|
The very first thing that should happen in a project is all of the logical
stuff. Analysis, scope, architecture and objectives.
If you skimp on the logical stuff the pain will increase as the project progresses and
money will just disappear into the project for no return.
If you do it well the project will flow smoothly and appear easy. The users will know
what they are getting and the project team wil know what they are building. Best of all
the management team will have a way of measuring success.
Everyone will wonder why you spent x weeks doing analysis work upfront when the
project is so simple...
Ironically it is managers and less-experienced developers who want to skimp on this
phase and just get on with the coding. Resist.
It may sound obvious, but the most fundamental failure you can make on an IT project is
to solve the wrong problem. Unfortunately this happens a lot. The more bureaucratic
organizations seem particularly keen to make this kind of mistake.
This area concentrates on making sure you get the logical bit right. It starts by
looking at the initial analysis in "What's the Problem",
moves on to look at possible solutions, prototypes and architectures.
Finally, ask yourself how many projects you have seen where the criteria of success
were not specified at the start of the project.
|