Of course, that was all nonsense. New projects are not created out of nothing. They don't spring from the head of Zeus fully grown, armour plated, and carrying a spear. Before it can even consider starting a project, a company needs to develop a substantial working environment. It needs people: analysts, programmers, project leaders, and testers; it needs stuff for the people: desks, chairs, terminals, big computers, little computers, conference rooms, coffee pots, and mouse pads featuring the company logo; and it needs money to buy all that stuff and money to pay the salaries, medical benefits, and pension and profit sharing contributions of all the people.

The stuff and the money are the easy part. The people are the hard part. Most companies haven't a clue as to how to recognize data processing talent and, without that, they will produce a stream of mediocre, badly designed systems that will cause endless problems for them. The sad part is, they'll probably never know why.

In addition to an environment, a project also has to have to have a reason for its existence and a decision to go forward and actually do it. How are those decisions made? Why is this project chosen rather than that one? Like it or not, this is data processing too. The medium might be slides and cost projections rather than databases and screens, but without it the project will never come into being. A lot goes on before that first box is drawn on the whiteboard...