Workforce Management Analytics: Related Metrics – Bar Charts Gone Wild
At the end of our last discussion, we saw that some of our metrics were affected as a result of the metrics we were trying to change (reduced travel time equated to better productivity) and I used the term “related metrics.”
Let’s stick with Productivity, and let’s say we are trying to measure jobs completed. I happen to think this is one of the worst metrics to measure, because it is subject to so much variability, but that’s what makes it a really good example. Here’s what I mean.
The number of jobs completed per day is subject to:
- How many jobs scheduled per day
- How many jobs worked per day
- How much idle time (time spend waiting for more work) per day
- How much travel time per day
- How accurate the planned duration is (if we plan for 2 hours and it takes 4, that’s bad)
Here’s a quick example. My most reliable and most skilled technician Joe completes an average of 4 jobs per day, so we have scheduled 4 jobs for Joe to work today.
- Job 1: Travel time to this job is doubled. Joe works this job and completes it. Due to the travel, Joe won’t have enough time in the day to complete his last job of the day.
- Job 2: We planned for 2 hours and it takes Joe 4 hours. Joe can no longer get to his 3rd job. Joe doesn’t have the part to fix job 2 so he cannot complete the job.
Joe completes 1 job today. Joe’s productivity dropped 25% because of related metrics. The result is flawed because related metrics impacted our core metric of productivity.
All of the metrics above are related to and will impact our job completion rate. Our example above yielded a negative response. In my example from last week, our related metrics provided the solution to our problem. While related metrics can help you achieve your goal, they can also compete with each other and sabotage each other. So on the day of service, you have high productivity planned and you get that impossible customer escalation…what do you do? Hope to see you back next week. It gets super messy from here on out.