Project Management Debacle

There is a story today about the Department of Transport royally screwing up a project to save the taxpayer money. http://news.bbc.co.uk/1/hi/uk_politics/7784868.stm has the details but in theory the scheme would save £57M (although it doesn't say over what time-frame). In the end, the scheme cost more than double, saved only half of the prospective amount and cost the taxpayer some £81M instead. To say that this is complete incompetence is too harsh to others who have made less ridiculous cock-ups in the past over large scale projects. What annoys me the most is that the lessons are never learned.

For instance, look at the passport computer system, the 2012 Olympics, the Millenium dome, the new Air Traffic Control system. All of them varying degrees of overspend. Of course once the projects start going awry, there is generally no option except to keep spending until completion, you can hardly write-off £50M and end up with nothing can you? I'm sure many articles have stressed the risk of large projects, the importance of breaking projects down into conceivable and manageable sub-projects etc but although this must be standard knowledge, the mistakes continue.

I want to propose something more radical for large scale projects. A large projects requirement specification that would be mandatory for the public sector and advisable or perhaps mandatory for the private sector too (their mistakes can impact the public after all). This requirement would end the days of "the timescales were too unrealistic" and "the budget was far too low". It would require such things as project managers being part of some parent organisation which demonstrates their ability and accountability. It could require that timescales be stated as minimum realistic and likely overrun. It would require that if a cost-saving exercise, costs be produced for the range of outcomes so that it is not given the go-ahead if only the best-case scenario is worth it (since projects rarely end up with the best-case outcome). If it is a money saving scheme then the requirement could say that permission would only be granted if the saving is at least X% of the cost over 75% of the outcomes etc. It could require that a Project manager signs his authority to say that the timescales are achieveable and if they are tight that money exists to ensure they do not slip.

Of course, this is a great idea and of course, it would never be implemented, partly because it is far too simple and partly because people would never agree what numbers would be appropriate for the spec. However, I personally think there are issues much deeper that would need to be addressed. We have a "cheapest price/shortest time" mentality for the most part and we don't trust people to give us honest prices for work. We therefore beat down their margins until they cannot afford to shoulder any contingency if problems arise. We can't accept that projects need to be planned well and might take longer overall but with a saving in cost and a higher confidence in the quality of the outcome. We cannot always plan things so that if the project must be shelved, that we might have at least gained some benefit in the part of the project that has been finished.