Organizational Design
The Span of Control Trap
The right number of direct reports is not a number. It is a design decision.
Leaders flatten the organization to increase speed, then discover that removing layers did not remove the work those layers were doing.
Key Takeaways
- There is no universal ideal span of control. The sustainable span is produced by the complexity of the work, team maturity, managerial load, and the operating conditions around the team.
- Flattening can reduce distance from the front line, but widening spans without redesigning decisions, information, and coordination simply moves the bottleneck into the manager.
- AI can remove administrative load, but it can also increase the volume and speed of exceptions. Redesign the managerial job before changing the number of direct reports.
The question sounds precise: how many direct reports should a manager have?
It usually appears during a reorganization, a cost reduction, or a push to become more agile. Someone presents a benchmark. The current average is six. High-performing organizations run at eight or ten. The arithmetic creates the feeling of design.
But a span of control is not a staffing ratio. It is a statement about how much coordination, judgment, coaching, and exception handling one role can absorb before it becomes the constraint. Two managers can each have eight direct reports and be doing entirely different jobs.
The Number Hides the Work
A manager supervising experienced people doing stable, similar work may sustain a broad span. A manager leading a cross-functional team through unfamiliar, interdependent, high-risk work may be overloaded with half as many people.
The difference is not managerial talent. It is the amount of work the design routes through the manager. Every unclear priority, disputed handoff, missing decision right, and invisible dependency becomes a meeting on the manager's calendar.
Span of control is the visible number. Coordination load is the hidden variable.
Why Flattening So Often Disappoints
The case for fewer layers is attractive. Information should travel farther with less distortion. Decisions should move faster. Employees should have more autonomy. Leaders should be closer to customers and the work.
Those outcomes are possible. They are not produced by deleting boxes. When a layer is removed, its work does not disappear. Prioritization, integration, coaching, resource allocation, quality control, and escalation still have to happen somewhere.
If decision rights and information flows are not redesigned, the work moves upward. Senior leaders gain more direct reports, more exceptions, and more status conversations. The organization becomes flatter on paper and more centralized in practice.
Four Conditions Determine the Sustainable Span
1. Similarity of work
Similar roles create repeatable coaching, comparable measures, and common operating rhythms. A manager can support more people when the work shares a logic. A portfolio of unrelated functions may require a narrower span even when every leader is capable.
2. Maturity of the team
Experienced teams with clear standards need less intervention. New roles, new leaders, or teams moving through major change require more attention. The same span can be healthy in steady state and irresponsible during a transition.
3. Interdependence and risk
The more the work crosses boundaries, the more integration someone must provide. High-consequence decisions also create review and escalation demands. Count the dependencies and exceptions, not just the names beneath the manager.
4. Strength of the operating system
Clear priorities, explicit decision rights, useful measures, and reliable information allow work to move without managerial translation. Weak operating conditions force the manager to become the connective tissue. Broad spans expose that weakness quickly.
AI Changes the Load, Not the Principle
AI can prepare briefings, summarize activity, surface patterns, automate routine approvals, and make operating information easier to see. That can release real managerial capacity.
It can also increase the pace of work and the volume of ambiguous exceptions. A manager who once reviewed ten decisions may now be asked to supervise one hundred machine-assisted actions. Reporting time falls while judgment load rises.
The wrong response is to assume every productivity gain should be converted into a wider span. The right question is what managerial work the technology removed, what new work it created, and which decisions still require human attention.
Diagnose Before You Redraw
Before changing layers or targets, study a sample of managerial calendars and escalations. Ask:
- How much time goes to coaching, decisions, integration, reporting, and preventable rework?
- Which decisions reach the manager because authority below is unclear?
- Which direct reports need distinct context, expertise, or operating rhythms?
- What coordination work would remain if a management layer were removed?
- What information could become visible without another meeting?
Then design the role and the system together. Clarify decisions. Standardize what should be repeatable. Protect time for coaching. Improve visibility. Only then set the span.
The Executive Question
The number of direct reports is easy to see, which is why it attracts attention. The organizational conditions that make that number workable are harder to see, which is why they are often ignored.
Do not ask whether the span is wide or narrow. Ask whether the manager is spending attention on the work only that role can do. If not, changing the number will rearrange the burden, not remove it.