There is no universal COO job description.

That is why the role creates so much confusion. One company's COO runs revenue. Another runs internal operations. Another owns finance, people, legal, and business operations. Another is effectively a president. Another is a founder's counterpart who handles everything from customer escalations to planning to recruiting. All of these can be valid. All of them can also be wrong.

The COO role is stage-specific. The question is not “What does a COO do?” The question is: “At this stage, what operating gap would prevent the company from executing its strategy?”

A COO at 30 people is not the same job as a COO at 300. A COO at 3,000 is different again.

At 30 people: the founder's operating counterpart

At roughly 30 people, the company is still close to the metal. The CEO often knows most customers, most employees, most product tradeoffs, and most risks. The company does not need a heavy operating model. It needs speed, focus, and just enough structure to stop dropping important things.

A COO at this stage is often a founder's operating counterpart. They turn the founder's context into execution without burying the company in process.

The work may include:

  • creating the first planning rhythm;
  • making hiring plans more disciplined;
  • building basic finance and people operations;
  • driving priority initiatives across the small team;
  • improving customer delivery or implementation;
  • turning founder decisions into follow-through;
  • creating lightweight dashboards;
  • removing recurring bottlenecks;
  • helping the CEO spend less time chasing loose ends.

The danger at this stage is overbuilding. A 30-person company does not need enterprise governance. It needs a few strong operating habits: clear weekly priorities, simple ownership, cash visibility, customer signal, a hiring plan, and a way to decide tradeoffs quickly.

The COO should protect speed while adding reliability. The strongest 30-person COO is usually allergic to ceremony but disciplined about promises: who owns what, by when, with which constraints, and what the founder must decide.

If the COO imports big-company process too early, the company becomes slower without becoming better. If they stay purely reactive, the CEO remains the operating system. The art is installing structure only where chaos is repeatedly expensive.

At 300 people: the cross-functional integrator

At 300 people, the company has real functions. Sales, marketing, product, engineering, customer success, finance, people, legal, support, and operations may all have leaders, teams, plans, and metrics. The company is no longer one conversation. Coordination becomes a scaling problem.

This is often the most classic COO stage.

The COO becomes the integrator across functions. Their job is to make sure strategy becomes plans, plans become resourced commitments, commitments become execution, and execution produces learning.

The work may include:

  • company planning and goal-setting;
  • executive operating cadence;
  • cross-functional initiative governance;
  • resource allocation and headcount tradeoffs;
  • decision rights between functions;
  • operating reviews and business reviews;
  • customer or enterprise delivery systems;
  • margin and efficiency programs;
  • scaling internal systems;
  • resolving repeated org seams;
  • making sure the CEO is not the default integration layer.

At this stage, the COO cannot personally fix everything. The company is too large. The COO must build a system where work moves through accountable owners.

The danger is becoming the chief bottleneck officer. Everyone brings the COO unresolved conflicts, unclear ownership, slipping initiatives, and political tradeoffs. If the COO says yes to all of it, they become indispensable in the worst way.

The 300-person COO must create operating leverage: clear forums, decision rules, executive accountability, resource discipline, and escalation paths that work even when the COO is not in the room.

At 3,000 people: the enterprise operating architect

At 3,000 people, complexity is no longer incidental. It is the environment.

The company has multiple business lines, regions, segments, product surfaces, layers of management, compliance requirements, planning systems, talent systems, governance forums, and financial tradeoffs. The COO may own large operating functions directly, but their highest leverage is often operating architecture.

The work may include:

  • enterprise operating model design;
  • annual and multi-year planning systems;
  • resource allocation across businesses;
  • business performance management;
  • global operations and regional coordination;
  • executive governance forums;
  • transformation programs;
  • acquisition integration;
  • cost structure and productivity systems;
  • risk, resilience, and compliance interfaces;
  • leadership operating standards;
  • enterprise decision rights.

At this stage, the COO is not primarily a fixer. They are an architect of how the company runs at scale.

The danger is bureaucracy. Large companies naturally create process to manage risk, but process can become a substitute for judgment. The COO must distinguish between control that improves reliability and control that only slows decisions.

A great enterprise COO increases organizational throughput. They make the company more legible, more accountable, and more resilient without making every decision climb the hierarchy.

Stage changes the COO's unit of work

At 30 people, the unit of work is often the urgent company problem: a customer escalation, a hiring bottleneck, a broken implementation process, a founder decision that needs follow-through.

At 300 people, the unit of work becomes the cross-functional system: planning, executive cadence, initiative ownership, tradeoff management, and resource allocation across functions.

At 3,000 people, the unit of work is the enterprise operating model: governance, delegation architecture, business unit interfaces, portfolio allocation, exception handling, and the company's ability to sense reality at scale.

Confusing the unit of work is how COOs become either too tactical for the stage or too abstract to be useful.

The same title hides different authority

Stage also changes authority.

At 30 people, authority may come from founder trust and proximity. The COO can walk across the room, clarify a decision, and move a priority.

At 300 people, authority must be more explicit. Functional leaders need to know when the COO is facilitating, deciding, escalating, or representing the CEO's decision rights. Without clarity, the COO becomes either toothless or resented.

At 3,000 people, authority must be institutionalized. The COO's role is embedded in governance, planning, budget, operating reviews, and executive accountability. Personality-based authority does not scale.

This is why “strong operator” is not enough. The company must design the role around the stage's operating reality.

The CEO's job changes too

The COO role cannot evolve if the CEO role does not.

At 30 people, the CEO may still be deeply involved in product, customers, hiring, and fundraising. The COO extends the CEO's operating range.

At 300 people, the CEO must stop being the default solver of cross-functional ambiguity. The COO can own the operating system only if the CEO reinforces that ownership.

At 3,000 people, the CEO and COO must divide enterprise leadership with precision. The CEO may focus more on strategy, capital, board, culture, market, and existential bets, while the COO runs operating architecture and execution discipline. But the split must be actively managed, not assumed.

A weak CEO-COO handoff creates organizational confusion at every stage.

Stage-specific questions

For a 30-person company, ask:

  • What chaos is repeatedly expensive?
  • Which founder bottlenecks are slowing execution?
  • What lightweight systems would improve reliability without reducing speed?
  • Is a COO needed, or would a chief of staff or head of operations solve the gap?

For a 300-person company, ask:

  • Where does cross-functional execution break?
  • Which decisions keep bouncing back to the CEO?
  • Are priorities reflected in resources?
  • Do executive forums produce decisions and follow-through?
  • Does the COO have authority across functions?

For a 3,000-person company, ask:

  • Is the operating model keeping pace with complexity?
  • Are decision rights and governance clear enough to preserve speed?
  • Does resource allocation reflect strategy across business lines?
  • Where has process become operating debt?
  • Can the company sense reality and respond at scale?

The COO role is not a template. It is a response to stage, strategy, complexity, and the CEO's leverage constraints.

Get the stage wrong, and the COO becomes either too heavy, too light, or too vague.