Read Amazon Unfiltered
Go deeper into the inside story of Amazon operations, pressure, performance, and the human cost of scale.
Learn MoreMetrics and Power
Metrics are supposed to clarify reality. At their best, they help leaders find problems, compare performance, reduce waste, and make better decisions. But inside many organizations, metrics become something more powerful and more dangerous. They become the reality leaders manage to, even when the numbers no longer describe the work honestly.
The first mistake is treating the metric as if it is the work itself. A productivity number is not the work. A defect rate is not the work. A utilization rate is not the work. A forecast is not the work. Each one is a representation of something happening inside the business, and every representation leaves something out.
This matters because people inside organizations quickly learn what the metric rewards. If leaders reward speed without understanding quality, people will protect speed. If leaders reward utilization without understanding fatigue, people will protect utilization. If leaders reward forecast accuracy without understanding bad inputs, people will defend the forecast. A broken metric does not merely report behavior. It creates behavior.
Corporate America loves clean numbers because clean numbers travel well. They fit into slides, dashboards, executive reviews, investor language, and performance summaries. They make messy work appear manageable. But the cleaner the number becomes, the more likely it is that important details have been stripped away.
A senior leader may see a trend line moving in the right direction and believe the system is improving. Someone closer to the work may know the improvement came from a temporary workaround, an exhausted team, a one-time intervention, or a definition change that made the number look better without fixing the problem. The metric says progress. The floor says pressure. Both can exist at the same time, but only one usually gets promoted upward.
When a metric misses, the first question often becomes who failed to execute. That question is not always wrong, but it is often incomplete. A miss may come from poor planning, flawed data, unrealistic assumptions, bad process design, unclear ownership, or a target that never matched reality. If leaders treat every miss as an execution problem, the system protects itself while workers absorb the blame.
That is how broken metrics become a form of power. They decide what is visible and what is invisible. They decide which explanations are acceptable and which sound like excuses. They decide who gets questioned and who gets protected. In a healthy organization, metrics should invite investigation. In an unhealthy one, metrics become a weapon pointed downward.
A flawed metric can move through an organization with surprising force. It can shape staffing levels, budgets, incentives, automation priorities, performance rankings, coaching conversations, and leadership reputations. Once it becomes accepted, people stop asking whether it is true and start asking how to improve it. That is the dangerous moment.
Improving a bad metric can make an organization worse. If the measure is too narrow, teams may optimize one part of the work while damaging another. If the measure ignores human cost, leaders may celebrate efficiency while increasing strain. If the measure treats exceptions as failures instead of signals, the organization may punish the very people who are exposing where the system needs repair.
The solution is not to abandon measurement. Large organizations need metrics. The solution is to treat metrics with humility. Leaders should ask what a number includes, what it excludes, who benefits from the definition, who is pressured by it, and whether the metric is still serving the work or simply protecting the system.
Good metrics create better questions. Broken metrics end the conversation too early. They make leadership feel informed while distancing it from reality. That is why the argument matters beyond one company, one warehouse, or one industry. In a world increasingly governed by dashboards, models, and AI, the quality of the metric becomes the quality of the decision.
The System Is the Boss is built around that warning. When organizations stop challenging their measures, they do not become more rational. They become more obedient to whatever the system has learned to count.
A metric is usually broken when people closest to the work stop believing it describes reality. That does not mean every complaint is correct or every target is unfair. It means leaders should pay attention when the people who understand the process say the number is creating distortion. If a metric causes teams to hide problems, chase cosmetic improvements, ignore important work, or punish honest exceptions, the metric is no longer serving the business. It is managing the business badly.
Another warning sign is when leaders can explain the number but cannot explain the work behind it. A director may know the rate, the target, and the trend, yet have no real understanding of the process changes, shortcuts, stress, or tradeoffs that produced the result. That gap is dangerous because it creates confidence without comprehension. The organization believes it is data-driven, but it may actually be dashboard-driven.
Good measurement should make reality harder to ignore. Broken measurement does the opposite. It gives leadership a clean story that can survive even when the work underneath it is telling a different truth. The fix begins when leaders treat metrics as evidence, not verdicts, and when they give frontline reality enough authority to challenge the report.
The practical test is simple. If improving the metric requires people to make the work less honest, less safe, less sustainable, or less useful to the customer, the organization is not improving. It is learning to satisfy the measurement system. That is not operational excellence. It is obedience dressed up as performance.
Go deeper into the inside story of Amazon operations, pressure, performance, and the human cost of scale.
Learn MoreExplore how AI, bad data, and broken metrics are reshaping authority and decision-making across modern work.
Learn MoreRead more essays on operations, AI, metrics, leadership, labor, and the future of work.
Articles