Because in many organisations, the metrics and KPIs end up choosing the destination.

I was discussing this recently with Ron Mo at BI Worldwide following his observations on AI adoption at Uber and Amazon.  In both cases, leaders introduced usage targets and leaderboards. Adoption increased. So did costs.

The problem wasn’t the technology. The problem was that the measure became the goal. Engineers did exactly what the system encouraged them to do. Measurement drove behaviour.

It wasn’t faulty choices the engineers made to drive token usage to the moon. Nor was it the metric; it’s a useful consideration in adoption, but not the whole story.

The problem was that no one stopped to ask what outcome they were actually trying to achieve.

If you don’t start with the outcome, the metric will choose one for you. Too often organisations start with:

“What should we measure?”

Rather than:

“What are we trying to achieve?”

Once the outcome is clear, we can define the behaviours that support it and the measures that indicate progress. Get that sequence wrong and the metric becomes the destination.

Measurement drives behaviour. But outcomes should drive measurement.

Metrics are important. But they were never meant to drive the bus. Their job is to tell us whether we’re heading in the right direction.

This writing is inspired by my earlier articles, Pragmatic Objectives and Where Discipline Begins: Clear Priorities.

Image: Oxyman – Own work