Once you start to develop a few KPI's the real questions start. The urge will be to want to do something at every time the KPI is below the average. It's estimated that 95% of variation in a system is caused by the system itself. So, since the whole reason for having the KPI is to know when something is out of whack, you'll be doing nothing 95% of the time. Want to know how to create a crisis? Try to find out why your KPI is "below average" every single time is goes below the line. The goal here is to know what is normal variation and what is out of whack variation. You are the best one to know this, but try to know it before you react. There are many statistical methods for coming up with the limits to normal variation, but for a quick a dirty start to managing your first KPI try to eyeball the line chart and estimate the usual high and low points for your system. Now, every period that you look at your KPI, you have to ask is this inside the normal system? If the answer is yes, then your response should be, ok, everything is normal and I should not do anything. So many times, managers want an answer to the reason the KPI is above or below the average every week when their team reviews their performance, when the real answer is everything is normal and that is why the number is what it is. It's up because its up and not because you are a star, or its down because its down and not because someone screwed up.
Once you start to develop a few KPI's the real questions start. The urge will be to want to do something at every time the KPI is below the average. It's estimated that 95% of variation in a system is caused by the system itself. So, since the whole reason for having the KPI is to know when something is out of whack, you'll be doing nothing 95% of the time. Want to know how to create a crisis? Try to find out why your KPI is "below average" every single time is goes below the line. The goal here is to know what is normal variation and what is out of whack variation. You are the best one to know this, but try to know it before you react. There are many statistical methods for coming up with the limits to normal variation, but for a quick a dirty start to managing your first KPI try to eyeball the line chart and estimate the usual high and low points for your system. Now, every period that you look at your KPI, you have to ask is this inside the normal system? If the answer is yes, then your response should be, ok, everything is normal and I should not do anything. So many times, managers want an answer to the reason the KPI is above or below the average every week when their team reviews their performance, when the real answer is everything is normal and that is why the number is what it is. It's up because its up and not because you are a star, or its down because its down and not because someone screwed up.
