Chief Operating Officer

A company of fifteen people has no use for this role. One person can meet investors, run interviews, and decide what ships this week, and the day still ends. There are barely any departments, so nobody has to work out whose decision something is.

Past forty people the arithmetic changes. Work starts appearing that belongs to no single team, and work with no owner keeps sliding. Here is how it played out at Pagedeck. Sofia, who runs sales, asked when a permissions feature an enterprise customer had requested could ship. Theo, who runs engineering, said the quarter was already full. Neither of them was wrong, which is exactly why nothing happened between them. The person who could decide was Grace, the founder and chief executive, and Grace spent two days of that week in investor meetings. Two weeks went by.

The chief operating officer (COO) exists so that this class of decision stops depending on one calendar. If the chief executive officer (CEO) owns the decisions that are hard to reverse, this role owns the ones that are reversible but need attention every week. Shutting down a market is the first kind; deciding what three teams do in which order while that market winds down is the second. Whether a company staffs the role separately depends heavily on its startup stage. Below roughly a hundred people the chief executive usually carries it; above that line most companies hire for it.

Most of the day goes to unblocking decisions and checking progress. Very little of it goes to inventing something new — the time lands on confirming that what was already agreed is actually moving. Five things come up over and over.

1. Checking progress against goals. If the quarter's targets are written as OKRs, this role looks at where each one stands and why anything that slipped, slipped. Reading the numbers is not the point; clearing whatever caused the slip is.
2. Deciding work that sits between teams. The permissions feature above is the shape of it: sales and engineering are both making correct arguments, and someone has to pick which one this quarter absorbs. These decisions rarely have a clean answer, and postponing them because the answer isn't clean just lets the cost accumulate. A feature deferred with a date attached is a decision; a feature deferred with nothing attached comes back as the same argument next month.
3. Keeping hiring on pace. When recruiting runs behind, product timelines run behind with it. If the plan was four engineers this quarter and two months have produced none, the job is to separate the causes — too many interview rounds, uncompetitive offers, or a job post nobody reads.
4. Running the cycle. The planning cycle has to open and close on time. That means teams bring goals two weeks before the quarter starts, and results get collected the week after it ends.
5. Turning repeats into rules. When the same problem shows up three times, it becomes a process. Spending approvals that used to route to the chief executive every time get split by amount instead.

The role connects other people's decisions more than it makes its own. Four counterparts come up constantly.

The chief executive, weekly. The split varies by company but the principle holds: outward-facing work and hard-to-reverse calls sit with the chief executive, and the machinery inside sits here. Where the line falls is worth writing down early, because the pair that never agrees on it ends up either duplicating each other's meetings or leaving a decision uncovered for weeks.

Functional leaders, over goals and resources. The chief technology officer (CTO), chief product officer (CPO) and chief financial officer (CFO) each run their own organization, and noticing where their plans fail to line up is this role's job. If product has promised a feature in March and engineering has booked March for a server migration, that conflict gets resolved in the room rather than discovered in March.

The board of directors, quarterly. The chief executive talks direction and funding; the chief operating officer reports what actually got done last quarter, in numbers.

Founders, in whatever arrangement the company has. When the founder is also the chief executive, the split above applies unchanged. When the founder sits inside product or engineering instead, they show up as one more functional leader.

"That's an operations issue, so let's settle it at the COO meeting this week."

No single team can decide it alone, so it goes to the forum that exists for cross-team calls.

"We don't have a separate COO yet — the CEO still covers it."

The company isn't large enough to justify splitting the role.

"Hiring has been behind for two months and the COO's team is going to write up why."

The problem spans the whole organization, so diagnosing it went to the coordinating role rather than to one department.

Around the time Pagedeck hit forty-five people, Grace started blocking Thursday afternoons for an operations meeting. Functional leaders bring whatever is stuck between teams that week. Sofia raises the enterprise permissions request, Theo says what would have to slip to fit it, and the group decides then and there whether it goes in. If it doesn't, they also decide what Sofia tells the customer.

What changed compared to the months before that meeting existed is not the quality of the decisions but the speed. Items that used to float for two weeks now close within one.

The most common misreading is treating the role as second in command. The title contains the word chief, and the person holding it sits with the chief executive more than anyone else does, so the ranking interpretation is easy to reach. But the authority here comes from scope, not seniority. Decisions about how the organization runs and executes belong to this role; decisions that change the company's direction or business model belong to the chief executive and the board. Read it as a ranking and functional leaders start escalating their own team's decisions upward, which recreates the bottleneck the role was meant to remove.

The second mistake is assuming every company should have one. It appears on other companies' org charts, so it looks like a gap. In practice the role pays off only when there are several organizations to coordinate. A twenty-person company with two teams that creates the position has added a step to its decision path and nothing else. When the title comes first and the work is found afterward, the person ends up absorbing the chief executive's overflow rather than running anything.