Why the Safest Decision in a Confused Company Is to Escalate

Consider a manager who receives a customer request they know how to resolve. The commercial impact is modest. The role description suggests that the decision sits within their authority. A similar issue has been handled locally before.

The manager still forwards it to a senior leader.

The last time they made the same kind of judgment, the answer was reversed after another executive objected. Nobody explained which boundary had been crossed. The customer was corrected, the manager absorbed the embarrassment and the organization moved on. This time, escalation is the safer choice.

From above, the behavior may look like weak confidence. From inside the system, it may be accurate risk management.

Companies often try to reduce escalation by telling managers to take more ownership. That message fails when the organization has taught people that formal authority is less reliable than senior reaction.

Authority is learned from consequences

An authority matrix can state who may decide. Employees learn the real model by observing what happens after decisions are made.

If a manager acts within the documented boundary and receives fair review, authority becomes credible. If senior leaders routinely reopen the choice, contradict one another or judge the decision using information that was unavailable at the time, the practical boundary moves upward.

People update quickly. They notice which executives must be comfortable, which customers trigger intervention and which kinds of mistakes are forgiven. They learn whether local judgment is genuinely expected or merely praised until it produces an unpopular outcome.

This creates an important distinction between nominal and usable authority. Nominal authority exists in the role description. Usable authority exists when a person can exercise judgment without unreasonable fear that the organization will redefine the boundary after seeing the result.

A manager may possess the first and lack the second.

Escalation then becomes a way to transfer not only the decision but the risk of interpretation. By asking a senior person, the manager obtains protection against invisible expectations. The company pays in speed and leadership attention, but the individual reduces professional exposure.

Confusion makes waiting rational

Several organizational conditions increase the safety of escalation.

The first is inconsistent precedent. Similar cases receive different answers depending on who is asked, how urgently the issue is framed or which stakeholder becomes involved. Employees cannot learn a stable principle, so they seek the current preference of the most consequential person.

The second is invisible consequence. Managers may know their financial limits but not understand which reputational, political or strategic factors make a case sensitive. A modest customer concession can become dangerous because the customer is important to a senior leader or because another function sees an implication nobody has shared.

The third is asymmetric review. A local decision that succeeds receives little attention. A decision that fails is examined in detail, often with hindsight. Escalation therefore has visible delay cost but lower personal downside. Acting has invisible speed value but potentially concentrated blame.

The fourth is unexplained reversal. Senior leaders may have good reasons to override a decision. When they do not explain the changed fact, violated threshold or broader consequence, the organization learns only that authority can be reclaimed unpredictably.

Under these conditions, asking managers to be bolder is asking them to ignore the evidence produced by the system.

Escalation volume is a diagnostic, not only a burden

Leaders understandably become frustrated by a queue of ordinary questions. Every escalation consumes attention and slows the company. But the queue contains information.

A pattern of escalations may reveal missing thresholds. It may show that two functions hold conflicting expectations. It may identify decisions that are formally delegated but behaviorally retained. It may expose a small number of customer, people or risk categories where the organization has never stated the actual boundary.

Treating the queue only as a manager-capability problem can make the signal disappear without repairing the cause. Employees may stop escalating openly and begin seeking informal sponsorship, delaying until the answer becomes obvious or making choices defensively within the narrowest possible interpretation. The senior queue shrinks while operating quality weakens.

A better review looks for recurrence. Which decisions repeatedly travel upward? Which managers escalate most, and what happened when they previously acted? Which senior leaders reverse delegated choices? Are the escalations concentrated around unclear policy, cross-functional conflict, exceptional customers or politically sensitive issues?

The purpose is not to shame the manager or the executive. It is to discover where the organization's real authority differs from its declared authority.

The common cure can deepen the problem

A leadership workshop on empowerment may improve language without changing the risk environment. Managers return with encouragement to decide, but the first controversial case still attracts intervention from above. The speech says ownership; the consequence says permission.

Similarly, raising approval limits does little if the meaningful constraint is not financial. A manager may be able to approve the commercial value yet remain uncertain about precedent, customer sensitivity or another function's informal veto.

Another common response is to refuse escalations. Senior leaders tell managers to "come back with a recommendation" or simply decide themselves. This can be useful coaching when the boundary is clear. It can be unfair when the manager is asking the organization to resolve a conflict in authority rather than supply an answer.

Leaders should distinguish between three kinds of escalation. One reflects a capability gap: the person needs support to make a decision they genuinely own. Another reflects a consequence threshold: the issue correctly belongs at a higher level. The third reflects design confusion: the manager cannot know whether the decision is theirs because authority, precedent or veto power is inconsistent.

These situations require different responses. Treating all escalation as weakness obscures the operating problem.

Make local action professionally safe

Reducing unnecessary escalation requires more than moving boxes in a decision matrix. The organization must make the boundary understandable and the review fair.

Decision thresholds should describe consequence, not only amount. A manager may act within a commercial range unless the choice establishes a new precedent, affects a named strategic relationship, creates a legal or safety boundary, or transfers material cost to another function. The exact criteria will vary, but the manager should be able to recognize when local judgment ends.

Precedent should also become accessible. When a recurring case is decided, preserve the rationale. Employees need to know not merely what happened but which fact made the answer appropriate. Otherwise, they imitate outcomes without understanding conditions or escalate every variation.

Overrides require discipline. A senior leader should state whether the original decision was unreasonable, whether new information emerged, whether an unstated boundary was discovered or whether the leader is consciously making an exception. This explanation protects the authority model. It allows the organization to learn without treating every correction as proof that local judgment was unsafe.

Review should separate a reasonable decision with an unfavorable result from negligence, threshold violation or concealment. If all bad outcomes are treated as bad judgment, managers will optimize for protection rather than initiative. Authority always includes the possibility of an imperfect decision. The control is not perfection; it is clarity of boundary, quality of reasoning and learning after consequence.

Senior leaders must stop answering the same question privately

Escalation remains attractive when access to senior leaders is the fastest route to certainty.

A leader may resolve an issue in seconds, but repeated private answers prevent the organization from building a public rule. The same question returns through another manager because the first answer did not become precedent, authority or process. Senior responsiveness becomes infrastructure.

A stronger response closes both the case and the design gap. The leader can ask: Why did this need to reach me? Which boundary was unclear? Who should own the next similar case? What should the organization remember from this decision?

Not every escalation deserves redesign. Some are truly rare and consequential. But when the same class of issue repeatedly reaches the same level, the company is paying senior attention to preserve ambiguity.

Before demanding courage, examine the risk model

A confident manager does not need permission for every choice. A prudent manager also does not ignore evidence that local decisions are punished unpredictably.

The challenge is to create an environment in which good judgment is both expected and usable. That requires authority connected to consequence, precedents that can be understood, genuine vetoes that are explicit and leaders who correct decisions without casually destroying the right to decide.

When those conditions exist, escalation becomes a deliberate tool. Managers seek help because the issue crosses a known boundary, not because the boundary may move after they act.

Before asking why people keep escalating, leaders should examine what the organization has taught them about the safety of deciding.

In a confused company, escalation is often not a failure of ownership. It is ownership of personal risk performed with uncomfortable precision.