Every metric you publish, someone will try to optimize. Not out of malice — out of rationality. If the organization rewards a number, the rational actor optimizes that number. This is not a character flaw. It's a design flaw.
When a metric gets gamed, the problem isn't the person gaming it. The problem is that you defined the wrong behavior and got exactly what you asked for.
Why Gaming Is Rational
Consider a team whose annual performance is based on hitting three OKRs. The OKRs are set in January. By March, it's clear one was built on faulty assumptions — the market shifted, a key partner walked. The honest move — "we need to update this" — requires political capital and an admission of error. The dishonest move — "we're still on track" — requires only silence.
Most people aren't villains. They're people operating in systems that punish honesty and reward the appearance of success. The gaming isn't the disease. It's a symptom of a system that makes gaming the rational choice.
The Underlying Principle
Goodhart's Law is not a slogan. When a measure becomes a target, it starts decaying as a measure. Treat that as a design constraint: pair targets with outcomes, guardrails, and human review. When a metric gets gamed, don't start by punishing the gamer. Ask what behavior the system rewarded.
