How to Fail at Software Development |
||||||||
16. Does anybody really want this thing? |
||||||||
| Score from 0 to 20 | ||||||||
| This is the most important question of all. This is just as deadly as the situation described in the previous question, but this one is much more common. More projects die of this condition than any other. If this question cannot be answered with a resounding, “Of course!” then not much else matters. Is there real support for this project, or is the whole thing just some sort of political football? If there is no support that runs all the way up through the levels of management, then you can just forget it. One way or another it won't make it. No software development project can survive anything less that one hundred percent support from the company. This is not an actual part of the project, in the sense that it can’t be controlled by anyone on the software development team, but without proper support from the rest of the organization there is no need to try to go on. One day, without notice, the plug could be pulled and it would all be over. No ceremony, no bang, not even a fizzle. Just a quiet vanishing. Or, barring that, without support it will be impossible to get the necessary facilities to do things properly. Software development takes time and money. If management disagrees over the value of the project, then there will be a compromise made with the funding. That's a slow death, but it is a certain death. But even if facilities and funding are not problems, without proper support of the actual users the end result will be a useless mess. It will be useless even if it comes in on time and under budget because it will be wrong. Dead wrong. Software developers can solve a problem, but they can't define the problem to be solved. And there is nothing worse than well-written, brand new, green, untested software that’s dead wrong. Chapter 17 of How to Fail at Software Development explains several situations that can cause management, at various levels, to do everything it can to kill a project. Even though it costs the company money. |
||||||||