Span of Control

At ten people the idea is unnecessary. One founder knows everyone's name and what they are working on this week, and decisions happen wherever the conversation happens. The trouble starts somewhere past thirty. If the same person keeps trying to see all of it the same way, approvals slip by a day each, and some people go two months without a real conversation with anyone above them.

Before there was a name for this, the situation got treated as a matter of character. That lead is diligent and stays close to the team; this one leaves people alone. Look closer and the first lead had four people and the second had fourteen. A manager with fourteen who runs a 30-minute one-on-one (1:1) every other week — a regular private conversation between a manager and one person — spends fourteen hours a month on those alone, and that time almost always comes out of their own hands-on work. The difference was in the number, not in the person.

Span of control is the name for that number. Once it has a name, it can be compared and planned. You can lay five team leads side by side and see how many people each one carries, and you can count in advance how many manager seats twenty new hires will require.

This one value decides how many layers the organization has. Divide total headcount by the span and you get the number of first-line managers; divide that result by the same span and you get the layer above it. The number of divisions it takes to reach one is the number of layers.

For a company of 45 people, the split looks like this.

1. Five people per manager — nine first-line managers, two managers above them to hold those nine, and the founder on top. Three rungs between the founder and the newest hire.
2. Ten people per manager — four first-line managers, and the founder handles those four directly. Same 45 people, one rung fewer.

Fewer layers means decisions come down faster, at the cost of spreading each manager thin. More layers means each manager carries fewer people, but a concern raised on the ground passes through three people before it reaches the founder, and it arrives blunted.

So the right number moves with conditions rather than sitting still. Three conditions let you widen it.

1. How standardized the work is — a support team repeating the same procedure every day runs fine with fifteen people to a manager. In a planning group where each item needs a judgment call, fifteen leaves the manager holding nothing but a queue of pending decisions.
2. How experienced the people are — eight seasoned people and eight new hires cost different amounts of time. A team weighted toward new hires carries roughly twice the real load at the same count.
3. How close together everyone is — sitting in one room settles a lot of small things in passing, while a remote colleague in another time zone needs a scheduled slot for each of those things.

Checking the arithmetic in hours makes the call easier. One-on-ones, writing reviews, and interviewing candidates add up to two or three hours per person per month. A manager carrying twelve people spends close to thirty hours a month on management alone, and if that person also owns half a workload, neither side holds up.

The number gets pulled out in four places.

It shows up in restructuring decisions about whether to split a team or merge one. A reorg redraws team boundaries and reporting lines: fourteen people under one lead means that team gets split in two, and three leads carrying three or four people each means they get merged into one so a layer disappears.

In hiring plans it moves together with headcount, the budgeted number of seats an organization is allowed to fill. If engineering has six seats approved and one lead already carries eight people, the arithmetic says one of those six has to be a manager seat rather than another engineer. Skip that step and you end up with more people and nobody with time to look after them.

It comes up when reviewing the org chart, the diagram of teams and reporting relationships drawn as boxes and lines. Counting the people attached to each box is enough to reveal that one company holds a lead with three reports and a lead with fourteen at the same time.

It also shapes career paths. An organization with a wide span has few manager seats, so it cannot promote every strong performer into a lead. That is why it opens an individual contributor (IC) track alongside, where people grow through depth of craft instead of taking on reports. Performance management hangs off the same number: whoever carries twenty reports writes twenty reviews in one quarter, and the sentences get shorter toward the end of the stack.

"Theo has fourteen people under him right now. Let's split that in two in this reorg."

One manager is carrying more than the role holds, so the team should be divided.

"If we set the span at eight, we need two more lead seats."

Headcount planning has to budget manager seats, not only the people doing the work.

"I only have five reports, so I run my one-on-ones weekly."

A smaller span bought more time per person.

Pagedeck first raised this number on reaching 45 people. Grace opened the org chart and counted: fourteen engineers reported to Theo, while Sofia carried three. Theo's one-on-ones had slipped from every other week to monthly.

Grace split Theo's group in two by product and folded Sofia's three into the neighboring team. The count of managers stayed the same, but the number each one carries settled between five and eight.

The most common reading is that fewer reports means better management. More time per person is a real effect, so the conclusion is easy to reach. But narrowing the span adds layers, and each layer adds one more approval and one more handoff. Staff 45 people at three reports each and four rungs sit between the founder and the newest hire.

Believing there is one correct number is just as common. Something like seven has circulated long enough that people treat it as the standard. That figure came out of factory organizations built on repetitive work. Inside a single company today, a support team is comfortable at fifteen while a planning group feels tight at six. Fix the number first and fit the organization to it, and you get one lead holding a queue of pending decisions while another has so little to do that they start reaching into their team's work.