IT projects are likely to exceed their cost and time estimates

Standish says that only 32% of projects deliver what they said on time and budget

The most often quoted study into overruns and their causes is the Standish Group’s “CHAOS Report” which was first published in 1995. It was based on replies from 365 respondents, covering 8,380 projects worldwide. Since then they have been running the survey every year. They say that they now have over 50,000 projects in their database.

The validity of the Standish reports has been disputed in recent years. Detractors argue that classifying something as an overrun is meaningless if the original estimates of cost, time and functionality were poor in the first place. This is, however, scant consolation for business people who use such estimates to justify projects in their business cases.

British Computer Society says that one in five projects overruns by 70%

But Standish are not the only ones to have found such miserable performance. In 2008 The British Computer Society published the results of research that looked a 214 projects, covering the period 1998-2005, across a range of sectors within the European Union. The findings are summarised below.

A staggering 23% of projects overran their budget by 70% or more. Add in the 24% of projects that were cancelled after significant money was spent and pretty close to half of the projects went very badly wrong. By the way, all of the projects had schedule overruns commensurate with the budget overrun.

Oxford University Business School says that 17% of projects have a cost overrun of 200%

This bleak assessment is further supported by the most recent research. A huge study of 1,471 projects worldwide, was carried out by Oxford University’s Said Business School. The results were published in Harvard Business Review in September 2011 and stated that:

the average project budget overrun was 27%,
but around 17% of projects had a cost overrun of 200% and a schedule overrun of 70%.

However one interprets the figures from these three studies, it is clear that IT projects often exceed their estimates. What is more, a significant proportion go very badly wrong. Here are three, of many, examples:

The Federal Bureau of Investigation’s “Virtual Case File” Project that was written off, after spending $170 million in 2005
British retailer Sainsbury’s Warehouse Automation project made it into production in 2003 but was later scrapped in 2005, with a reported write-off of £290 million
The CEO of Anglo-French Clearing House LCH.Clearnet departed when his company wrote off €47.8m, after scrapping a failed three-year project to build an integrated clearing platform

People are always over-optimistic about project success

We like to think that we are realistic about success. But actually, we are hard-wired towards optimism. The latest research from neuroscience and psychology shows that we all have bias towards unrealistic optimism, even when counter-evidence is present.

IT projects are complex and difficult to estimate. This moved the academic, Douglas Hofstadter to formulate what has become known as:

Hofstadter's Law

It always takes longer than you expect, even when you take into account Hofstadter's Law.