Dashboards are built in the hope that, on seeing the numbers, people will act better. But there is an uncomfortable truth in almost every company: most dashboards are viewed, commented on and then forgotten, changing nothing. The goal of data was never to decorate screens — it was to change decisions and behaviors. And the difference between a dashboard that achieves this and one that does not is not in the technology. It is in the design of the metric itself.
The problem: metrics that inform but do not move
There is a huge distance between knowing a number and acting on it. A dashboard can show perfectly that the satisfaction rate dropped, and still no one does anything — because it is not clear who is responsible, what to do, or whether that is good or bad. The metric informed, but did not move. Designing metrics that change behaviors means ensuring every number comes with an answer to "so what?" and "who?".
Principle 1: a metric with no owner moves no one
The first mistake is showing numbers that belong to no one. If an indicator worsens and there is no person or team that feels it is their responsibility, the number floats without consequence. Metrics that change behaviors have a clear owner — someone who looks at them and feels they have to act. Without that responsibility, the dashboard is a show everyone watches and no one stars in.
Principle 2: a number with no comparison says nothing
"Sales: 1.2 million" is a dead fact. Compared to the target, to last month, to the same period last year, it becomes a story: we are beating the goal, we are falling, we are accelerating. It is the comparison that gives meaning and triggers action. A dashboard full of numbers with no reference is a wall of facts nobody knows whether are good or bad news.
Principle 3: show the action, not just the state
The most powerful metrics do not only describe where we are — they point to what to do. Instead of "we have 200 customers at risk", an actionable metric shows "these 200 customers at risk, sorted by value, to contact this week". The first generates worry; the second generates a to-do list. The closer the dashboard is to concrete action, the more likely the action is to happen.
Principle 4: beware of what you incentivize
What you measure shapes behaviors — sometimes in unexpected ways. If you measure a call center only by the number of calls answered, agents will rush the calls and quality drops. The metric changed behavior, but in the wrong direction. Designing metrics that change behaviors requires anticipating how people will react to being measured by them, and choosing indicators that pull toward the right behavior, not the easy number.
A concrete case
A customer support team had a beautiful dashboard with dozens of indicators nobody used to decide. They reworked it around a single question: "which customers need attention today?". Instead of showing averages and totals, the new dashboard listed the concrete cases requiring action, with an assigned owner and a comparison to what was normal. The change was immediate: the team stopped "looking at the numbers" in the weekly meeting and started acting on cases every day. Response time improved not because technology was added, but because the metric started pointing to action instead of just describing the state. The same data, designed differently, changed the team's behavior.
Less is more: the discipline of focus
A dashboard with forty indicators dilutes attention and moves no one. One with three right indicators, well designed, with owner and comparison, changes how people work. The temptation is to show everything you can measure; the discipline is to show only what leads to action. Every extra indicator that does not change a decision is stealing attention from those that do.
In practice
Look at your most important dashboard and test each number: who is the owner, what does it compare to, and what action does it trigger? The indicators that do not pass this test are decoration — and they are hiding the ones that matter. A good dashboard is not the one that shows the most data, it is the one that changes the most decisions. Which of your indicators actually changed a behavior this month?
