In 1898, Frederick Winslow Taylor walked onto the floor of a Bethlehem Steel plant with a stopwatch and a theory in his head. The theory was simple: measure everything, and efficiency will follow.

Workers, the cheeky scamps, had other ideas.

The moment they understood they were being timed, they slowed down. Not out of laziness, but out of logic. A slower baseline meant a manageable quota.

A manageable quota meant a bearable life.

Taylor had built a measurement system. The workers had immediately figured out how to break it by following it. Go workers.

Goodhart gets one sentence

A British economist named Charles Goodhart formalised this problem about seventy years later:

When a measure becomes a target, it ceases to be a good measure.

We named a law after him. As we do.

It’s obvious in retrospect and invisible in practice.

Everyone agrees with it in a conference room. Nobody applies it on a Monday morning when the dashboard is due.

Four stars, for the wrong reasons

Walk into almost any restaurant obsessing over its TripAdvisor score, and you’ll see Goodhart’s law running the kitchen.

The food is fine. The service is okay.

At the end of the meal, the waiter will ask, with slightly pleading eye contact, whether “you’d mind leaving a review?”

The restaurant stopped optimising for the meal and started optimising for the score. The score and the meal used to point at the same thing.

They don’t anymore.

Nobody set out to break anything

Here’s what Taylor, Goodhart, and the TripAdvisor waiter all have in common:

None of them set out to break anything.

They were responding rationally to the system they were actually in.

This isn’t a story about bad people making bad decisions. It’s a story about normal people making completely understandable ones (normal).

Marketing’s version of the stopwatch

Which brings us to marketing, where this pattern has been industrialised.

MQLs that bear no resemblance to actual buying intent.

Follower counts that correlate with nothing.

Email open rates that, post-Apple Mail Privacy Protection, are essentially numbers we all agreed to keep reporting.

The dashboard looked great. The pipeline did not.

Ask why, and you get to the real problem.

It isn’t dishonesty. It’s that ambiguity is genuinely uncomfortable, and a number, any number, feels like solid ground.

The nearby answer

Brand health is hard to measure. Mental availability is harder.

The slow accumulation of trust across thousands of small interactions is essentially unmeasurable in any given quarter.

But impressions are right there. Clicks update hourly. Cost per lead has two numbers and a division sign.

It feels like an answer.

The near answer crowds out the right answer. Not because anyone is foolish, but because uncertainty isn’t shown on a dashboard.

Why smart people keep doing it

There’s a concept in psychology called cognitive ease.

When something is simple to process, it feels true.

When something is difficult to process, it feels suspicious, or at least inconvenient.

Metrics exploit this without trying to. A clean number in a clean cell produces cognitive ease.

The messy, unmeasurable truth about whether your brand is genuinely growing in people’s minds produces the opposite.

It feels like an excuse. Even when it’s the answer.

This is why smart people keep doing it. Ambiguity feels like a personal failing. A number in a spreadsheet feels like evidence.

Standing at the edge of the rule

The workers on Taylor’s factory floor weren’t saboteurs. They were doing what people always do when faced with an arbitrary measure of their worth: they found the edge of the rule and stood just inside it.

Marketing does the same.

Holding it honestly

None of this means abandoning measurement. It means holding it more honestly.

A click-through rate tells you something. It doesn’t tell you everything.

The gap between those two statements is where judgment lives.

The best marketers tend to carry both things at once.

Genuine curiosity about the numbers alongside genuine scepticism about what the numbers are actually capturing. That’s not a contradiction; that’s just pragmatic sense.

The restaurant across town

The TripAdvisor restaurant is still out there. Four stars. Slightly too much eye contact. The food, fine.

Somewhere across town, there’s a place with three and a half stars and a six-month waiting list. The owner hasn’t worked out the review solicitation game. She’s been too busy making food worth eating.

Goodhart would have loved her.

This article was written with the assistance of AI.