In the last 40 years we have seen computer systems reach into every part of our lives and today most of us carry several computers around with us all the time. They are in our mobile phones, our sat-nav systems, our cameras and camcorders. They are even in our purses and wallets where we keep our "chip" & pin cards. Their processing power has multiplied almost beyond measure and their prices continue to tumble.

But one thing hasn't changed. If they don't do what we want them to do then they can rapidly become very expensive headaches. They can even destroy successful businesses.

In the 38 years since I started working with computer systems I have developed a very robust view about how they should be developed and how they should work. Computer systems should be developed on the basis that there is a clear understanding by both the user and the developer of what the system is being designed to do. But if the user doesn't know what he wants then it is unreasonable of him to expect the developer to design and build something that meets his need. Equally if the developer doesn't really understand what the user needs then he can't possibly design and build it.

And the systems should do what they were designed to do, but all too frequently they don't!

A few years ago I successfully challenged Microsoft in the public arena (otherwise known as Watchdog!) for the very simple reason that their AutoRoute Express Software didn't recognise that Motorway Service Areas were directly accessible from the motorway! The software directed you to leave the motorway at a nearby junction and then navigate you by country roads to the location of the Service Area - but of course there was no public access at this point! Was it fit for purpose?

Today, when serving as an Expert in a Civil Litigation regarding any Retail Computer System the same questions come up time and again:

• Did the user know what they wanted?
• Did the system supplier know what was needed?
• Did they deliver what was needed?
• Did it work?

In the two largest cases that I have dealt with, both involving more than £1m, the reason for the project failure was not that the system supplier didn't deliver a system that met the user's needs; it was that the user didn't know what they wanted in the first place.

But, on the other hand, almost all of the small and medium sized cases have been caused by the system supplier either not understanding what was needed or failing to deliver a system that worked.