Software project management is a complex subject. And apparently, in Belgium, we’re not very good at it – we often overshoot our deadlines.
The question is: where does the problem lie ? In my experience, analysts and developers are just as good as anywhere else.
There are two aspects to it:
- setting deadlines
- managing scope
Both of them are tough.
Especially in consultancy, there’s a tendency to agree with the deadlines the customer ask for, without discussion. The fear to lose a customer is a strong one, especially since the slump of 2002-2004. The first ones to go in case of recession are the consultants. And the competition is fierce. So top managers and sales people have trouble with the words ‘there’s no way we can do this in three months’.
Managing scope: there are very few projects where the scope is clearly defined from the start. And there are also very few projects where the customer won’t want to add a few features.
If you want to keep to your deadlines, Just Say No, or wave the Change Request flag. And this doesn’t go down well with the customer.
Some methods have been developed to deal with this: extreme programming, agile development, etc. These adapt the scope as they go, fleshing out a bare-bones skeleton. The deadlines are guaranteed by making sure they have a ship-worthy application every step of the way.
Strangely, these techniques are rarely applied (again, in my experience). They’re less easily applicable when dealing with legacy systems. And there’s a tendency for large companies to stick with good ol’ waterfall (though i’ve been told Scrum is applied at KBC).
At my workplace, one project is managed with Scrum. I talked to the project leader. He was very enthousiastic about it, and offered to give us a talk about how they applied it to their project.
Strangely, this didn’t interest my colleagues that much. There’s a sort of blasé attitude ‘no method is magic’, and ‘oh god, another hype’. While i partially agree – a badly applied method is almost worse than no method at all – it’s never a bad thing to soak up new ideas. Some of them might be worthwhile.
So in sum, we can do better. Both in straight-forward dealing, and in willingness to reconsider existing methods. There’s no magic receipe – but better tools yield better results.
At Boulevart.be we have introduced Scrum. One huge project is already in an advanced stage and it works fine so far.
An answer to your remark about why we can’t we stick to deadlines might also be due to change. When a customer asks for changes, even the smallest one, isn’t taken into account rearding deadlines. Change management isn’t applied.
Yes, i agree with you there.
In fact it might be interesting to know what they use for their comparison. Initially planned deadline ?
I’ll look into it.