Under the Banner of “Self-Running”: What to Delegate and What to Protect
“Self-running organization” — this is a keyword that many managers yearn for and also struggle with. The ideal is clear: break away from centralized decision-making, delegate authority to the front lines, and gain speed and creativity. However, this act of “delegation” harbors a quiet danger. Authority, once given, is not easily taken back. By fixing roles and expectations to people, an organization can become rigid and lose sight of its original purpose.
KDDI’s initiative to transform its IT department into a “self-running organization” is an excellent example of confronting this challenge head-on. The case of the company, introduced in DIAMOND Harvard Business Review, is not merely a success story but a valuable record of the “experiment of delegation.” From this, we can decipher the design philosophy for delegating authority in a “reversible” manner.
The Greatest Risk of Delegation: The “Personalization” of Roles
When aiming for a “self-running organization,” the first mistake many companies make lies in the very idea of “granting authority to a person.” Giving budget approval authority to the excellent Mr. A. Delegating final interview authority for hiring to Team B. This seems like a rational decision at first glance. However, reversibility is lost here. This is because authority becomes tied to “the person of Mr. A” or “the proper noun of Team B.”
What is noteworthy in KDDI’s case is the discernible conscious approach to this point. Reports mention “redefining the value of the IT department,” suggesting the core of the transformation was the department’s role itself and its performance metrics (likely internal customer satisfaction, development speed, etc.). This is an approach focused not on “specific individuals” but on “the department’s work and outcomes.”
The first principle for making delegation a reversible experiment is to “focus on the work and scope of decision-making, not the person.” If delegating budget authority, clearly define “for which scope of work and up to what amount discretion is granted.” This should be designed as a “temporary function” attached to the work of that position, not because Mr. A is there. From the outset, it’s necessary to build in the possibility that if the person changes, the function transfers to someone else, or is once reclaimed and redesigned.
Decide the Evaluation Period and Observation Points for “Self-Running” in Advance
Simply saying “Go ahead, run on your own” and handing over authority ends up as irresponsible abdication. For a large corporation like KDDI, it was likely a multi-year transformation program with phased goals and evaluation metrics. Even for SMEs, this “setting of an evaluation period” and “clarification of observation points” is the lifeline of reversible delegation.
For example, suppose you delegate a certain level of pricing authority to the sales department. If this decision is an experiment, the following elements must be decided in advance.
- Evaluation Period: Set as a 3-month trial.
- Observation Points: Average gross profit margin and its variance, changes in average customer spend, time required for sales staff to make decisions.
- Exit Conditions: If the gross profit margin falls below X%, or if price variance becomes larger than expected, risking damage to brand image.
- Reversion Method: Authority is temporarily reclaimed by headquarters. Based on observation data, revise the decision criteria (manuals, approval rules) and then either attempt delegation again or scale down to a different form (e.g., discretion for only some products).
In this way, by “designing with the assumption of failure,” delegation becomes a fearless experiment. KDDI’s IT department reform also likely repeated phase-by-phase verification based on observable metrics like “cloud migration rate” and “in-house development speed.” The goal was probably not the vague state of “being self-running,” but the clear, evaluable state of “these specific metrics have improved this much”—a state that allows for course correction if not achieved.
Resisting the “Psychological Fixation” That Destroys Reversibility
The biggest factor that makes delegation irreversible is not contracts or systems, but “psychological fixation.” “That task is Mr. A’s responsibility now.” “We entrusted this decision to Team B, so we can’t interfere.” When such tacit understandings permeate an organization, reclaiming authority becomes psychologically very costly, even if performance is poor. The feeling of “betraying trust” hinders rational judgment.
To prevent this, it is essential to share the premise from the outset with all involved that “this is an experiment.” When a manager says “I entrust this to you,” the message must not be “I’m giving this to you forever,” but “Let’s try this new way together for this term to achieve results.”
In a retail business I was involved with, we conducted an experiment to delegate product ordering authority to individual store managers. At that time, management clearly stated, “This is a 3-month trial. If it doesn’t work well, we’ll rethink the system together again.” As a result, some stores experienced skewed inventory, leading to a decision to temporarily revert authority to headquarters. However, the store managers accepted this not as a “failure” but as “lessons learned from trying,” and it was positively used in discussions to create a better system next time.
Even in KDDI’s large-scale organizational transformation, this framework of “experiment as a learning process” likely greatly contributed to building internal consensus and ensuring psychological safety.
“Self-Running” is Not a Destination, But an Adjustable Cruise State
A “self-running organization” is neither complete autonomy nor abdication by management. It can be described as an “adjustable cruise state” where, within a properly designed scope of authority and responsibility, the front lines make decisions, the results are observed and evaluated, and course correction is possible as needed.
KDDI’s case is a challenge to realize this state in a critical function like the IT department. The key to its success likely lay in the design philosophy of building “reversibility” into the very act of delegation, even before technology or mindset.
For SME leaders, there’s no need to aim for a company-wide “self-running” transformation all at once. Start with one department or one business process. When doing so, do not forget:
- Authority is delegated to the “scope of work,” not to a “person.”
- Always set the evaluation period and observation points (quantitative metrics) in advance.
- Share the premise that “this is an experiment” with everyone, making failure an opportunity for structural learning.
Delegating authority must never be a one-way, irreversible act. It is a “reversible experiment” that can be repeated as many times as needed to gauge the organization’s capabilities and aptitudes. Perhaps the most important lesson from large-scale transformations at major corporations like KDDI is that, regardless of scale, this “design philosophy as an experiment” reduces the fear of change and enables sustainable organizational evolution.
Comments