A quick overview of percentiles
Using a percentile methodology allows you to evaluate the “long tail”, the extremes within your workforce that can add-up overtime (and become very costly). Optimizely did a write-up that does a wonderful job illustrating the dangers of averages:
“While the average might be easy to understand it’s also extremely misleading. Why? Because looking at your average response time is like measuring the average temperature of a hospital. What you really care about is a patient’s temperature, and in particular, the patients who need the most help.”
If we were to take this and apply it to our workforce; When you’re dealing with thousands of workers over the course of your season, the average can become extremely misleading. Why? Because looking at the average employee hourly rate, or average bins/hour is like measuring the average temperature of your entire packhouse throughout the course of a year. What we really care about is an individual CA room’s temperature, and in particular, anomalies in temperature that we need to address.
This same concept applies to your workforce, and percentiles allow you to focus your efforts on those that need the most attention. This is where percentiles come in, as a method to center your attention on the extremes of your workforce that can accumulate to tens of thousands (or even hundreds of thousands) of dollars each season.
Above: A typical percentile distribution found at PickTrace apple customers
In the above diagram, you’ll see a fairly standard finding across PickTrace apple customers while in harvest. The “contribution” column represents the percentage of contribution towards yield that the corresponding percentile contributed