"Dysfunctional Performance Through Dysfunctional Measures"

Andy Neely, Mike Bourne, and Mike Kennerly

Journal of Cost Management

September/October 2003, pp. 41-45

The authors of this article provide several examples of how poorly designed measures can lead to inappropriate action and behavior by employees. One example given is of an airline that measured baggage-handling performance by assessing the lapsed time between the plane landing and the time the first bag hit the conveyor belt. The employees soon learned to quickly grab a very light bag and rush it to the belt. Then they unloaded the rest at their leisure.

The article recommends that one should "deconstruct" a performance measure and think about its essential elements before trying to implement it. The components for measures to look at are:

• Title—What should the measure be called? Does the title explain what the measure is? Does it explain why the measure is important? Is it a title that everyone will understand?
• Purpose—Why is the measure being introduced? What is the aim of the measure? What behaviors should the measure encourage?
• Target—What level of performance is desirable? How long will it take to reach this level of performance? How does this level of performance compare with the competition? How good is the competition currently? How fast is the competition improving?
• Formula—How can this dimension of performance be measured? Can the formula be defined in mathematical terms? Is the formula clear? Does the formula explain exactly what is required? What behavior will the formula induce? Are there any other behaviors that the formula should induce? Is the scale being used appropriate? How accurate will the data generated be? Are the data accurate enough? If an average is used, how much data will be lost? Is the loss of data acceptable? Would it be better to measure the spread of performance?
• Who measures—Who, by name, is actually responsible for making this measure?
• Source of data—Where will the data to make this measure come from?
• Frequency of measurement—How often should this measure be made? How often should this measure be reported?
• Who owns the measure—Who should be held accountable for performance against this measure? Whose neck is on the line if performance does not improve?
• What do they do—What actions will they take to ensure that performance along this dimension improves?

Back to top of page