Delegating responsibility and still deciding every step leaves it unclear who is directing the work. The relationship between ruler and general allows this contradiction to be examined, provided executive autonomy is paired with limits and accountability.
The relationship between Sun Tzu and micromanagement starts with a problem: giving orders without knowing the conditions under which they must be carried out. Within Sun Tzu applied to leadership, the comparison makes it possible to examine how decisions are split between the board and executive management, not to place the CEO beyond any oversight.
The ruler and the general: what the passage questions
In chapter 3, sections 12 to 15 of Lionel Giles's translation, the ruler may harm the army by ordering movements it cannot carry out, managing it without understanding its conditions, or employing its officers without taking the circumstances into account. Section 17 links the general's capability to the ruler's lack of interference.
The commentary on the hierarchy of attack in chapter 3 places those warnings within the work. The difficulty described is intervention disconnected from the situation. Bringing it across to business does not prove that every board question is an intrusion, or that every executive decision must be accepted.
The analogy also does not establish legal powers. An ancient ruler is not the same as a board, a board of directors or a group of investors. Here, the proposal is to examine decisions within an already established division of responsibilities, without replacing it or drawing legal obligations from the treatise.
For this reading, we will call micromanagement the repeated replacement of delegated judgement with instructions about details, without clarifying who retains responsibility. Reviewing a deviation is different: it asks about its causes and the options available before deciding whether to intervene.
Agree on what is decided, what is consulted and what is reviewed
A clear delegation should describe the expected outcome and the leeway to achieve it. “Improve the service” leaves questions open. A more precise formulation identifies which problem must be solved, which resources are authorised and which conditions cannot be changed without a new decision.
As an organisational proposal, it is worth separating the team’s day-to-day decisions from those that change the commitment made. Assigning tasks within an approved scope can be delegated; changing that scope requires the agreed process. The limits must be known in advance, not appear later when a solution displeases the person supervising.
Reporting, consulting and seeking authorisation are not synonyms either. A manager may communicate a decision they were empowered to make. Turning every communication into a pending approval nullifies that leeway in practice. Conversely, presenting a change as mere information does not make it a decision that was previously delegated.
Review needs data that are proportionate to the question. To understand the status of a delivery, it may be relevant to check what is done, what is missing and what is blocking progress. Reviewing every message exchanged does not necessarily answer those questions better, and it adds work that also needs to be justified.
The executive must explain the difficulties, including those that challenge their plan. The five virtues of the business leader raises questions about that behaviour. Invoking autonomy to hide a deviation is incompatible with a delegation that depends on shared information.
The agreement must also explain what happens when there is a significant change. It may establish that a blocked dependency is reported as soon as it is detected, without waiting for the next review. This distinguishes a necessary alert from a routine query about every step. The aim is to resolve the exception, not to return all decisions to the board.
An example of oversight that does not replace the team
Imagine a fictional company deciding to renew its customer service system. The board approves the objective and the available resources. Executive management takes on coordination, while the team organises testing and transition. It is agreed that the existing service will remain operational until the new one has been confirmed as able to handle the expected enquiries.
During the work, a board member directly asks a developer to add a feature for a demonstration. Another tells the project lead not to change the scope. The developer receives incompatible instructions, and the timetable no longer matches the work actually requested.
The problem is not solved by asking the team for more initiative. First, it must be clarified whether that request changes the brief. In the example, the proposals are passed to the responsible person, who explains what they would require and which delivery would be pushed back. If the change is approved, the commitment is updated and communicated to those who must carry it out.
The board can keep asking about the results. The difference is that it stops introducing tasks through parallel channels while holding the executive accountable for a plan it has already altered. Management, in turn, cannot dismiss every proposal by saying it belongs to the technical detail: it must explain its impact on the agreed objective.
Suppose the tests reveal a limitation that delays the transition. The review compares the new evidence with the acceptance conditions and decides what must be checked before continuing. Changing the sequence for that reason is not the same as capriciously disobeying an order; it requires showing which assumption no longer holds.
Attention to context links with the six ground conditions in chapter 10. In this adaptation, the point is to ask what information the decision-maker needs. That question does not authorise the executive to blame the environment for all problems, nor to avoid reviewing their own choices.
If repeated delays appear without explanation, decisions are made outside the agreed scope, or information is incomplete, it is appropriate to review how the delegation is working. That review may ask for clarification, check results and redefine limits. Reducing a margin temporarily needs an identifiable reason, not a personal preference for controlling every action.
The same criterion applies between the executive and their managers. Care and discipline in talent management addresses how to maintain clear expectations without turning oversight into constant surveillance. Demanding autonomy from the board and then denying it to the team repeats the contradiction at another level.
Check whether each manager can explain what they decide for themselves, what they must consult on and how they report the outcome. If the answers differ, the division of responsibilities needs to be clarified before more autonomy is demanded.
View all articles in Leadership with Sun Tzu: how to review a team’s direction
