Does it sometimes seem that praising an outstanding performance is not as effective as calling out a bad performance? Is that true? You’d think people would want more praise than they do the discomfort of being called out. Let’s take a look at what happens when we praise a good performance and call out a particularly bad performance.

I am going to assume people getting feedback are qualified. That is to say they are capable of doing the work being asked of them. So on average the individuals can deliver outcomes meeting the requirements.

As an example let’s look at a widget producer. Every shift the widget production operators are supposed to produce 100 widgets. For the most part the widget production varies between 94 and 106 widgets. However every once in a while one of the widget operators has a very bad day and produces 78 widgets. You call the widget operator into your office and give them a good kick in the pants. You expect better from him or her. The same shift a widget operator hits 123 widgets. You go out to the widget area and commend the high producing widget operator, noting the need to continue the high level of performance.

The next day the low performing widget operator average 98 widgets for the next five days. It is not to target but it is a whole lot better than 78. The high performer on the other hand averages 103 widgets, far below the 123 they produced just five days ago. What is going on?

What you witnessed in both the great and poor performing widget outputs is a deviation from the mean. In one case the departure was above the mean. The high performing operator (Z) may have been better than the average operator but on one day they had all the stars lined up. The widget process ran well with little down time and when they hit 115 widgets they skipped their last break knowing they had a great run going. They could not reproduce those circumstances and performance returned to a level closer to the mean (X).

On the other hand the low performing operator (Y) may have been on the lower side of average but on one day everything went wrong. The process was balking, materials did not get to the floor and things went exceptionally poorly. The next five days circumstances were closer to normal (X).

The leader sees this as positive feedback not sustaining great performance and a good chewing out as creating improved performance. This reinforces a behavior that may not be as cause and effect as it seems.

Most of us do not produce widgets or oversee widget production. We do lead teams that are leading functions, departments and business units that in turn are producing some kind of outcome. We may want to consider how feedback is being provided on our teams.

When we see an employee has had a bad day do we chew them out and watch what happens the next day? Do we observe and take time to determine if it really was just a bad day or a trend that needs be addressed? For exceptional performances do we praise them and tell them to keep it up?

When we are observing performance we have the opportunity to call a huddle and use examples of performance as learning opportunities. Are there things we can do to our process to repeat strong performances? Are their things we can do to prevent performances that do not meet expectations?

Have we taken the time to develop performance measures and share those with our teams? Do we have ways of tracking progress that are objective and visible? Do we convey our feedback in a way that promotes the target outcomes we are seeking?

So before we give that feedback let’s ask ourselves the question, what is really going on here? Will our feedback change anything or are we just looking at a single data point in a broader distribution?

Joe Thompson