Why Management Exists: The Coordination Problem

The Foundations overview promised that we would build the idea of management from the ground up, without assuming it’s just “being the boss.” This page delivers the foundation everything else rests on. Before we can talk about what a manager does, we have to answer a stranger question: why does the role exist at all? Plenty of work gets done with no manager anywhere in sight. So what changes to make one necessary?

The short answer, which the rest of this chapter unpacks: management is not a natural feature of work. It’s a response to a specific problem that appears the moment a goal is too big for one person. That problem has a name — coordination — and once you see it clearly, the manager’s job stops being a mystery and becomes almost obvious.

One person needs no management

Start with the simplest possible case: one person, one goal.

Imagine a single baker who wants to open at 7 a.m. with fresh bread. She wakes early, mixes the dough, shapes it, bakes it, opens the door. There is no confusion about who does what, because there is only one who. There are no hand-offs, because nothing is handed off. There are no meetings, no status updates, no “wait, I thought you were doing that.” Her plan lives in one head, and that head also does the work.

Notice what’s absent here: there is no management, and nothing is missing. She doesn’t need a strategy document to align herself with herself. She doesn’t need a meeting to find out what she’s doing today. The information, the decision, and the action all live in the same person, so they’re automatically in sync.

This is the crucial starting point. Management is not a universal ingredient of getting things done. A solo worker — a freelancer, a lone shopkeeper, an artist — can pursue a goal at a high level for years and never once need a manager. Whatever management is, it is not required by work itself.

So what changes?

The goal outgrows the person

Now the baker succeeds. Demand triples. She physically cannot mix, shape, bake, serve, and handle the till fast enough. The goal — “open at 7 with enough bread to sell all day” — has outgrown what one pair of hands can do.

She hires help. And the instant she does, something new appears that was never there before: the work now has to be split across more than one person, which means it has to be joined back together. The dough one person mixes has to be ready when another person is free to shape it. The loaves have to come out of the oven before the front-of-house runs out to sell. Two people are now doing one job, and the two halves have to meet.

That meeting-back-together is the thing that didn’t exist with one person. It has a cost, and a name.

ONE PERSON                        SEVERAL PEOPLE
-------------------------         -----------------------------
goal ── plan ── work             goal ─┬─ plan ─┬─ work (A)
  (all in one head)                    │        ├─ work (B)
                                       │        └─ work (C)
no hand-offs                           └─ and now: keep A, B, C
no alignment needed                       pointed the same way,
nothing to sync                           handing off cleanly,
                                          not colliding

The coordination problem

Here is the heart of it. When you add people to a goal, the work grows in proportion to the headcount — two people can do roughly twice the tasks. But the connections between people grow much faster.

With 2 people, there is exactly 1 relationship to manage — one line of communication, one hand-off to get right. With 3 people, there are 3. With 4, there are 6. With 5, there are 10. The pattern isn’t linear; it accelerates. A team of ten people has forty-five possible pairs who might need to talk, agree, or hand something off — and that’s before you count the ways any of those exchanges can be misheard, forgotten, or contradicted.

People   Pairwise links between them
  2            1
  3            3
  4            6
  5           10
 10           45

This is the coordination problem: as headcount grows, the number of hand-offs, dependencies, and possible miscommunications grows far faster than the headcount itself. You add people to go faster, but each new person also adds friction — more places for signals to cross, more decisions that need someone to reconcile them. Left unmanaged, a growing group doesn’t just fail to speed up; it can slow down, because everyone spends more time figuring out what everyone else is doing than doing their own part.

Coordination is the overhead created by splitting a goal across people. It is not optional and it does not go away. The only question is who handles it.

The evidence — why bigger teams don’t scale linearly

This isn’t just arithmetic; it shows up in real work. Studies of software teams, going back to Fred Brooks’s observations in the 1970s, generally find that adding people to a late project can make it later, because the coordination cost of bringing them in outweighs the extra hands — at least for a while. More broadly, research on team size tends to find that beyond a certain point, adding members yields diminishing and sometimes negative returns on output per person. The exact numbers vary by task, but the direction is consistent: connections cost, and they cost more the bigger you get.

What coordination must supply

If coordination is the problem, we can be precise about what a solution has to provide. There are three things a group needs that individuals cannot supply for themselves — because by definition they’re about the relationships between people, and no single person owns those.

1. Shared direction

Each person can decide what they think the goal is. But nothing guarantees those private versions match. One baker optimizes for the freshest bread; another for the fullest shelves; a third for the shortest queue. All reasonable, all slightly different, all pulling the work in different directions. Someone has to hold one version of “what we’re actually trying to achieve” and keep everyone pointed at it. A person can’t align themselves with a picture they’ve never been shown.

2. Division of labor

With one person, the split is trivial: you do everything. With several, who does what becomes a real decision — and a shared one. If two people both assume they’re mixing the dough, no one shapes it. If neither assumes they’re on the till, customers walk. The division has to be decided once, made explicit, and adjusted as things change. No individual can unilaterally decide the whole split, because it depends on what everyone else is doing.

3. A way to resolve conflicts

Reasonable people, working hard, will disagree — about priorities, methods, whose task comes first when both are urgent. Two nurses on a ward both need the one available bed for their patient. Two engineers disagree on how to build a feature. These aren’t failures; they’re the normal texture of shared work. But a disagreement between peers can stall indefinitely, because neither side can end it. Coordination needs a way to make a call and move on.

The three things coordination must supply
-----------------------------------------
Shared direction   → one version of the goal, everyone aimed at it
Division of labor  → who does what, decided and made explicit
Conflict resolution→ a way to make a call when peers disagree

Look at those three needs. Holding the shared direction, deciding the division of labor, and resolving conflicts — that is a job. It’s a job that no individual doing the actual work can do for the group, because it’s about the group. The moment that job becomes big enough to need someone’s real attention, you have invented management.

That’s the whole derivation: management is the answer to coordination overhead. It is created by the problem, not by the org chart.

Management is the answer to overhead, not a reward

This reframing matters, because most people meet the word “manager” through the org chart, where it looks like a rung on a ladder — a thing you’re promoted into as a reward for being good and senior. That story gets cause and effect backwards.

The role doesn’t exist because someone deserved a title. It exists because a real, structural problem — coordination overhead — showed up and needed an owner. If your organization had no coordination cost, it would need no managers, no matter how senior anyone got. The job is defined by the problem it solves, not by the seniority of the person who happens to hold it.

We’ll build on this heavily in The Manager’s Real Output Is the Team’s Results: once you see the manager as the owner of coordination, it’s clear their output isn’t their own tasks — it’s whether the group delivers.

The common trap — treating management as a prize, not a problem to be solved

The most common mistake new managers make flows straight from the org-chart story: they think the promotion is the point, so they keep doing their old expert work and treat the coordination as an annoying tax on their “real” job. But the coordination is the real job now. Ignoring it doesn’t make it disappear — it just leaves the overhead unhandled, which is exactly the failure the next section describes. If you find yourself resenting the meetings and hand-offs, it usually means you haven’t yet accepted that owning them is the work.

Before and after: two people versus ten

The clearest way to feel this is to watch the same goal at two sizes.

Two people. A café run by two friends. The whole coordination problem is solved by conversation. “You take the front, I’ll bake.” “We’re low on cups — grab some when you’re out.” One relationship, one line to keep clean, and they can keep it clean in real time, all day, just by talking. No one is “the manager,” and nothing is missing — because with two people, coordination is cheap enough to handle on the fly. This looks exactly like the one-person case with a bit of talking added.

Ten people. Now the café is a busy restaurant: cooks, servers, dishwashers, a host, a bar. Forty-five possible pairs. If everyone just “talks when they need to,” the failures are quiet and deadly. The kitchen plates a dish nobody picks up because the server thought another server had that table. Two people both ordered napkins; nobody ordered gloves. Friday’s rush hits and no one decided who covers the bar when the bartender takes a break. Nothing dramatic breaks — there’s no single visible catastrophe. The place just runs worse than the sum of its people: slower, tenser, dropping things, with everyone privately sure they’re doing their part.

The fix is not “hire better people” or “tell everyone to communicate more.” The fix is that someone has to own the whole — hold the shared direction (what kind of night are we running?), set the division of labor (who’s on what station, who covers breaks), and make the calls when priorities collide (VIP table versus the big party). That someone is the manager, and the ten-person restaurant needs the role for a reason the two-person café simply didn’t have.

In practice — the failure is invisible until you name it

The dangerous thing about the coordination problem is that unmanaged, it rarely produces an obvious disaster you can point to. There’s no alarm that says “coordination is failing.” Instead you get a slow drift: missed hand-offs, duplicated effort, decisions that never get made, a general sense that the team is busy but not effective. People blame each other, or “the process,” or bad luck. The real cause — that no one owns the whole — stays hidden precisely because it’s nobody’s individual fault. Half of learning to manage is learning to see this overhead so you can take responsibility for it.

How it connects

This chapter is the root of the whole part. The Shift That Feels Like a Demotion explores what it’s like to move into this role and stop doing the hands-on work. Management vs Leadership vs Doing the Work separates coordination from the other things leaders do. Authority vs Influence asks how the owner of coordination actually gets people to move. And What a Manager Actually Does All Day turns these three needs — direction, division, and conflict resolution — into the concrete daily activities that fill a manager’s calendar.

Try this

This week, find a group effort you’re part of — at work, in a club, a family project, anything with three or more people — and map its coordination cost. List every hand-off (where one person’s work has to reach another before either can finish) and every point where two people might disagree about priority or ownership. Then ask: who currently owns keeping all of that aligned? If the answer is “no one” or “everyone assumes someone else,” you’ve just found unmanaged coordination overhead — and probably the source of the group’s quiet friction.

Reflect

1. Think of a time a project went smoothly with just one or two people but got messier as more joined. What specifically got harder — the work itself, or the connections between people?
2. In a group you’re part of now, roughly how many “pairs” of people need to coordinate? Does the friction you feel match that number better than it matches the headcount?
3. Of the three things coordination must supply — shared direction, division of labor, conflict resolution — which is weakest in a group you know? What goes wrong as a result?
4. Where have you seen management treated as a reward for seniority rather than an answer to a problem? What effect did that have on how the role was done?
5. Can you recall a “quiet failure” — no dramatic disaster, just a group performing worse than the sum of its people? Looking back, was anyone actually owning the whole?

Show reflections

1. The useful realization is that the work usually didn’t get harder — the coordination did. Naming that separates “we need more skill” from “we need someone owning the connections,” which are completely different fixes.
2. This is the coordination problem made personal. If a five-person group feels far more than “2.5x” harder than a two-person one, that’s the pairwise growth (ten links, not five) showing up in your own experience.
3. Pinpointing the weakest of the three is diagnostic: missing direction shows up as people pulling different ways; missing division shows up as gaps and duplication; missing conflict resolution shows up as things that stall and never get decided.
4. Good answers connect the “prize” framing to a concrete cost — a manager who kept doing their old job, or who avoided the hard coordination calls because they saw the title as an achievement rather than a responsibility.
5. The goal is to practice seeing invisible overhead. The strong insight is that quiet failures are usually nobody’s individual fault, which is exactly why they persist — until someone takes ownership of the whole.