When a measure becomes a target

Organizations use numerical measures to make performance visible, but a measure can change behavior once rewards depend on it. The effect is not always the improvement that managers intended. An easily counted activity may represent only one part of a broader goal.

Imagine a student support office that rewards staff for reducing the average length of appointments. Shorter meetings might indicate that staff have made explanations clearer. They might also mean that complicated questions are being postponed or passed to another office. The appointment length alone cannot distinguish these possibilities. If staff concentrate only on that number, students may leave quickly while their problems remain unresolved.

One response is to combine several kinds of evidence, including whether the issue was resolved and whether students needed to return. However, adding measures also increases the effort required to collect and interpret information. A useful system therefore needs a manageable set of indicators tied closely to its purpose. Numbers can help reveal a problem, but they should not replace judgment about what successful support actually involves. The aim is to use measurement to inform the work, while checking whether the incentives attached to a measure are pulling behavior away from the original goal.