Why Many 'Big' Systems Projects Are Doomed Almost From The Start

There have been a number of reports recently of escalating, and in some cases, quite extraordinary costs for computer systems.

This type of 'Project Risk' seems to be unavoidable; wherever a 'big systems' project is to be found, there are often large unforeseen additional costs, late delivery and functionality that was not entirely, well ... functional.

Although many of these systems are 'big', what they actually do is often not particularly complicated. It is often asked how such huge overruns can happen, and indeed why they continue to happen.

There are many reasons for this, but there are often common factors that soon become apparent. One of these appears to be the role played by consultants in many of these projects - at this point I have to tread carefully, since I am now myself a consultant.

At one point in the past, I was looking into some possibilities for getting some technology work done for a large financial institution, and was bombarded with approaches from various consultants, all eager to earn what could have been quite a significant amount of money.

I met with a number of these, and some common traits were immediately apparent. In particular they were all immaculately turned out, and the technical buzz words flew. Almost psychedelic Powerpoint presentations appeared with remarkable consistency.

These people could clearly 'Talk the Talk', but could they actually 'Walk the Walk' ? - I started asking them technical details about how and why their solutions would actually work.

At this point they would invariably explain that they were not really the technologists as such, but more on the sales side. So I would ask them if they could arrange for their technologists to come and see me.

Some time later, the 'technologists' would usually turn up, and launch into the same blizzard of technical buzzwords. I decided to turn things around and ask them some basic technical questions - the kind of questions that I might have asked a candidate for an intermediate level systems development post. The result was that in most cases they failed abysmally.

In one case, one of the services that they was being offered related to the development of what are known as 'COM' components. I asked if they could tell me what COM actually stood for - they had no idea, and the chequebook stayed firmly in my pocket, as it usually did.

I was perhaps a little out of step in believing that, while 'presentation' is important, 'substance' is actually more so. I was reminded of a quote from Thomas Moore - he was defending a 'Jester', who was perhaps not very good at playing the fool, with the words:

“It is less injurious to the commonwealth when wise men go in fools coats in jest, than when fools go in wise men's in earnest”.

A point which we think should be taken to heart by those involved in the procurement of complex computer systems.

Incidentally, COM stands for the: Component Object Model.