For those who don’t already know, FIST stands for Fast, Inexpensive, Simple and Tiny. It’s a set of values that can be applied across the spectrum of decision making (organization, requirements, architecture, design, testing, etc), for all kinds of technology development efforts. Whether you’re building an aircraft carrier or a hand-held radio, there is always an opportunity to chose between fast and slow, inexpensive or pricey, complex or simple, and large or small.
I recently came across some research by The Standish Group that strongly supports the FIST approach. In fact, they make me look downright conservative. They write:
Our research has shown smaller projects are consistently more successful because of reduced confusion, complexity and cost.
… smaller projects experience fewer cost overruns.
… shorter time frames… increase the success rate.
… the more expensive a project becomes, the less likely its chance of success.
Time is the absolute enemy of all projects.
Our newest data suggests that we need to further reduce the amount of resources to increase the success rates even more. … no more than four people, for no longer than four months at a cost of less than $500,000. We find that the less the features, the greater the yield.
All I can say is yes, yes, a thousand times yes!
5 comments:
Wow!! very interesting they are saying that in such concrete terms.
Related story to this topic: I was in a meeting today where a superior of mine said "To do this faster we need more people and more money". I decided this would be a good opportunity to make a stand and said "Or more money and more peopel will make us go slower". I got a blank stare followed by a "well, done right, more money and people will make us go faster". I decided to concede the point because anything counter point I might have would have fallen into his "not done right" category. In retrospect, I might have been more successful if instead of saying "more" could be slower, I offerend how "less" could be faster. lessons learned I guess...
I'm so proud, Pete! Nice job standing up and saying "Hey, that's backwards!"
As for that dude, I contend that adding more people and more money is, by definition, NOT doing it right.
Sure, there is some sort of threshold, some minimum value below which there are simply not enough resources to get the job done... but that floor is WAY lower than most people think. In the DoD, we hardly ever get there.
Exactly
Another tack is to say HOW MANY extra people you'll need.
Say you want to go from 12 to 20 programmers... Go check the Fred Brooks Million Man maths on that.
So if you are goig to add 8 programmers you'll also need to add at least 1 team leader, maybe 2 analysts, and a tester or three...
Agreed - Brook's book is brilliant (mythical man month, right?)
Post a Comment