What the Leaderboard Does Not Tell You

Competitive surfing is scored by judges on wave selection, difficulty, innovation and flow. It is a good system and it still produces the same complaint every year: the scoring rewards a particular style, so athletes optimise for that style, and the sport bends around the scoreboard.

Every measurement system does this. Yours is doing it right now.

Goodhart's law has a surf equivalent

The formal version is that when a measure becomes a target, it stops being a good measure. The practical version is that your team is smart, motivated, and pointed at whatever number is on the wall, and they will move that number whether or not moving it helps.

Point a team at signups and they will find channels that produce signups from people who never return. Point them at time in app and they will make the app harder to leave, which users experience as friction. Point them at ticket close rate and tickets close faster, sometimes without the problem being fixed.

None of this is cynicism. It is what happens when competent people optimise honestly against an incomplete target.

The three questions that separate real metrics from theatre

Would this number go up if the business were dying? Total registered users only goes up. Total content created only goes up. Cumulative anything is a ratchet that cannot report bad news, which makes it useless as a steering instrument.

Does moving this number require someone to get value? Signups do not. Page views do not. A second week of use does. A second purchase does. The metric should be downstream of the customer getting what they came for, because that is the only thing that eventually produces money.

Can I act on it this week? Monthly recurring revenue is real, honest and almost impossible to act on directly, because it is the sum of a dozen upstream behaviours. You need it on the wall, but you steer with the thing you can influence in seven days.

Most dashboards fail at least two of the three, and the failure is usually invisible because the numbers are technically accurate.

The number under the number

Here is where competitive surfing gets useful again. Wave selection is a judged criterion, and it is the one that separates the top of the field. Not turns, not airs. Which wave you chose.

The startup equivalent is the composition of your growth, and it is almost always hiding under a headline number that looks fine.

Two companies both add a thousand users this month. The first added them across four channels with a forty percent second week return. The second added nine hundred from one viral post with a four percent return. The headline metric is identical. One of these companies has a business.

So for every headline number, ask two follow ups: where did it come from, and what did those people do next. Source and retention. The headline tells you the size of the wave. Source and retention tell you whether you actually rode it.

What to put on the wall

Keep it to four, and make them a chain rather than a collection. A chain tells you where the problem is. A collection just tells you that there is one.

• One reach number. People who encountered you. Honest about the top of the funnel.
• One activation number. People who reached the moment of value. Defined precisely, ideally as a specific action rather than a signup.
• One retention number. People still there in week four. This is the one that decides whether the company is real.
• One money number. Revenue or the closest honest proxy.

Four numbers, each one dependent on the last. When something goes wrong you can see which link broke, which turns a vague panic into a specific problem.

Next to each, write the one thing you can do this week to move it. If you cannot write that sentence, the metric belongs in a monthly report, not on the wall.

The practical audit

1. List every metric currently on a dashboard. All of them, including the ones nobody looks at.
2. Run the three questions on each. Could it go up while dying? Does it require value delivery? Can you act on it weekly?
3. Delete anything failing two. Genuinely delete it. A dashboard nobody trusts is worse than no dashboard.
4. Build the four link chain. Reach, activation, retention, money.
5. Define activation precisely. Argue about this properly. It is the definition that matters most and the one teams handwave.
6. Add source and retention splits to your headline growth number. This is usually where the unpleasant surprise lives.
7. Review quarterly for gaming. Ask directly: how could someone move this number without helping a customer? Then check whether anyone is.

Takeaways

• Your team will optimise for whatever is on the wall. Choose it as if that is guaranteed, because it is.
• A metric that cannot report bad news cannot steer anything. Cumulative counts are ratchets.
• The headline number hides the real story. Always split by source and by what those people did next.

Audit your dashboard this week and delete the metrics that fail two of the three questions. Then go chase a high score somewhere the leaderboard is honest, because it only measures one thing.