Statline article from The Quality Magazine, February 1994
Remember that what you measure affects behaviour
Teresa Dickinson and Geoff Robinson
How companies choose to measure their performance affects how people behave. This is often subtle. The changes to behaviour may be difficult to foresee.
Here is an example.
The management of a company which provides installation and maintenance services for its products was concerned that many customers complained that the company didn't provide service when they wanted it. The customers complained that servicemen often called at inconvenient times, were late, or did not show up at all.
Management decided that delays between requested times and actual service times should be measured, and that the percentage of services provided when required by customers should be reported to management weekly for every region.
What did management intend? They thought they would get some real data on how bad the problem was, including some indications of where and when the problem was worst. Also, they thought the servicemen would become aware that their performance was being monitored and so, somehow, lift their game.
This idea seemed to be working. The figure shows that the percentage of services provided when required by customers increased rapidly over the first few weeks that the measurement was reported. The desired consequence of measuring performance seemed to have been realised.
But what was happening `in the field'?
The servicemen, naturally, concentrated their efforts on improving the measure used to judge their performance. How did they do this?
After a few weeks the management noticed something else. The company was getting more complaints than usual about faulty installations. Checking back over service dockets confirmed that the number of repeat services to fix recently installed products had increased.
The apparent improvement in performance was illusory. In order to improve the measured aspect of performance the servicemen had made compromises which had degraded other, unmeasured aspects of performance.
An entirely negative aspect of the measurement was that recording this data reinforced the 'them-and-us' barriers between management and the servicemen. The servicemen perceived that they were not being trusted to do their job.
Examples of measurement causing undesired behaviour include the following scenarios from other organisations.
How do you minimize the problem?
The worst disasters can be avoided by even a small amount of effort to be wary of possible behavioural consequences of measurement.