The question “how many people can a manager lead?” seems to ask for a number. In practice, it calls for a capacity analysis.
The latest data from Gallup helps put the problem into perspective. In 2025, managers in the United States had an average of 12.1 people under their leadership, but the median remained at approximately five to six. Two-thirds managed fewer than ten people. The difference between mean and median occurs because a smaller portion of very large teams pushes the average upwards. Therefore, 12 reports should not be interpreted as an organizational design goal.
The most useful conclusion is another: there is no universally safe span of control. The sustainable number depends on the work that the manager performs, the degree of autonomy of those led and how much individual monitoring, coordination and decision-making the team requires.
What span of control measures and what the number hides
Span of control, or breadth of control, is the number of people who report directly to a manager.
If a director has six managers as direct reports and these managers together lead 60 professionals, her formal span is six, not 60.
This indicator is important, but incomplete. Two managers with eight reports can face radically different loads. One can coordinate experienced professionals, carrying out standardized work and making decisions autonomously. The other can lead newly hired people, in different functions, with strong dependencies between activities, while remaining responsible for their own technical deliveries.
An academic review published in the Journal of Management in 2025 reinforces this contingency logic. The reviewed literature associates smaller spans with greater possibilities for relational coordination, coaching and feedback, while technologies that increase access to information can allow more autonomy and larger spans. The same work highlights that managerial attention is limited and that interdependencies between activities change coordination needs.
So counting people is just the beginning. It is necessary to estimate the actual management burden.
How many direct reports serve as reference?
A well-known reference from McKinsey divides managerial work into five archetypes. The model, published in 2017 and based on the consultancy's experience with organizational designs, should not be treated as a standard, but remains useful for showing why different teams allow for different amplitudes.
| Manager context | Reference span | Dominant trait |
|---|---|---|
| Player-coach | 3 to 5 | Manager still produces a lot, work with little standardization and intense need for guidance |
| Coach | 6 to 7 | Frequent development, some standardization and professionals in the process of gaining autonomy |
| Supervisor | 8 to 10 | Best known and standardized work, with moderate autonomy |
| Facilitator | 11 to 15 | Very standardized and led processes that require less individual intervention |
| Coordinator | 15 or more | Highly standardized or automated work, with a manager focused on coordination and exceptions |
These ranges do not mean that an engineering team must necessarily have five people or that a standardized operation can have 20 without consequences. They work best as hypotheses to discuss work design.
The central point of the model is that larger spans are more defensible when the manager does less individual work, processes are more predictable, team members perform similar activities, and the team quickly becomes self-sufficient.
2026 search didn't find a magic number
Gallup published a meta-analysis on span of control in January 2026 that brought together 312 studies, 92,252 work units and 897,971 employees, distributed across 104 organizations, 26 sectors and 46 countries. Teams were classified as small, with 4 to 9 people; medium, with 10 to 19; and large, with 20 or more.
In small and medium-sized teams, the relationship between engagement and results such as productivity and turnover appeared more consistently. In teams with 20 or more people, the relationship between engagement, productivity and turnover varied more according to sector and operational context.
This does not prove that nine people are better than ten, nor does it establish 20 as a limit. The methodology itself was not designed to produce this type of cut. It shows something more relevant for those who are designing a structure: the larger the span, the more important the conditions under which the work takes place.
Large teams can work. It's just less safe to assume they will work just because another company uses the same ratio.
The first limit is usually the manager's attention span
Gallup found another piece of data especially relevant for managers who accumulate leadership and execution: 97% of managers surveyed in the United States said they performed some individual contribution work, and the average was to dedicate 40% of their time to these activities.
When this portion exceeded 40%, the engagement of the managers themselves tended to be lower, and the difficulty increased as the number of people led increased. Gallup does not propose 40% as a universal limit, but the data works as a warning sign for organizations that expand teams without taking deliveries out of the manager's hands.
Deloitte reached a similar diagnosis through another route. In his research on the future of management, published in March 2025, he observed that almost 40% of managers' time was being consumed by administrative tasks and solving immediate problems. The organization also reports on a business experiment in which the number of professionals per manager almost tripled: day-to-day operations continued to function, but coaching and development suffered, accompanied by lower engagement and higher turnover.
This is an important risk of excessive span: the first signs do not need to appear at operational delivery. The manager can maintain deadlines and indicators while silently reducing feedback, development, conflict prevention and individual attention.
Feedback provides a concrete test of capability
A practical way to test a manager's ability is to ask whether the cadence of relevant conversations continues to fit into the agenda.
In Gallup research, receiving meaningful feedback at least weekly was strongly associated with engagement. The organization also verified that these conversations do not need to take up an hour: frequent meetings of 15 to 30 minutes can be sufficient when dealing with recognition, priorities, collaboration and the professional's strengths.
This transforms an abstract discussion about “eight or twelve employees” into an operational question: after expanding the span, can the manager still maintain useful individual conversations, make decisions in the necessary time and monitor development without relying on nights, weekends or recurring cancellations?
If the answer is no, the structure has already exceeded the available capacity, even if the organization chart appears efficient.
The CIPD also recommends that organizations review managers' workload and ensure time for meaningful one-on-one conversations. Its 2024 analysis relates good people management not only to a reasonable span, but also to workload, autonomy and available resources.
When is it safer to increase the number of reports
Enlargement tends to be less risky when multiple conditions appear at the same time:
- professionals already work autonomously and need little daily supervision;
- the activities are relatively similar or the processes are well defined;
- routine decisions are delegated and do not need to constantly go up to the manager;
- the manager can reduce technical or administrative work as he receives more people;
- individual feedback continues to occur with frequency and quality;
- onboarding, performance issues or relevant changes are not concentrated simultaneously;
- there is data to quickly notice deterioration in decisions, monitoring or development.
The presence of just one of these conditions is not enough. Automating reports, for example, can save administrative time without reducing the complexity of human decisions that continue to reach the manager.
When expanding reporting is probably a false economy
The clearest sign of excessive span is not the absolute size of the team. It is the systematic replacement of leadership with screening.
This happens when individual meetings are canceled first, conversations become only about status, decisions are left waiting on the manager's agenda, problems are discovered late and performance evaluations become periodic large-scale efforts.
The player-coach manager who remains responsible for his own technical goals also deserves caution. Adding people without removing some of that responsibility does not create managerial capacity; it just distributes the same hours between more demands.
Heterogeneous teams increase the difficulty. Leading eight professionals who execute the same process does not produce the same workload as eight specialists from different areas, each with their own decisions, stakeholders and development needs.
Similarly, an organizational chart can show eight direct reports and hide several matrix relationships, project leads, ongoing hiring and functional responsibilities. For sizing, it is worth looking at the formal span and also this effective attention span.
How to decide whether a manager can accommodate more reports
Before changing the structure, record for a few weeks how management capacity is already being consumed. The goal is not to fill out a perfect timesheet, but to figure out which responsibilities will compete for space when new reports arrive.
Observe at least five dimensions: frequency and quality of individual feedback; time consumed by decisions and escalations; delayed development and performance actions; portion of the agenda dedicated to individual execution or administration; and situations in which the manager becomes an obligatory point of passage.
Then, simulate the new structure. If three additional people join, what activities will the time needed to keep up with them come from? If the answer is “the manager will need to organize himself better”, there is no capacity created yet.
The safest logic is to remove, delegate, standardize or automate load before significantly increasing the span.
When there are conditions for expansion, an incremental change allows us to verify whether the new structure preserves the mechanisms that justify the existence of management. Compare the before and after using simple indicators: percentage of professionals who received significant feedback, delays in decisions, cancellations of individual conversations, development actions overdue and distribution of the manager's time.
Engagement and turnover also matter, but they tend to react more slowly. Waiting for these indicators to worsen to discover that the span has become too large is managing in the rear view mirror.
The number must be a consequence of the work design
For complex teams that depend on coaching, the references found point to smaller spans. For mature, standardized, autonomous operations, larger spans may work. Recent evidence also shows that teams of 20 or more people exist and can produce good results, but their results are more context dependent.
Therefore, the decision should not start with “we need to reach ten reports per manager”. It should start with “what work do we expect this manager to do and how much of that work can be done sustainably?”.
The appropriate span is one in which the manager is still able to deliver the three functions that the agenda places at the center of the decision: sufficient feedback, timely decisions and individual development.
If increasing the number of reports requires sacrificing one of them, the organization has not gained efficiency. You just transferred a cost that was visible in the organization chart to a less visible problem in the operation.
