David over at 37 Signals has a great post today about the problems with averages. It boils down to this: making decisions based on averages (average wait times, average turn-around, average production) means your making decisions on flawed data.
If a call center answers its first 900 calls in 1 minute and its next 100 calls in 20 minutes, it's average pick-up time is just under 3 minutes. Not too bad. Problem is, 10% of its customers waited a painful 20 minutes. Not too good.
Reading David's post will give you some better metrics than averages for measurement. I just want to point out that no underdog brand ever got anywhere by focusing on averages.
Underdog brands focus on outliers.
They measure where they are having the most success and they do more if it. They find where they are failing miserably and they learn from it. And though it may not be wise to get mired in analyzing extreme outliers--Black Swans--you can usually learn a lot more from outliers than averages.