Flyvbjerg & Budzier 2011: The IT Project Black Swans
The most-cited source for the “average IT project runs 27% over budget” figure, and for the finding that matters more: one in six IT projects is a catastrophic outlier. Here is the sample, the numbers, and the citation.
What the study found
Across 1,471 IT projects, the average cost overrun was 27%. But the average hides the real danger: one in six projects was a Black Swan, with a cost overrun of 200% on average and a schedule overrun of almost 70%. Flyvbjerg and Budzier's point is that IT projects are not unusually risky on average; they produce a disproportionate number of these extreme outliers, and it is the outliers that break budgets and companies.
Source: Flyvbjerg B. & Budzier A. (2011), “Why Your IT Project May Be Riskier Than You Think”, Harvard Business Review, September 2011.
Why the 27% average is the wrong number to plan around
A 27% average cost overrun sounds manageable. You could add a contingency line and move on. That is exactly the trap the study warns against. The distribution of IT project outcomes is not a tidy bell curve around 27%; it has a long, heavy right tail. Most projects land near or a little over budget, and then a minority blow up completely.
Those blow-ups are the Black Swans: one in six projects, overrunning by 200% on cost and almost 70% on schedule on average. A single Black Swan can dwarf the combined overrun of dozens of well-behaved projects, and it is large enough to threaten the organisation running it. The authors' conclusion is blunt: by managing to the average, most managers and consultants have been measuring the wrong risk.
The practical implication is that IT project budgeting should be a tail-risk exercise, not an average-plus-contingency exercise. That is the same logic behind reference class forecasting and Monte Carlo cost-risk modelling: size the budget against the P80 or P95 outcome, not the mean.
What “Black Swan” means here
The term is borrowed from Nassim Nicholas Taleb: a rare, high-impact event that conventional averages and forecasts fail to anticipate. Applied to IT, a Black Swan project is one whose overrun is so far out on the tail that it is a different kind of event, not just a worse-than-usual project. In this dataset the Black Swans averaged a 200% cost overrun, meaning they cost roughly three times their approved budget.
Flyvbjerg later generalised this pattern across every sector in his project database, where IT sits among the worst tails: the 18% of IT projects that go more than 50% over budget overrun by 447% on average. The 2011 IT study is where the fat-tail argument was first made in a management outlet.
Where it sits among the large-sample IT studies
This is one of four credible large-sample sources on IT project overruns. They measure different things, so quoting the right one matters.
| Study | Sample | Headline finding |
|---|---|---|
| Flyvbjerg & Budzier (HBR 2011) | 1,471 IT projects | 27% average cost overrun; 1 in 6 a 200% Black Swan |
| McKinsey-Oxford (2012) | 5,400+ large IT projects | 45% over budget, 7% over time, 56% less value |
| Standish CHAOS (2020) | 50,000+ project records | 31% successful, 50% challenged, 19% failed |
| Flyvbjerg project database (2023) | 16,000+ projects (all sectors) | IT mean overrun 73%; worst 18% overrun 447% |
Sample and methodology
The analysis covers 1,471 IT projects, comparing budget and schedule at approval against actual outcomes. It is the largest sample of its kind published in a mainstream management outlet at the time. Overruns are measured against the original approved budget, which is why they run higher than figures measured against later, re-baselined budgets.
The key methodological contribution is not the average but the shape of the distribution: the authors show the overruns follow a fat-tailed, power-law-like distribution rather than a normal one. That is what makes averages and standard contingency percentages unsafe for IT, and it is the finding worth citing.
How to cite
Primary source: hbr.org
The HBR article is paywalled. A free working-paper version is available on SSRN and arXiv (1304.0265). Cite the HBR article as the primary source.
Frequently Asked Questions
What is the average cost overrun for IT projects?
In Flyvbjerg and Budzier's analysis of 1,471 IT projects (Harvard Business Review, September 2011), the average cost overrun was 27%. But the average understates the risk: one in six projects was a Black Swan, with a cost overrun of 200% on average and a schedule overrun of almost 70%.
What is a Black Swan IT project?
A Black Swan is an IT project whose cost overrun is an extreme outlier. In this study the Black Swans averaged a 200% cost overrun and almost 70% schedule overrun. The term, from Nassim Taleb, describes rare high-impact events that averages hide. The study's point is that IT produces a disproportionate number of these outliers, and they, not the average, are the real risk.
How is this different from the McKinsey-Oxford 2012 IT study?
They are separate studies. Flyvbjerg and Budzier (HBR 2011) analysed 1,471 IT projects and found a 27% average overrun with one in six a 200% Black Swan. The McKinsey-Oxford study (2012) analysed more than 5,400 large IT projects and found 45% over budget, 7% over time, and 56% less value than predicted.