Success or Failure?
An article by Colin Thompson in Computing (20/4/06) argued that to concentrate on the failure of IT projects could result in IT projects not addressing the real needs of organisations. Rather, success should be judged on how well these projects enabled organisations to move forward. It might seem that Project Doctors would have a vested interest in opposing such views but actually we whole-heartedly endorse them ... but don't think this is limited to IT. The problem, of course, lies in the definition of project success - the well used examples of the Dome and Concord illustrate how views on success can differ markedly.
Risk Management not Risk Avoidance
It is certainly true that an undue focus on risk avoidance can paralyse projects. Building in too many checks, risk reduction activities and mitigation strategies will make any project longer and more costly than it needs be. Risk management expenditure must always be proportionate to the risk if we are not to emulate the poor woman who "swallowed the spider to catch the fly ... etc.". Projects are of their essence risky, so risk aversion is not an appropriate mind-set for project managers - but neither is risk oblivion. Risk management is what is called for.
A rounded view of success
The solution surely lies in having a rounded picture of what success means for a project. The balance will be different for each project. It will, for most, have elements of delivery to budget, schedule, and specification - but it must also include the facilitation or realisation of benefits (which addresses Colin Thompson's point) and may well include criteria which are more difficult (though not impossible) to measure, such as customer satisfaction and improvements to customer working practices or environment. The priorities assigned to these different criteria must be clearly understood by all those involved as, in the real world, one may need to be sacrificed to some extent to enable another to be achieved.
And then there is the little question of whose definition of success counts ... but that is for another day.