An experience over these last few weeks reminded me of the 80/20 rule or the Pareto principle. In software development this is the observation that the first 80% of requirements are easy, the remaining 20% are difficult. And most importantly it’s usually the important requirements, the ones that realise the most benefits that are in your 80% (as you usually focus on the key requirements first).
As a number of software companies understand, once you’ve got your 80% then deploy the software. Yes, we all would like things 100% perfect. But there comes a time in building software that the extra effort to realise that last 20% is just not cost effective. In fact you may be totally wasting your time on working on requirements that are simply not that important or worse completely irrelevant.
There is of course a risk that you may have an important requirement in the final 20% of requirements. But you could put a lot of effort into those last 20%, significantly delaying the completion of a project, and find none of them were that important.
If you deploy at 80% you get most of the important requirements out there available to the end user faster and you are more likely to identify the remaining important requirements quickly afterwards. You deliver to your customer and can see in the field which remaining requirements are important, de-scoping the rest (and thus reducing the length of your project).
Another great example of the 80/20 rule that I think is relevant to law firm IT departments (or any IT department) is one that I found in Tim Ferriss’ book “The 4-Hour workweek”. That is to fire the 20% of customers who take up the majority of one’s time and cause most trouble.
Now I don’t recommend firing or ignoring the 20% of your customers in the law firm that are the most troublesome! But rather you do not allow your time to be totally taken up by looking after this 20%, rather focus your efforts on the 80% who are as important but perhaps not as vocal in their complaints!
I’m sure I have come across software where they have been following this the other way around (i.e. wait until it is 20% right and then launch!)
I have seen this principle quoted the other way around in relation to clients of professional services businesses as well – that it is likely that 20% of your clients generate 80% of your profit. This doesn’t work for all businesses (not even all law firms), but if it does apply it can be useful to bear in mind when decided where to focus your time and resources.
Now I’m not sure if your first sentence is in jest? There certainly are software companies that release clearly before they’ve finished any testing! It’s one thing releasing once your key requirements are done, but never release something that clearly doesn’t fulfil any.
The overall driver though should be get a working version with the key requirements in front of the people who will use it as soon as possible. That way you can get feedback quickly and review the requirements, focussing on the important and discarding the unimportant.