Goodhart's Law (When a Measure Becomes a Target)

Charles Goodhart's principle that when a measure becomes a target, it ceases to be a good measure. Originally about monetary policy (when central banks target a specific money supply indicator, the relationship between that indicator and economic outcomes breaks down), it applies universally: any metric used as a goal will be optimized at the expense of what it was designed to represent. Hit your revenue target by pulling forward deals. Hit your response time target by giving quick but poor answers. Hit your publication target by splitting one paper into three. The measure optimizes itself; the underlying reality drifts.

When to use it

When KPIs are being achieved but business outcomes aren't improving; when incentive structures are producing gaming behavior; when designing performance management, OKR, or accountability systems; when a metric that used to be useful seems to have lost its predictive power.

How it can help

Design measurement systems with Goodhart's Law as a design constraint, not a surprise. Practices: (1) Use metrics as diagnostics, not targets, when possible. (2) When metrics must be targets, use a balanced set that's harder to game simultaneously. (3) Pair quantitative metrics with qualitative assessment. (4) Rotate metrics periodically to prevent gaming infrastructure. (5) Monitor for the tell-tale sign: metrics improving while the underlying reality stagnates or declines. The deepest insight: measure what you value, but don't let the measure replace the value.