On a recent trip to England, I had brunch with Alex Budzier, a researcher at Oxford University, whose work with professor Bent Flyvbjerg has shown that IT-based strategies have unusual patterns of failure and success. Alex was a senior associate at McKinsey before joining academia.
Alex shared statistics over scrambled eggs. “We found a strange pattern with IT projects. They have distinctly different patterns of failure – it’s a fat tail, not a normal distribution. More of the really big projects fail, more often. They often cost 2-4 times what is expected. And what’s more, nearly 50% of IT projects are under funded… they are starved.”
Alex continued, “People get fired for failure, these big projects are high risk for careers. I’m looking into why a room full of experts that get paid good salaries to be rational are still not making rational decisions.” Alex and I agreed that this was peculiar, and discussed how politics, policies, psychology and group decision-making influence risk and forecasts in business cases.
Alex’s key points on business case, projects, policy and politics:
- Smart people are still part of the social system of the organisation, approval and language affects them
- Big projects are politically good for careers when they are kicked off even if they risk careers later on
- Early warning systems for project success/failure need to consider more than average performance or % of on-time delivery.
My point to Alex was that instead of looking at success or failure as a dead certainty, let’s use uncertainty in our favour. Statistical analysis and predictive algorithms can identify which project is more likely to succeed than another. This helps executives focus attention to where it is needed most.
Alex and I traded stories. One was about the IT project at a Middle Eastern bank that was successfully delivered on time despite a merger (Alex mentions this in his Harvard Business Review article in the Sept 2011 edition). Alex wondered what the secret of the project’s success was. I shared what a friend on the project had said: “You should have been at the Steering Committee Meetings. The Sheikh would come in with his big bodyguards – they had shining, sharp scimitars in their hands. It kept people really focused.”
If you don’t want scimitars brandished at your project meetings or your career on the chopping board, get into predictive algorithms!
Do more for strategy, projects and prediction, see Joanne at www.whatisaproject.com. Joanne Flinn is the author of “The Success Healthcheck for IT Projects” (Wiley 2010).