It's no secret that journalists are fans of the scary statistic, mostly because nothing is more effective at grabbing a reader's attention.

Think we're exaggerating? Try this stat from Accenture on for size: Only 29 percent of IT projects are considered successful. According to Accenture, the average IT project exceeds its projected cost and schedule by 56 percent and 84 percent, respectively.

VitalSmarts and the Concours Group attempted to identify the major issues contributing to project problems and came up with five: unrealistic deadlines or budgets; sponsors who are not actively engaged; not enough participation in the priority-setting process; team members and team leaders who do not acknowledge problems; and team members who are unable or unwilling to support a project.

So how to solve these problems? There are no easy answers, but a recent profile of Chevron that appeared in Computerworld New Zealand offers some interesting ideas.

A framework called Project Everest helps ensure that projects that yield the biggest business benefits for Chevron receive the right level of investment, as well as special management attention. All projects under the Everest umbrella are top-down initiatives, says Chevron's CIO, so that competing needs of local and regional business units don't hamper more strategic initiatives. Thus, business drives IT and not the other way around.