How to Measure an Agile Team
The long standing question of metrics, measurement and performance in Agile teams has people in quite a muddle. This is one of the most common themes of questioning on the forums and boards. People want to use velocity as a measure, or record detailed numbers about who did what tasks.
The answer to this question is to throw out all this stuff. It takes up a load of time and tells you nothing.
I have said this before, but I have tested it in action and I can say that I genuinely know when a Sprint is going to be a success based on how joyous the team is. Smiles, jokes, playfulness – these all indicate that things are going to go well. I also believe these signals are evident well before your burn-down chart tells you the story.
It really is this simple. If your team look a bit deflated, or are unusually quiet or aren’t having some fun – you’re going to miss your Sprint.
So watch out for unhappy people and find out what it will take to make them enjoy the work again. This is how you’ll develop a team that can crush features.
Chicken and Egg
Does a happy team give you a successful Sprint, or does a successful Sprint give you a happy team?
Of course to some extent this can be quite circular. If you have a happy, well-performing team you will get a successful Sprint and this will contribute to how good people feel. You could measure subjective well being on a team to see where you get movement first – I suspect that once the team is happy with the method of working, the amount of work expected and their working environment you are more likely to have a successful Sprint. If people were marking tasks on the Agile board with smilies, you’d see happy faces before you saw a successful Sprint. Of course, this could be gamed by giving unhappy people a very small amount of work to make the Sprint appear successful.
Another interesting way to measure this is in reverse. You can measure a successful team and see what changes first – do they fail a Sprint first, or does the mood change first?
The evidence I present to the case of subjective well being pre-empting success or failure of a Sprint comes from Jim Benson’s Why Plans Fail: Cognitive Bias. It is a great read, so go ahead and dive in for yourself for all the details – but Jim describes a very powerful effect not only in measure subjective well being (and discovering problems before they manifest themselves in burn-down charts) but also in how to improve it by celebrating your victories to avoid skewed emotional responses due to availability heuristic.