Gary Hamel once did the arithmetic that org charts are designed to keep abstract. Assume a span of control of 1:10, generous by many companies' standards. An organization of 100,000 employees will need about 11,111 managers, because managers of managers also need managing, layer stacking on layer. Assume, as Hamel does in his Harvard Business Review illustration, that a manager earns three times the average of a line employee, and direct management costs approach a third of the payroll, before counting the managers' assistants, the systems built to serve the reporting cascade, or the cost of decisions slowed on their way up and down the stack.
The point of the exercise is not the specific numbers, which Hamel offers as illustration rather than measurement. The point is that spans and layers are multiplication, and multiplication compounds. Every step narrower on average span, from ten to eight to six, adds managers at every level and often a whole additional level, each one a salary, a meeting generator and a relay station where information loses fidelity in both directions.
What sets the span
The traditional answer to "how many people can one person manage?" depends entirely on what the manager is assumed to do. If the job is to inspect work, approve exceptions, relay information and adjudicate conflicts, the span must be narrow, because all four activities consume the manager's personal hours. Every one of those four, though, is a variable, not a constant. Work that is inspected by peers or by automated checks needs no inspecting manager. Exceptions handled by published rules need no approver. Information that lives in documents needs no relay. Conflicts that adults resolve directly, or through named processes, need no standing judge.
This is why span is a consequence, not a starting point. An organization that leaves inspection, approval, relay and adjudication welded to the management role has chosen narrow spans and thick layers, and will pay the multiplied bill. One that strips those functions out can run spans the traditional chart considers impossible.
The existence proof
Buurtzorg, the Dutch neighborhood-nursing organization, is the cleanest documented case of the alternative arithmetic. Research by Martela and Nandram in the Journal of Organization Design describes more than 900 self-managing teams, capped at twelve nurses each, roughly 14,700 employees as of January 2021, with two directors, a back office of about fifty people, and no middle managers or supervisors anywhere in the structure. Team support comes from about 21 coaches serving nearly 950 teams, more than 45 teams per coach, and the coaches hold no formal authority and no responsibility for team productivity. A KPMG study commissioned by the Dutch health ministry around 2015 found the organization's overhead at about 8% of revenue against a sector average near 25%. The researchers flag the boundary conditions honestly: small teams, low interdependence between them, motivated front-line professionals, highly customizable service. But as a demonstration that tens of thousands of employees do not require thousands of managers, the case is difficult to argue with.
The repricing has started
The market is now running its own experiment on the middle layer. Deloitte's 2025 Global Human Capital Trends reports that US employers were posting 42% fewer middle-management vacancies at the end of 2024 than in spring 2022, and cites Gartner's prediction that by the end of 2026, a fifth of organizations will have used AI to flatten their structures, eliminating more than half of their current middle-management positions. A prediction is not a fact, and flattening by cost-cutting is not the same as redesigning the functions that made the layers feel necessary. A company that deletes managers while leaving inspection, approval, relay and adjudication in place has not flattened; it has orphaned those functions and will hire the layer back within two budget cycles.
That is the real lesson of the arithmetic. Spans and layers are outputs of a deeper design decision about where coordination lives. Organizations that move coordination into teams, rules, documents and software get wide spans and thin management as a byproduct, the way Buurtzorg did. Organizations that simply decree wider spans get exhausted managers with twenty direct reports and the same welded-on duties. The multiplication table is patient. It will invoice either way.
Sources
- Hamel (2011). First, Let's Fire All the Managers. Harvard Business Review. Illustrative arithmetic: at 1:10 spans, 100,000 employees imply ~11,111 managers and direct management costs near a third of payroll. https://hbr.org/2011/12/first-lets-fire-all-the-managers
- Martela & Nandram (2025). Buurtzorg: scaling up with hundreds of self-managing teams but no middle managers. Journal of Organization Design. 900+ teams, ~14,700 employees, 2 directors, ~50 back office, ~21 coaches without authority; KPMG (~2015) found ~8% overhead vs ~25% sector average. https://link.springer.com/article/10.1007/s41469-024-00184-y
- Deloitte (2025). 2025 Global Human Capital Trends. US middle-management postings down 42% (end-2024 vs spring 2022); cites Gartner's prediction that 20% of organizations will use AI to flatten structures by end-2026. https://www.deloitte.com/us/en/insights/topics/talent/human-capital-trends/2025.html