The magic number that does not exist

Every reorganization reaches the moment someone quotes the magic number: «a manager can handle at most 7±2 people». It is one of organizational lore's most durable legends — and one of the least useful, because reality offers serene team leads with fifteen reports and drowning managers with five. The difference is not in the number: it is in how much managing each person costs.

The sustainable span of control is a budget of hours, not a universal constant: the hours the week leaves for management, divided by the hours each report requires. The team, the seniority, the type of work change — and the «right number» moves by a factor of four within the same working week.

This calculator does the count with five numbers: current reports, real weekly hours, the management cost per report, the fixed managerial overhead and the share of individual work the role still requires. Out come the maximum sustainable span, the share of time absorbed by management, the remaining hours and a dimensionless saturation index saying how close to the limit you are working.

The budget of hours: where the span comes from

The count has the structure of every budget: start from the week (H hours), subtract the committed items, divide the rest by the unit cost. The committed items are two: the managerial overhead — cross-functional meetings, upward reporting, administration — which exists identically with one report or ten, and the expected individual work: the share of the role that must remain personal production, because almost no manager is paid only to manage.

What remains is the time for people, and divided by each person's management cost — 1:1s, feedback, approvals, escalations, coordination — it gives the maximum span: R* = (H·(100−q)/100 − o) / g. In the reference: 42 hours, minus 25% of individual work, minus 8 of overhead, divided by 2.5 hours per person = 9.4 reports.

The formula explains the legend better than the legend itself: with a middle manager's typical values the result often lands between 7 and 9 — there is the «magic number». But a team of autonomous seniors (g=1) or a squad ramping up (g=4) turns the right number into 23 or 6: the textbook photographed an average, not a law.

The same week, five management costs

The table holds the reference week fixed — 42 hours, 8 of overhead, 25% individual work — and changes only the management cost per report. The week is the same in every row: what changes is how much managing each person costs.

The factor of five between the first and last rows is the answer to the wrong question («how many reports can a manager have?») and the birth of the right one: what team are they managing? The same person, with the same energy, sustains 23 seniors or 5 juniors — and reorganizations that move heads without looking at g discover the difference at the manager's expense.

The table also says why the span changes over time without anyone touching the org chart: a maturing team lowers g and widens the capacity; hiring three novices shrinks it at once. The sustainable span is a dynamic quantity disguised as a structural parameter.

How this calculator works

The direct reports are the people answering to this manager — not the extended team. The weekly hours are the real ones, not the contractual ones: budgets are made on facts. The management hours per report are the item to estimate with care: recurring 1:1s, feedback preparation, approvals, escalations, the coordination attributable to each person — one month of calendar tells the true number.

The fixed managerial overhead is what exists regardless of the number of reports: leadership meetings, upward reporting, budgeting, administration. The expected individual work share is the part of the role that must remain personal production — the tech lead's code, the sales manager's key accounts: zero is allowed, and it is the case of the «pure» manager.

The calculator returns the solved budget: how many reports the week really sustains, and at what saturation point the current one sits.

How to read the results

The maximum sustainable span is the number to bring to org-chart discussions: not an opinion but a budget. The share of managerial time says what job this person is actually doing — at the reference's 72.6%, a manager with a quarter of a contributor.

The remaining hours are the direct comparison with the promise: 11.5 hours against the 10.5 the 25% would require — the margin is half a day. The saturation index ρ = R/R* is the verdict: below 0.75 capacity abounds (and the question becomes whether that layer is needed); between 0.75 and 1 the work sits near the limit; between 1 and 1.25 the threshold is crossed and the valve that blows is the individual work; beyond 1.25 the overload is structural and no personal discipline fixes it.

The valve that blows first

Managerial overload has a predictable sequence, and knowing it is half the diagnosis. First goes the individual work: the manager beyond threshold does not stop managing — they stop producing, and the promised contribution slides to evenings and weekends until it slides away entirely. It is the silent phase: the organization does not notice, because the managing continues.

Then go the 1:1s: thinned, shortened, cancelled «just this week». It is the expensive phase, because the management cost does not disappear — it transforms: less prevention today, more escalations tomorrow, and g rises exactly while the time to pay it falls. Managerial overload is a mortgage at a rising rate.

The last phase is visible on the org chart: team turnover, queued decisions, the manager becoming the bottleneck of every file. The calculator serves earlier — when the index crosses 1.0 and the choice is still between removing a report, delegating approvals or revising the individual contribution expectation.

The org-chart neighbours

The management cost per report lives largely in the calendar, and the meeting cost calculator is the right microscope: the 1:1s and the team recurrences are exactly this page's g and o, seen from the price side. The two pages complete each other — this one says whether the week holds, that one what filling it costs.

The other neighbour is the cost of employee turnover: the saturated manager is one of the best-documented engines of resignations — phase two's thinned 1:1s and missed feedback reach the income statement as replacements to pay for. Preventing saturation costs less than paying its effects.

The practical sequence for HR: compute R* for role archetypes (not per person), compare it with the real org charts, and treat deviations above 1.25 as operational risks — because that is what they are.

Where the numbers come from

The management cost per report is measured from one month of calendar: recurring 1:1s plus the feedback, approval and escalation time attributable to each person, divided by four weeks. Eyeball estimates are almost always low — five-minute interruptions end up in no calendar but add up to hours.

The overhead comes from the same calendar: everything that would stay identical if the team had a single person. The expected individual share is not a wish but an agreement: it must be asked of whoever evaluates the manager, because it is exactly the promise they will be judged on — and the misalignment between the share the boss expects and the one the calendar allows is this exercise's most frequent discovery.

A warning on the average g: it hides variance. The struggling report can cost four times the autonomous senior, and a single critical case moves the real span more than any average — the count is redone when the team's composition changes, not once per tenure.

Limits of the estimate

The model is linear: it assumes each report costs g hours independently of the others. In strongly interdependent teams coordination grows more than linearly — conflict combinations rise with the square of the people — so for interwoven squads the real span sits a bit below R*: the number is an optimistic upper bound.

The cost per report is an average that seasonality moves: annual reviews, budgeting and hiring peaks inflate g for whole weeks, and an average week that holds can coexist with quarters that do not. Finally, the model measures time, not quality: a span sustainable on paper with fifteen-minute 1:1s is sustainable only on paper.

A parametric estimate for information purposes: it supports organizational design and the manager-HR conversation; it replaces neither the assessment of real workloads nor structural choices, which remain judgement calls.

The individual contribution ladder

The first table holds the week fixed — 42 hours, 8 of overhead, 2.5 hours per report — and moves the share of individual work the role requires.

The last row is a warning in the shape of a number: whoever must stay 75% productive can manage one person. Yet org charts are full of experts «with a small team of four» and unchanged technical expectations — a role the formula declares impossible before it even starts, and which indeed burns out its occupants with statistical regularity.

The table also gives the organizational price of every technical promise: going from pure manager to manager-contributor (25%) costs 4.2 reports of capacity. It is not a reason to ban hybrid roles — it is the reason to size them knowing what is being bought and what is being paid.

The overhead ladder

The second table moves the fixed managerial overhead — the hours that exist with one report as with ten — all else equal.

The index column tells a story org charts do not see: the same manager with the same nine reports goes from sustainable to double overload without anyone touching the team — it is enough for cross-functional meetings and reporting to grow from 4 to 20 hours. It is the most common way middle managers sink: not because of the people they manage, but because of the organization around them.

The useful conversion: every 4 extra hours of overhead remove 1.6 reports of capacity. When the company adds the committee, the weekly report and the new ceremony, it is implicitly narrowing every span — and the table lets you say so with a number in the very meeting where the committee is proposed.

The player-coach: the org chart's most fragile role

The player-coach — half contributor, half manager — is the role growing organizations love most and the table's most fragile. Loved because it looks free: promote the best technician, keep their output, add the management. Fragile because it adds up two jobs competing for the same hours with opposite priorities: managerial urgency always wins the day, the promised production always loses the week.

The formula makes the conflict visible before it becomes a biography: at 50% expected contribution, 5.2 reports of capacity remain — not the 8-10 the org chart typically assigns. The role works only in two honest versions: with genuinely few reports (3-5), or with the technical share renegotiated downward as the team grows.

The dishonest version — full share and full team — produces failure in instalments: first the code that slips, then the 1:1s that get skipped, finally the burnt-out expert leaving with two failed trades instead of one successful one. The calculator costs ten seconds; that trajectory costs a year.

Flattening or layering: the levels dilemma

Flat organizations promise speed: fewer layers, less broken telephone, decisions closer to the work. But this page's mathematics imposes the counterpart: fewer layers means wider spans, and wide spans hold only if g falls — senior teams, standard processes, real autonomy. Flattening without lowering g does not remove hierarchy: it replaces it with saturated managers.

Layering has the opposite cost: every added level lowers the spans but lengthens the chain — and the coordination overhead between levels grows, that is, everyone's o rises. The right structure is not an ideology: it is the (levels, span) pair the population's real g sustains.

The practical test before every restructuring: compute R* with the measured — not wished — g of the teams involved, and compare it with the spans the new org chart draws. If the design requires spans of 15 with squads costing 3 hours each, the reorganization has already failed on paper: better to find out there.

The levers to widen the span

If ρ exceeds 1 and removing reports is not an option, the lever is lowering g — and the paths are known. Delegating approvals: every signature moved down a level removes minutes from every file and hours from every week. Standardized 1:1s: fixed agenda, cadence adapted to seniority (weekly for juniors, fortnightly for seniors), honest duration.

Built autonomy: self-service documentation, decisions de-escalated with written criteria, the senior acting as first informal reference for two colleagues — not one more hierarchy level, but half an hour a week saved across three heads. And overhead protection: every committee the manager stops attending returns capacity to the whole team.

The sequence matters: first lower g and o, then widen the span — in the reverse order you get the saturated version, which burns the trust the healthy version needs. The wide span is the prize of organizational maturity, not the shortcut to it.

Mistakes that distort the budget

The first is the contractual week instead of the real one: a manager working 42 hours budgeted on 40 gets every margin wrong — and one budgeted on 50 to accommodate chronic overload normalizes exactly what the count should expose.

The second is the eyeballed g: the spontaneous estimate counts the 1:1s and forgets the interruptions, approvals and escalations — often half the total. The third is the wishful individual share: the one the manager would like, not the one their boss evaluates them on; the misalignment between the two is the exercise's most frequent discovery, and it is resolved in a conversation, not in a spreadsheet.

The fourth is the average hiding the critical case: eight autonomous reports plus one in deep crisis do not make g=2.7 — they make a week dominated by one person. For teams with an active critical case, the honest count is: that report at full cost, the others at normal cost — and often the conclusion is that the critical case IS the load, and the rest of the team is already running on its own.

Frequently asked questions

How many direct reports can a manager handle?

There is no universal number: the sustainable span is R* = (hours for management)/(cost per report), and it runs from 5 to over 20 within the same week. The famous «7±2» photographs a middle manager's average values, not a law.

What is the saturation index?

ρ = current reports / sustainable span. Below 0.75 capacity abounds; between 0.75 and 1 you work near the limit; between 1 and 1.25 the valve that blows is the individual work; beyond 1.25 the overload is structural.

How do I estimate the management hours per report?

From one month of calendar: recurring 1:1s plus feedback, approvals, escalations and coordination attributable to each person, divided by four. Typical: 1 hour for autonomous seniors, 2–3 for mixed teams, 4–5 for ramping juniors.

What counts as managerial overhead?

Whatever exists even with a single report: leadership meetings, upward reporting, budgeting, administration. It is the fixed item of the hours budget — and when it exceeds a quarter of the week it becomes the first lever.

Why include the individual work share?

Because almost no manager is paid only to manage: the tech lead writes code, the sales manager keeps key accounts. That promise consumes hours of the week — and ignoring it produces «sustainable» spans that are not.

What happens when reports exceed the sustainable span?

A predictable sequence: first the individual work slides (silent phase), then the 1:1s thin out (expensive phase: less prevention, more escalations), finally turnover and a bottleneck visible to everyone.

Is a very low index a good sign?

It depends: below 0.75 the week holds effortlessly, but the question becomes organizational — is that margin a deliberate reserve or a hierarchy layer the structure does not require? Unused capacity has its own cost too.

Is the right span fixed over time?

No: it is dynamic. A maturing team lowers g and widens the capacity; three junior hires shrink it at once; reviews and budgeting compress it for weeks. It is recomputed when the composition changes, not once per tenure.

Does it apply to remote or hybrid teams?

Yes, with a caveat: remotely g tends to rise (more explicit coordination, fewer informal corridor corrections) and the communication overhead grows. The formula is identical; the typical values shift upward.

How is the count used in reorganizations?

For role archetypes, not per person: compute R* for the recurring profiles, compare it with the real org charts and treat saturations above 1.25 as operational risks. It is the way to discuss structures with a budget instead of opinions.

Does the model hold for strongly interdependent teams?

With caution: the model is linear, but in interwoven teams coordination grows more than linearly with the number of people. For highly interdependent squads the real span sits a bit below R*: the number is an optimistic ceiling.

Does this tool replace workload assessment?

No: it is an hours budget for information purposes, giving structure to the manager-HR conversation. The quality of the management, the critical cases and the org-chart choices remain judgement calls.

Updated July 2026