How to Hire a Fractional COO: Scope, Signals, and Engagement Shapes

If you run a marketing agency and you are searching for how to hire a fractional COO, you have usually hit a particular ceiling. The work is not the problem and the team is not the problem. Every operational decision still routes through you, and you have run out of hours to route them through.

I sell this seat, so read what follows with that in mind. The basis for it is twelve years running delivery operations for marketing agencies, 79 completed engagements and a 4.9 out of 5 client rating, including roughly five years as one agency owner’s right hand, running her entire delivery operation. What follows is the employer’s side: what the seat owns, whether you are ready, what to test before you hire, and what to put in the agreement. I will also be specific about when not to hire one, because the version of this article that never says that is a sales page.

What does a fractional COO actually own?

Start here, because COO is the loosest title in the agency world, and that vagueness is the most common reason the hire disappoints. In an agency of five to fifty people, the seat covers the operating layer: how capacity gets planned, how work moves between people, which tools the business runs on, and the recurring decisions sitting in your inbox waiting for a yes.

What the seat does not cover matters as much. Strategy, client relationships and anything irreversible stay with you. Someone who starts redirecting the company rather than running it has taken a job you did not offer.

So before you interview anyone, decide in writing whether the seat includes hiring and firing, financial ownership, or client-facing escalation. The title settles none of those questions, and two candidates can both answer honestly while describing completely different jobs. If you are not certain this is even the right seat, I have written separately about the difference between a COO, an operations consultant and an integrator, and the scope I take is set out on my fractional COO services page.

How do you know your agency is ready for a fractional COO?

Readiness for this seat is not measured in chaos but in where decisions pile up. You are close when several of these are true at once:

• You are the escalation point for decisions that do not need you, and your team has learned to wait rather than guess.
• You have capable leads or project managers, but nobody owns the system they work inside, so each of them solves the same problem privately and differently.
• Capacity is a feeling rather than a number. You discover you were overloaded at the point something slips.
• You cannot take two consecutive weeks off without the operation degrading while you are gone.
• Growth is limited by delivery rather than by sales. You hesitate before taking on new work because you are not confident the machine will hold.

Now the other direction, because the fair answer is not always to hire. Two conditions should stop you. The first is that the pain is really at the project level: dates, status, client updates, coordination between people who are already busy. That is a project manager, and a more expensive seat will not fix it faster. The second is harder to admit. If you are not prepared to let someone else make operational calls without checking with you, this hire will frustrate both of you, because the value is delegation rather than an expensive assistant asking permission.

If you cannot tell which description fits you, that is a diagnosis problem rather than a hiring problem, and it is cheaper to solve as one. A fixed-fee Agency Ops Audit tells you which seat you are short of before you commit to a retainer or a salary.

Should the seat be fractional or full-time?

This is not a comparison of two prices but of two structures. A full-time operations executive is a fixed cost that starts before the workload justifies it: the hours, plus the search, the ramp-up quarter where you are training rather than being helped, the employment overhead, and the notice period at the other end. None of that argues against the hire. It argues for making it when the seat genuinely fills a week.

In most agencies of five to fifty people it does not. The operating layer is concentrated work: capacity planning, the weekly rhythm, tooling decisions, and a set of calls that need a senior head but not a full one. Paying for forty hours of seniority to cover a slice of a week is the expensive option rather than the safe one, and the fractional version usually removes the ramp-up problem too, because the person has done the job before rather than learning it on your clients.

My own arrangement is 10 to 15 hours a week on a $4,000 to $6,000 monthly retainer with a 90-day minimum. I post the numbers so you do not have to sit through a call to learn whether this is in range.

Full-time wins on three conditions, and if two hold I would tell you to hire an employee. The operating layer already fills a full senior week. The role needs someone in the room, in your time zone, for live client escalation. You want a successor rather than a service. The useful role for a fractional operator then is to build and document the system your new executive inherits, so their first quarter is spent running an operation instead of inventing one.

How do you tell a real operator from a good talker?

One test matters more than the rest: has this person run agency-shaped operations, or have they run operations elsewhere and read about agencies? Agency shape is specific. Many small concurrent client projects instead of a few large internal ones. Capacity that is also revenue, so an idle week and an overloaded week are both expensive. Scope that expands quietly inside a fixed fee. Someone who ran operations in a product company may be excellent and still spend a quarter learning why your Thursday looks the way it does.

Four questions worth asking

• Which decisions did you make without checking with the owner in your last engagement, and which did you always take back? You are testing whether they have ever actually held decision rights, or only recommended things to someone who held them.
• How would you work out whether we have the capacity to take two more clients? Listen for a method rather than a promise. A real answer involves current commitments, who is actually available, and what the last three months looked like, not confidence.
• What would you deliberately leave alone in the first month? Anyone planning to change everything in week one has missed that your imperfect process is currently holding live client work.
• If we stopped after 90 days, what would we still have? The answer should be documentation, systems inside tools you own, and a team that can run the rhythm. If it is that things revert, you are renting order rather than buying it.

Red flags worth walking away from

• A framework you are required to adopt. If every answer routes back to their methodology, the methodology is the product and your agency is the installation site.
• A migration plan before a diagnosis. I work inside ClickUp, Asana, Monday.com, Airtable and Notion, and build automations in n8n, and I would still distrust anyone who wants to move your team before they understand how your delivery runs.
• Vagueness about hours and availability. Fractional describes a bounded commitment, not an open one, and an operator who will not define the boundary has not thought about how they will hold it.
• No answer on documentation or handover. The version of this seat that leaves nothing behind is the version you pay for twice.

What should the agreement specify?

An engagement that disappoints is more often underspecified than badly executed. Four things belong in writing before the first week.

• Decision rights. A short list of what is theirs, what is yours, what is joint. This one document decides whether the hire works.
• Hours and how they are spread. Ten concentrated hours across a week beat fifteen scattered ones, because the operating layer runs on cadence.
• A minimum term, with a reason attached. Stabilising an operation takes about a quarter, and a month-to-month arrangement pushes an operator toward visible quick wins rather than the unglamorous work that holds.
• Ownership and exit. Everything built lives in accounts you control, and documentation is a deliverable rather than a favour asked for at the end.

What should the first 90 days produce?

Agree the checkpoints before day one. A vague first quarter is how a strong hire gets judged unfairly and a weak one keeps the seat.

• Day 30. You can see the next two weeks of delivery without asking anyone. The map of how work actually moves exists, and it is written down rather than in someone’s head.
• Day 60. The weekly rhythm runs whether or not you attend, and a deadline at risk reaches you as a decision with options rather than as news that it was missed.
• Day 90. The operating layer has an owner who is not you, documented well enough to survive that person being unavailable. The honest test is whether you could take two weeks off.

So should you hire a fractional COO?

My verdict, with the bias declared once more. Below roughly ten people, or when the pain is project-level, this is the wrong seat, and buying it will not fix what is actually broken. Between ten and fifty, with a capable team, decisions queuing behind you, and a real willingness to hand some of them over, fractional is usually the right shape: the work is a slice of a week, and the person can arrive already knowing the job. Once the operating layer fills a week and needs presence rather than cadence, hire full-time and hand them a documented system instead of a mess. Whichever way you go, hire on evidence, and ask any candidate for the record behind their claims. Mine is twelve years, 79 completed engagements and a 4.9 out of 5 client rating.

Where to start before you hire anyone

You do not have to make this call blind. Every engagement of mine starts with the Agency Ops Audit: $1,500 fixed, two weeks, a written 90-day roadmap and a 60-minute readout call. It answers the prior question, which is what is actually breaking and which seat would own it. If the answer is a full-time employee, or a project manager, or nobody yet, I will say so.