In Part 1, I focused on variation and subsequent adjustment of a physical system such as a machining operation. But this doesn’t apply only to systems where you are making a physical product. This applies to any system with outputs that can be measured. Consider sales activities. A very common practice is to give a “salesperson of the month” award. But what is being rewarded? Is it really outstanding salesmanship or is it something else? If you don’t understand the variation, you may just be rewarding the “outlier of the month”. A better approach would be to analyze your historical data and try to understand the variation. If you have at least a handful of salespeople and a few years of data, create a histogram of monthly sales values and see what that looks like. My reading of Deming reminded me that you could even create a control chart for this, assuming you have enough data.
Side Note: I have yet to encounter a business that used control charts to understand whether team member performance is in control when considering scolding or rewarding behaviors in a non-manufacturing environment. If you do, I’d love to see how you apply this, so please ping me.
The same logic applies when you are considering reprimanding someone for poor performance. The following scenario is based on another of Deming’s examples. Consider that you have a jar with a mix of red and black beads in a 1:10 ratio, respectively. The employee’s job is to “manufacture” red beads by pulling them out of the jar. There is a quality mandate in place that says that they must produce 1 red bead for every 3 black beads. Needless to say, not many employees meet the quality standard. Those that do are rewarded, and those that don’t are punished. The poor performers are “retrained” to remind them that they need to stop producing so many black beads. Does this make any sense? No? Then why do so many companies operate in this fashion? I bet if you think about it, you can come up with an example at your company that falls directly in this camp.
The “adjustments” in these scenarios are often new rules/policies that either don’t make any difference in performance (e.g. increase the sales bonus) or make the situation worse (e.g. firing a good employee that produces too many black beads). The mistake here is confusing a common cause (system variation) for a special cause (the employee isn’t performing well against the target). Unfortunately, this is a tragic situation that I see all too often. That is, employees are chastised for poor performance when the expectations are unreasonable given the work environment. Deming also said: “A bad system will beat a good person every time.” If you are a manager, you should understand the variation and expected performance of your system. It’s very possible that your employees are exemplary, but you are forcing them to work in a bad system.
In Part 3, I will provide guidance on when you should adjust and how this concept relates to the difference between problem solving and improvement.
Don’t adjust,
Dave