Imagine a CEO walking into a leadership meeting and saying, “Let’s skip the dashboard today.”
Half the room would assume something is broken. The rest would wonder if the CEO has lost faith in data. For years, we have treated dashboards as proof of seriousness. We build them before the product is ready. We debate KPIs before we have a strategy. Every decision starts with a chart. At some point, dashboards shifted from being tools to being signals of good management.
But what if we’ve misunderstood their purpose? What if the real purpose of metrics is not to become more important over time, but progressively less?
That sounds close to heresy in product circles. But it matches what I see in the healthiest organizations. Metrics are not the goal. They are scaffolding. They help us learn. If they are still at the center of every conversation years later, it is a sign the organization has not learned as much as it thinks.
Metrics help us make decisions when things are unclear. They are not executive reports or team scoreboards. They are tools to reduce uncertainty, test assumptions, and check if our mental model of the business is improving.
That distinction changes everything.
When a product is new, uncertainty is the default. We do not know what customers value. We do not know if onboarding works, if pricing is right, or if the strategy addresses the real problem. At this stage, metrics are essential. They are headlights on a dark road. Each experiment teaches us something. Each measurement either confirms or challenges an assumption.
But learning should eventually change the organization itself.
Think about an experienced physician. They don’t immediately order every possible laboratory test before speaking to the patient. Years of experience have given them a mental model of how the human body behaves. The tests are still important, but they are no longer the starting point. They confirm or challenge the doctor’s understanding. They don’t replace it.
The same should be true for product organizations.
The most mature teams can often predict what the dashboard will show before they look. Someone spots friction from a redesign. Someone else notes that customer interviews have turned more negative. Another sees support tickets about the same issue creeping up. When the activation metric drops, nobody is surprised. The dashboard just confirms what the team already knows.
That is organizational learning.
Many companies never reach this stage. They become attached to dashboards instead of using them to build understanding. Every Monday looks the same. Twenty graphs on a big screen. Green means good. Red means bad. Someone explains why one KPI increased by 1.8% and another decreased by 0.6%. People nod, a few action items are noted, and the meeting ends.
Very few people ask the question that actually matters: “Why do we believe this number represents success in the first place?”
When that question fades, dashboards stop supporting thinking and start replacing it.
Ironically, the further executives move away from customers, products and engineering teams, the more they tend to rely on metrics. Numbers become a compressed version of reality. They fit neatly into quarterly business reviews. They are easy to compare and easy to summarize. But they are still abstractions. Eventually, leadership begins managing the abstraction rather than the system it was meant to represent.
We’ve all seen how that story ends.
Support teams chase ticket resolution time instead of solving real problems. Engineering teams count story points rather than deliver outcomes. Editorial teams watch clicks instead of building trust. Marketing celebrates traffic while conversion slips. Nobody is acting in bad faith. People respond to the incentives the system creates.
Every metric sends the same message: “This is what success looks like.”
This makes metrics one of the most powerful management tools. They do not just observe behavior. They shape it. They influence priorities, conversations, promotions, budgets, and daily decisions. In many cases, they reshape the organization more than the organization reshapes them.
This is exactly why strategy should always come before measurement.
The first question is not, ‘What can we measure?’ It is, ‘What kind of company are we trying to build?’ Only then do we build a model of how that company creates value. Once we have the model, choosing metrics is easier. They are not random numbers from the analytics tool. They are signals that show if our understanding still holds.
Many organizations reverse the order. They start with what is easy to measure and let those measurements define the strategy. Over time, they become skilled at optimizing things that were never important.
This is risky because the most valuable things in an organization are usually the hardest to measure.
Trust between teams. Learning from failures. Editorial credibility. Customer confidence. Leadership quality. The ability to adapt.
These things do not fit into a dashboard. They do not produce a weekly trend line. Yet they often decide whether a company succeeds or quietly declines over years.
This is why I think the strongest organizations are not really data-driven. They are model-driven.
Their advantage is not collecting more numbers. It is building a shared understanding of how the business creates value. Metrics still matter, but their role shifts. They do not discover reality. They confirm or challenge the organization’s understanding of it.
But there is one more ingredient. None of this works without trust.
Dashboards do not create trust. Poorly designed metrics often destroy it. If every KPI is a performance score, people stop experimenting. They stop admitting uncertainty. They learn to optimize numbers rather than improve systems.
The best organizations use metrics differently. They do not ask, ‘Who caused this number?’ They ask, ‘What does this number tell us about the system?’ The conversation shifts from blame to learning, from defending results to improving the model together.
This is the paradox we often miss. Metrics were never meant to be the center of management. Their purpose is to help an organization build something more valuable.
A shared understanding of how the business works, and the mutual trust to question that understanding whenever reality disagrees.
Dashboards do not build great companies. People who trust each other enough to learn together do.