To keep pace, organizations are learning to use more advanced risk management techniques. Despite these improvements, they continue to ignore basic problems in the way projects are planned and delivered.
“To Engineer Is Human: The Role of Failure in Successful Design.” – Henry Petroski
Human Imagination and Complexity
Most IT disasters result from failure of human imagination. These major misses are often a result of unattainable project objectives, challenges in the management of complexity and not proactively managing risks. The Canadian government’s MiDAS unemployment system and Australia’s Centrelink “Robodebt” welfare systems serve as additional examples of early automated decision-making failures that have caused significant distress and financial strain on affected individuals.
Yet the long-term impacts of these failures aren’t just found on budget spreadsheets. We can’t forget about thousands of employees and their families who have experienced tremendous emotional trauma and distress. They experience many of the same problems faced during the Phoenix build. The psychological effects on survivors can be devastating. This not only leads to greater employee disengagement, but erodes the public’s trust in government processes.
“Anyone can make a mistake, but only an idiot persists in his error.” – Cicero
Navigating Future Challenges
IT leaders and organizations are under tremendous pressure to reinvent the way they build software. The inevitability of these software failures points to a deeper, systemic problem outside the bounds of tech fatigue. Despite the industry’s focus on innovation, most IT failures today do not stem from attempts to push the boundaries of computing art but rather from struggles at the mundane edges of project execution.
To move forward and enact long needed change, organizations need to tackle these fundamental issues head on. To counteract the dangers of writing code in the dark, prioritize good project management practices. Promote open lines of communication between all stakeholders and establish achievable milestones to help achieve desired outcomes. Beyond that, industry and government need to invest in training and development programs that help enhance the skills of existing IT professionals.
