A September 2003 Harvard Business Review article, “Why Good Projects Fail Anyway” by Nadim Matta and Ronald Ashkenas (free summary here), says that the high failure rate of major projects — not just IT projects — suggests that either these projects are inherently unmanageable or else something is wrong with the standard approach to project management.
Matta and Ashkenas argue that the standard project management model is designed to control “execution risk” — the risk that designated activities won’t be carried out properly — by means of project plans, timelines, and budgets, but ignores two other equally important risks:
- White space risk: The risk that some required activities will not be identified in advance, leaving gaps in the project plan.
- Integration risk: The risk that disparate activities will not come together at the end of the project.
As a result, the traditional approach has a decent chance of success only when the end product is thoroughly understood.
Matta and Ashkenas outline an alternative model which uses a series of miniprojects or “rapid-results initiatives.” This reduces the probability that critical activities will be excluded from the plan and improves the probability that all the elements can be integrated at the end.