Measuring operational performance without vanity metrics

Most organisations measure what is easy to count rather than what matters. Activity metrics, the number of meetings held, the volume of emails sent, the percentage of tasks marked complete, are easy to produce and easy to report. They are also easy to improve without improving anything that matters.

The difference between an activity metric and an outcome metric ¶

An activity metric measures something the organisation does. An outcome metric measures something that changes as a result. The distinction sounds simple but it is easy to blur in practice. 'Number of client calls completed' is an activity metric. 'Percentage of clients who renewed' is an outcome metric. Both are easy to track; only one tells you whether the calls are working.

How to choose metrics that will not be gamed ¶

Any metric that is reported upward will eventually be optimised for rather than used as a diagnostic tool. That is not a character flaw; it is a rational response to being measured. The way to slow that process is to choose metrics that are hard to improve without actually improving the underlying thing.

Cycle time is a good example. The time from a request being received to it being resolved is hard to improve without actually making the process faster. Compare that to 'percentage of requests acknowledged within 24 hours', which can be improved by sending an automated acknowledgement that does nothing.

• Choose metrics that require real change to improve
• Pair each metric with a qualitative check every quarter
• Review the metric set annually. What you measure should change as the organisation changes
• Limit the dashboard to five metrics; more than that nothing gets attention

What to do when the metric is telling you something you do not want to hear ¶

A metric that consistently shows underperformance is doing its job. The temptation is to adjust the target, change the measurement method, or add context that explains the gap. Sometimes that context is legitimate. More often it is a way of avoiding the question the metric is asking.

The more useful response is to treat the underperforming metric as a diagnostic prompt. What would have to be true for this number to improve? That question usually points directly at the operational problem that needs attention.

A small set of metrics that people trust and refer to regularly is worth more than a comprehensive dashboard that no one believes. Start with three and add only when you have a specific reason to.