Analytics
Metrics vs. decision systems
A metric without a decision owner and an action framework is just a number on a screen.
Most companies have more metrics than they can use. Dashboards are full of charts. Reports are generated weekly. KPIs are reviewed in meetings. And yet, decisions are still made on instinct. The gap between the dashboard and the decision is the gap where most analytics budgets go to die.
The gap between having metrics and having decision systems is organizational, not technical. Decision systems require three things that most analytics setups lack: a clear decision owner for each metric, an action framework that specifies what to do when a metric moves, and a feedback loop that measures whether the action improved the outcome. Without those three, a metric is decoration. With them, a metric is a decision.
The decision owner is the part most often missing. A metric without an owner is a number that moves and triggers nothing, because no one is responsible for moving it back. The question to ask of any metric on a dashboard is: who changes behavior when this moves, and what do they do. If you can't answer both, the metric is reporting, not decision infrastructure.
The action framework is the second missing piece. Most dashboards show the metric and stop. A decision system shows the metric and the response: if this drops below X, we do Y. Without the framework, every metric movement becomes a fresh debate about what to do, which is slow, political, and usually resolved by whoever speaks last. The framework pre-decides the response, which is the whole point of having the data.
The feedback loop is the one that makes the system improve. You act on a metric. Did the action work? Most companies never close the loop, because the metric moved, the action was taken, and everyone moved on to the next fire. Without the loop, the team never learns whether its interventions actually work, and so it keeps intervening the same way, with the same results, indefinitely.
There's a reason this is hard. Dashboards are a tooling problem, and tooling problems get funded and shipped. Decision systems are an organizational problem, and organizational problems require accountability, which is harder to mandate than a dashboard. The company that builds the dashboard looks productive. The company that builds the decision system looks slow, because it's spending time on the unsexy question of who decides what when a number moves.
The diagnostic we use is simple. Pick the three metrics that most influence the business. For each: name the owner, name the action thresholds, and describe the last time the metric moved and what happened. If any of the three answers is missing, you have a dashboard, not a decision system. Most companies fail this test on all three metrics.
There's a cultural dimension worth naming. Decision systems require the discipline to actually change behavior based on what the data says, which means overruling the senior person's instinct when the data disagrees. That's politically hard. The companies with real decision systems have leaders who have explicitly given the data that authority. The companies without it have dashboards that get overruled whenever the data is inconvenient, which is the moment the data mattered most.
Building decision systems is harder than building dashboards. It requires cross-functional alignment, clear accountability, and the discipline to actually change behavior based on what the data says. But it's the only way to turn analytics investment into better business outcomes. The dashboard is the easy part. The decision is the work.