A KPI stops being useful when improving the number no longer means improving the business.

For example, increasing website traffic looks positive on a dashboard, but if conversions remain flat and customer acquisition costs keep rising, traffic alone isn’t telling the full story.

I think there are three signs that a KPI needs to be reconsidered: it no longer reflects the current business objective, teams start optimising the metric at the expense of the actual outcome, or it fails to explain what action should follow.

There’s also a difference between a KPI becoming irrelevant and a KPI being poorly defined. Sometimes the metric itself is still valuable, but it needs more context. Revenue, for instance, becomes much more informative when viewed alongside margins, retention, and acquisition costs.

A good KPI should help you make a better decision, not just report a number. If teams can hit the target without creating meaningful value, it’s probably time to rethink what they’re measuring.

I think a KPI stops being useful when teams continue tracking it even though it no longer helps them make a better decision.

For example, a team might focus on reducing customer support response time. The numbers improve, but if customers still need multiple follow-ups to resolve their issues, the KPI is telling only part of the story.

That’s why I’d look at whether the metric still reflects the outcome the business cares about, whether people are changing their decisions based on it, and whether improving the number is actually improving performance.

Retiring a KPI can be difficult when teams have spent years building reports and targets around it. But keeping a familiar metric just because it’s easy to track can be more misleading than replacing it.

A useful question to ask is: if this KPI improved by 20% tomorrow, would we know what business outcome actually got better? If the answer is unclear, it may be time to rethink the metric.