And this on idealgov - procuement.
Well I think that IT procurement in this country could be made at between a tenth and a twentieth of existing cost using the sorts of ideas discussed above. But this is more a job for mathematical financial market modelling. Some of the issues are big controlled contracts against many small contracts left to self organise, strong defining principals against amorphous requirements creep, efficiency of of need propagation through a system, top down, bottom up or middle out. And so on. I think these things can be modelled, but I also think we can all take a stab at what would be sensible types of answers. As a first step it would be all round helpful if the government simply had less money to spend. With less churn and lower expectations smaller players could begin to operate and show their intrinsic strengths. Let me put it another way. We think there has been a financial crisis and if you are like me you probably tend to think of it being fairly restricted in time and sector. But, actually, no. The reality is that there has been an orgy of spending going on, not least by government in IT. The reality is that it is not possible to spend huge sums of money efficiently, it generates churn, and that churn has been building up over the last fifteen years. When there are very large sums of money to spend (let's spend 4 billion on the NHS, oh it's going to cost 12, no matter, I can pay) it is very difficult to even contemplate spending it through any other party than the very largest players. They, of course, couldn't cope with this either and ended up losing money! That is what I mean by the inefficiency of churn. I remember, about six years ago, attending one of the early meetings for the NHS IT project where the promise was made from the Civil Servant responsible that there was ample room for small companies to become involved. In fact this was the reason for the meeting. But once the green light was given for this avalanche of spending their fate was sealed, because all expectation became ridiculous at that point. The only way this would have worked would have been if expectation had been lowered. This is the first fifth of saving that agile can bring - simply don't do the last 20% of requirements. And don't expect the project to be finished, software doesn't get finished, it gets overtaken by changed requirements. So all of this requires a very different out look than what has been traditional for government procurement. Software is no the same as running a railway, commissioning a power station or building a battleship. Civil Servants need the think space to consider small and organic, they need the information and financial theoretic models to guide how they make these decisions and we all need to accept that this is, as yet, a young industry and that procurement, too, is in its infancy. Finally the Civil Service needs the power to act free from government pressure. I believe that Thatcher made all the encouraging signs to big business and then Blair insisted on a very big spend in the context Thatcher had provided. It has been a disaster. A firmer, more confident and knowledgeable, Civil Service would not have allowed this. And it is only the Civil Service that can change the existing landscape. With pressure from the outside. One further point. Can any one demonstrate that the huge sums spent in gov. ICT have been necessary? Without that demonstration it stands that it could be done for less. Now you see why so many people, both in and outside government, would want to keep this unscrutinised. No one wants to admit to wastage on any scale. But how can the future savings be made without that admital? -- AsaltlIdeal (2010-03-02 00:12:40)