Received wisdom says that complex advanced technology is the prime cause of failure in most large programs, but the seeds of failure are often sown right at the start of a programme in poor decisions made by CEOs and their boards. "There are very, very rarely technology issues in failed programs - you can always track issues back to failure being 'designed-in' at the start of a programme because of decisions the senior management team either ducked or didn't make at the time or were ignorant of," according to David Hilliard, who is often called in to rescue failing programs in his role as CEO at Mentor Europe, a telecoms consultancy firm.

Even the basic task of identifying the really critical programs from the mass of decisions that land on a CEO's desk every day is usually fudged. "It's a fact that companies always want to run an infinite number of programs concurrently so being selective about what you want to do is really important, and segmenting the truly important programs is something that many companies find difficult," Hilliard says. "The first questions a CEO needs to ask are 'can we afford to fail?' and 'is it business critical?' The reality is those questions are rarely asked."

Many plans start simply as an ambition by a CEO anxious to project an impression of activity, ambition and expansion. These start life as a statement of an objective, actual planning starting afterwards. Unfortunately, the resulting plan often differs radically from the original aim but the CEO may be unwilling to change it and be seen as lacking in vision or, worse, weak. So, the program goes ahead with unrealistic objectives and wildly optimistic timetables.

The next mistake that is often made is to add the program to the day-to-day workload of managers. "Many companies fail to make the distinction between a business-as-usual sort of program and something they haven't done for three or four years," Hilliard says. "They try and time-share with other things that are going on in the business and that never works."

Large programs should be regarded as stand-alone operations with their own objectives, timelines and budgets, Hilliard says. "The only way to manage a business critical program is to have a dedicated core team kitted out with people who have done it before and who can bring in people with additional skills, as required. The evidence is overwhelming that if you don't do that you are likely to have a big delay and a massive cost overrun or a complete collapse."

Another factor that CEOs often overlook in setting up major programs is oversight of subcontractors. "Most big programs these days rely on at least one third party supplier, and it is always amazing to me how little management they get. It is often quite a long way down the road before problems spring up and then you find yourself in long, turgid meetings asking them for recovery plans," Hilliard says.

Simple overconfidence can also be a factor, Hilliard believes. "Frequently managers are naive, thinking failure can't happen to them."