Designing How to “Use” Experts

Assumed Reader State (Before)

Are you delegating decisions to experts in legal, accounting, tax, or security matters, allowing choices to solidify in an atmosphere of “if the expert says so, we have no choice”? At the same time, you might feel uneasy as decisions slow down, business flexibility decreases, and accountability becomes ambiguous. The result is a dilemma: despite having more experts, the quality of management decisions isn’t improving.

Agenda Setting (What is the decision?)

The decision at hand is how to design experts not as “decision-makers” but as “devices to advance management decisions.” This is crucial for management because misusing experts leads to the outsourcing of judgment, weakens the leader’s design capability, and causes a loss of governance agency. The starting point is understanding that this is not a problem with the experts’ competence but a design flaw in how they are used.

Conclusion Summary (Upfront)

Experts are not “entities that provide the correct answer.” An expert is a support device for clarifying the options and conditions of a management decision. The correct design principle is to have them produce materials for comparison and evaluation without handing over decision-making authority. This is not about undervaluing experts; it’s about positioning their insights where they can deliver the most value.

Clarifying Premises (Facts & Constraints)

The business objective is to keep moving forward with management decisions even in an uncertain environment. Three key constraints exist here. First, no one other than the management bears ultimate responsibility. Second, experts tend towards local optimization (optimizing within their own field). Third, expert advice depends on underlying assumptions. Given these premises, letting experts “decide” is structurally irrational.

The Structure That Creates Expert Dependency

In many organizations, the following structure becomes entrenched. Management asks, “Is there a risk?” and the expert answers, “There is a risk.” As a result, only the most conservative option remains, and decision-making is effectively outsourced to the expert. In this scenario, risk management can lead to the loss of business opportunities.

The Proper Way to Use Experts

In organizations that correctly “use” their experts, the way questions are framed is decisively different. Instead of asking “Can we do this?” (Yes/No), they ask, “What are the conditions for making this viable?” The role division is clear: experts present multiple options with their conditions, trade-offs, and comparative risk levels, while the final choice always rests with management.

Division of Labor as a Management Decision

For effective decision-making, the following division of labor (role allocation) is essential.

• Management’s Role: Determine the purpose and priorities. Decide the acceptable risk level. Make conditional choices.
• Expert’s Role: Organize the constraints. Translate viable patterns. Structure the decision-making materials.

The moment this agency (who decides) is reversed, the organization’s governance begins to collapse.

Common Failure Patterns

There are three main failure patterns organizations often fall into regarding their relationship with experts.

• Complete Outsourcing: Letting the expert decide.
• Title Dependency: The weight of an opinion is determined by the person or their position.
• Responsibility Illusion: Believing the expert will take responsibility.

In all these cases, management is under the illusion of “using” the experts while, in reality, being used by them.

After (The Manager After Reading)

A manager who has internalized proper governance and risk management thinking can use experts as design support devices, not decision-making devices. The quality of questions posed to experts changes, and the agency for management decisions is never lost. As a result, experts cease to be constraints on management and become powerful leverage for increasing the precision and speed of decisions in a highly uncertain environment.