Numbers tell a story that anecdotes can’t. This is the data — organized, sourced, and analyzed so you can use it.
Numbers tell a story that anecdotes can’t. When a single project fails, it’s easy to attribute it to bad luck, a difficult client, or an unusually complex technology stack. When the same failure patterns repeat across thousands of projects, across decades, across every industry and organization size — that’s not bad luck. That’s a systemic problem.
We’ve compiled the most current and authoritative data on IT project failure rates, root causes, and financial impact. These are the statistics every project manager should have command of — both to understand the scale of the challenge they’re managing and to make the case internally for doing things differently.
The Overall Failure Rate
The headline number has been remarkably stable for forty years: approximately 70 percent of IT projects fail to deliver their intended value.
The Standish Group’s CHAOS Report, the longest-running study of software project outcomes, consistently finds that fewer than one in three large IT projects (those over $1 million) is completed on time, on budget, and with the originally specified functionality. The 2020 edition found that only 31 percent of projects were “successful” by those criteria.
For large IT projects — those with budgets exceeding $15 million — McKinsey’s research puts the failure rate significantly higher, with the average large project running 45 percent over budget and 7 percent over schedule while delivering 56 percent less value than predicted. The 2025 MIT Sloan report found a 95 percent failure rate for generative AI pilot projects not showing measurable returns within six months.
The Financial Impact
$97 million lost per $1 billion invested. That’s PMI’s estimate of the average dollars wasted during poor project performance.
$260 billion wasted annually on failed software projects in the United States alone, according to estimates drawing on Bureau of Labor Statistics data combined with project failure rates.
189 percent average budget overrun for projects that fail to stay on budget — meaning a project budgeted at $1 million that fails will typically cost closer to $2.89 million.
31.1 percent of software projects are cancelled before completion — representing total write-offs of investment with zero return.
Root Causes: What the Data Says
Requirements problems are the leading cause. PMI’s Pulse of the Profession found that 74 percent of organizations cite inaccurate requirements or a related issue as the primary cause of failure. An Info-Tech Research Group analysis found that 70 percent of digital transformation failures trace back to requirements issues. When combined, roughly 49 percent of digital transformation project failures are attributable specifically to requirements problems.
75 percent of rework is requirements-driven. Reworking issues that should have been done right the first time consumes almost half of all IT department staff time.
Only 49 percent of organizations have adequate requirements resources. PMI’s research found fewer than half of organizations have the people and processes in place to do requirements management properly — yet 87 percent recognize improvement is needed.
47 percent of unmet project goals are attributable to poor requirements management. Nearly half of all the ways projects fail to meet their goals trace back to requirements. Not technology. Not timelines. Requirements.
Industry-Specific Data
Failure rates aren’t uniform across industries. Healthcare IT has historically had some of the highest-profile failures — the $11 billion NHS National Programme for IT was abandoned in 2011. A KLAS Research study found that only 33 percent of healthcare organizations felt their EHR implementations fully met expectations. Government IT consistently underperforms private sector benchmarks. Manufacturing organizations implementing ERP systems face failure rates some research puts as high as 75 percent, with the average ERP implementation running 178 percent over budget and 230 percent over schedule.
Project Size and Failure Rate
Project size is one of the strongest predictors of failure. Small projects (under $1 million) succeed roughly 60 percent of the time. Medium projects ($1M–$10M) at roughly 52 percent. Large projects ($10M–$100M) at roughly 24 percent. Mega-projects (over $100 million) at roughly 0 to 6 percent, depending on the definition of success. The bigger the project, the more likely it is to fail either in whole or in part.
This isn’t primarily about complexity, though complexity plays a role. It’s about requirements. Large projects have more stakeholders, more organizational units, more people whose needs need to be surfaced, integrated, and kept current — and more opportunities for the requirements process to fail.
What Organizational Maturity Data Shows
Organizations at the highest levels of project management maturity — those with formal, repeatable requirements processes, strong executive sponsorship, and active portfolio management — achieve success rates significantly above the industry average. The top quartile achieves success rates above 70 percent, against an industry average of 35 percent.
The gap between the best and worst performing organizations is not explained by technology, budget, or industry. It’s explained by process maturity — specifically, the rigor of how requirements are defined, managed, and communicated across the business-IT boundary.
→ We share these numbers not to demoralize, but because the data points to something important: IT project failure is not inevitable. The organizations achieving 70 percent success rates have built processes that address the actual root causes of failure — starting with requirements.
Explore the course → bridgingbusinessit.com/the-course/
