McKinsey-Oxford 2012: Large IT Projects Run 45% Over Budget
The most-cited number in IT project performance. We give the source URL, sample size, definition of "large", and the full headline findings so you can quote it accurately.
What "large" means in this study
The McKinsey-Oxford study defined a large IT project as one with an initial budget greater than 15 million US dollars. Projects came from public and private sectors, across geographies, drawn from a McKinsey project database supplemented by Oxford Said Business School data.
The study explicitly excluded small in-house IT work (anything sub-15 million) because those projects have very different governance and risk profiles. Large IT projects, by contrast, tend to have multiple vendors, long timelines, regulatory complexity, and political stakes: the exact conditions that produce overruns.
The full headline findings
| Finding | Value |
|---|---|
| Average cost overrun | 45% |
| Average schedule overrun | 7% |
| Average value shortfall vs predicted | 56% less |
| Share of software projects more than 50% over budget | 33% |
| Share of black swans (overrun above 200%) | 17% |
The black-swan finding (17% of large IT projects overrun by more than 200%) is the most concerning number in the dataset. It establishes that large IT projects do not just slip: a meaningful minority become outright disasters that can threaten the parent business.
How to cite
Source URL: mckinsey.com (delivering large-scale IT projects)