Alex: A factory measures output per hour. More units, bigger bonus. So workers run the machines faster than they should, and quality collapses.
Jordan: A school measures test scores. Higher scores, more funding. So teachers teach to the test, and real learning dies.
Alex: Wells Fargo measured new accounts opened. More accounts, bigger bonuses. So employees opened millions of fake accounts.
Jordan: Every one of these started with a reasonable metric. And every one ended with the metric eating the goal.
Alex: Welcome to Incentives, Decoded. I am Alex.
Jordan: And I am Jordan. This is chapter thirteen, Goodhart's Law. Last time, you learned the principal-agent problem, why the people you hire do not want what you want. Today, we learn what happens when you try to fix that with measurement. Spoiler: the measurement becomes the problem.
Alex: Goodhart's Law says when a measure becomes a target, it ceases to be a good measure. People do not improve the thing. They improve the number.
Jordan: This is the law that explains why every metric you have ever been judged by felt like a game. Because it was.
Alex: Let us start with the most expensive example. Wells Fargo, twenty sixteen. The bank set aggressive sales targets. Eight products per customer. Every employee, every branch, every day.
Jordan: The metric was new accounts opened. The goal was deeper customer relationships. Sounds reasonable. Measure the thing you want, reward the people who deliver it.
Alex: But the target was impossible through honest means. So employees invented customers. They opened three point five million unauthorized accounts. Fake emails, fake PINs, real fees charged to real people.
Jordan: The bank paid three billion dollars in fines. The CEO resigned. Thousands of low-level employees were fired, the same employees the impossible target had cornered.
Alex: Nobody set out to commit fraud. They set out to hit a number. And when the number becomes the job, the job becomes the number. The goal, actual customer relationships, was not just missed. It was destroyed.
Jordan: Here is why this always happens. A metric is a proxy. It stands in for something you actually care about but cannot measure directly.
Alex: Test scores stand in for learning. Units per hour stand in for good work. New accounts stand in for customer trust. The proxy is always simpler than the real thing. And simpler means gameable.
Jordan: The moment you reward the proxy, smart people stop pursuing the goal and start pursuing the proxy. Not because they are dishonest. Because you told them to. You said this number is what matters. They believed you.
Alex: And the cruel twist? The better the metric, the faster it dies. A vague metric is hard to game. A precise metric is a precise instruction for how to cheat without technically cheating.
Jordan: This is why the factory worker runs the machine too fast. She is not a bad employee. She is a perfect employee, optimized for exactly what you measured.
Alex: Now turn it around. What are you measuring in your own life?
Jordan: Your step count. Your followers. Your revenue. Your weight. Your inbox zero streak. Pick one. Now ask: am I improving the thing, or gaming the number?
Alex: The dieter who drinks water before the weigh-in. The student who memorizes instead of understanding. The manager who hits the quarterly target by borrowing from next quarter. All optimizing the metric. All missing the point.
Jordan: Social media is the purest example. The metric is likes. The goal was connection. Now everyone performs connection for the metric, and actual connection is rarer than ever. You know this feeling. You have posted for the number.
Alex: The scariest part? You do this to yourself. No boss required. You pick a metric, you chase it, and slowly the metric replaces the dream that created it.
Jordan: So here is the move. Measure, but measure like someone who knows Goodhart's Law.
Alex: One. Measure what matters, not what is easy. If the real goal is customer love, do not measure handle time. Measure callbacks, complaints, the things that are harder to fake. Easy metrics are easy to game.
Jordan: Two. Use multiple metrics that pull in different directions. Speed and satisfaction. Quantity and quality. A single metric is a single point of failure. Three metrics form a triangle that is much harder to cheat.
Alex: Three. Rotate the metrics, or add one nobody can easily game, like a spot check. If people do not know exactly how they are scored, they cannot optimize the score. They have to actually do the job. And the meta-move? Judge by judgment, not just numbers. Talk to customers. Read the work. Look with your eyes. The best measure of quality is still a human being paying attention.
Jordan: Never confuse the scoreboard with the game.
Alex: The number is not the thing.
Jordan: Next time, in chapter fourteen, The Cobra Effect, we see what happens when incentives do not just get gamed, they backfire completely. Pay for dead cobras, and people will breed cobras. The real story is even stranger.
Alex: Thanks for listening to Incentives, Decoded, chapter thirteen. Watch your metrics.