Every perverse incentive starts as a metric someone was proud of.
Stage 1: a real problem. Stage 2: a measure that proxies the problem. Stage 3: the measure becomes the target. Stage 4: behavior optimizes the measure, not the problem. Stage 5: the original problem returns — usually worse.
Hanoi's rat bounty paid per tail. Citizens bred rats. Wells Fargo paid per account opened. Employees opened 3.5 million fake accounts. The pattern is the same — the measure decoupled from the goal.
The fix is not better measurement. The fix is staying close enough to the original goal that you can re-anchor the measure when it starts to drift.
The Structural Flip
Treat every metric as temporary. The day it stops moving with the goal is the day you change it.
