The cost of complexity (aka KISS)

Human beings seem to like making things complicated, but when it comes to business systems this can be bad news.

System reviews often start with a simple, but pressing, requirement eg.

• get a better idea of what sales are being worked on;
• what marketing works;
• store all information centrally.

Then best practice says you should:

1. list your requirements;
2. assess what is available to meet those requirements;
3. meet with a shortlist of vendors;
4. evaluate and make your decision;
5. implement.

I don't disagree with any of the above but the work involved must be in proportion. There are traps to fall into at each stage:

1. List your requirements - if you sit down with a blank sheet of paper you will almost certainly produce a lists of 'must haves', 'wants' & 'would likes'. It's worth understanding the three categories as the longer the list, the more complex & costly the finished system.
2. Assess what is available to meet the requirements - if stage 1 went well then you have a good solid list. If you spend too long looking at the options then you will generate a mountain of information that will take even longer to filter through. Three or four options should usually be enough and please consider your budget at this stage. If a system is too expensive then it probably isn't right for you.
3. Meet with a shortlist of vendors - it sounds silly but the most common trap here is to have a short list that isn't short. If you invite 15 vendors to the party then nobody gets what they need - least of all you.
4. Evaluate and make your decision - if there is an obvious choice then that's OK, you don't need to look to make it more complicated. If it's not obvious, then a team meeting, a white board and simple scoring system might do the trick. List your requirements and score each system as to how well it meets them - 1 = doesn't, 5 = completely. This will highlight where you need more info and may give you an answer. Please don't rely on the vendor to do this exercise as they may be biased!
5. Implement - This is where it is crucial to keep the list from stage 1 handy. Systems people call it 'project creep' - new requirements added along the way will cause sleepless nights as well as cost and time over runs. Set your goals, make sure everyone sticks to them and have some measurable objectives if possible.

There is always a reason why an IT project goes wrong and being over ambitious or complex is up there at the top of the list.