Why Do So Many Projects Fail?

The numbers

Research from the Standish Group, which studied over 50,000 projects around the world, shows a clear picture: most projects do not go as planned.

66%

of technology projects end in partial or total failure.

Source: Standish Group CHAOS 2020

45%

Large IT projects run 45% over budget and deliver 56% less value than planned.

Source: McKinsey & University of Oxford, 2012

88%

of business transformations fail to achieve what they set out to do.

Source: Bain & Company, 2024

Failure rates across project types

Why do projects fail?

The same causes come up again and again across all studies:

• Poor planning, 39% of failures start here (PMI, 2024)

• Weak communication, linked to 57% of all failures (ZipDo, 2023)

• No clear goals, teams without a defined success measure are far more likely to fail (PMI, 2025)

• Projects that are too large, large projects succeed less than 10% of the time (Standish, 2020)

What actually helps

PMI's 2025 Project Success Report surveyed 2,841 project professionals worldwide. Three things made the biggest difference:

Define success first

Write down what done looks like before work starts. Teams with a clear goal score 59 points higher on PMI's success scale.

Keep projects small

Small projects succeed 9× more than large ones. Break big work into smaller pieces.

Communicate often

Most failures involve poor communication. Regular check-ins and shared updates prevent most problems.

Key references

• Standish Group. (2020). CHAOS 2020: Beyond Infinity. [50,000+ projects]

• McKinsey & Oxford. (2012). Delivering large-scale IT projects on time, on budget, and on value. [5,400+ IT projects]

• Bain & Company. (2024). Business transformation research.

• BCG. (2020). Most Large-Scale Tech Programs Fail.

• RAND Corporation. (2024). Root Causes of Failure for AI Projects. [65 data scientists interviewed]

• PMI. (2025). Step Up: Redefining the Path to Project Success. [2,841 professionals globally]

• ZipDo / Beta Breakers. (2023). Essential Software Project Failure Statistics.