Skip to content

The buyer’s guide

How to Hire a Fractional COO: The Complete Buyer’s Guide

Hiring a fractional COO is a five-step process: define the mandate, source candidates, vet for real scale experience, structure the engagement, and agree how it will be measured. Most bad hires fail at step one, before a single interview. This guide walks through each step — including the red flags and reference questions most buyers never think to use.

The fractional COO market has grown faster than buyers’ ability to judge it. Anyone can print the title; there is no licence, no register, and no standard scope. That puts the burden of diligence on you — and the good news is that diligence works. An operator who has genuinely run operations at scale can describe, in specifics, what they built, what it measured and what happened next. One who has not will speak in frameworks and adjectives. The entire hiring process below is designed to force that difference into the open before you commit — while the stakes are still an interview, not a retainer.

Sequence matters more than speed. Companies that hire well define the mandate first — the two or three operating problems the person must own — and only then look at candidates. Companies that hire badly collect impressive résumés first and retrofit a mandate to the person they liked most. That order produces expensive drift: a senior advisor attending meetings, no agreed measures, and a quiet parting some months later with nothing installed. Write the mandate before you open a single conversation, and every later step — sourcing, vetting, structuring, measuring — becomes a test against it. That one page is the cheapest insurance available in this market.

I write this as a practitioner, not a marketplace. My own background is nineteen years inside operations, most recently as Senior Director, Business Excellence at Publicis Groupe — quality and delivery across 500+ clients, 2,000+ teams and more than USD 750 million in annual media spend. I have sat on the operator’s side of these hiring conversations many times, and this guide reflects what the serious ones have in common: a clear mandate, evidence over polish, artefacts over attendance, and measurement agreed before the fee is. Use it to hire anyone in this market, including me — the point of a process built on evidence is that it never asks you to take anyone’s word.

In depth

What you need to know.

Step one: define the mandate before you meet anyone

A mandate is not a job description; it is a short statement of the two or three operating problems this person must own and the change you expect to see. “Billing approvals take two months; we need them at a fraction of that.” “Quality is slipping under load and clients are noticing.” “Every decision routes through the founder and the queue is growing.” Write it in one page, with the current numbers where you know them. This does three things: it tells you what experience actually matters, it gives every candidate the same problem to respond to, and it becomes the baseline the engagement is later measured against. If you cannot write the mandate, hire a diagnostic first, not a COO — you are not ready to brief one.

Where to find fractional COO candidates

Three sourcing channels exist, each with a trade-off. Personal and investor networks produce the highest-trust candidates — a referral from a founder or operating partner who has seen the person work is worth more than any profile — but the pool is narrow. Fractional-executive platforms and talent marketplaces offer speed and choice; the screening varies widely, so treat platform vetting as a starting filter, not a substitute for your own. Independent practitioners found through their published work — writing, frameworks, talks — let you inspect how the person actually thinks before you ever speak. In practice, run two channels in parallel and hold every candidate to the same mandate and the same vetting bar, whatever the source. The channel signals convenience; it does not signal quality.

Vetting for real scale experience, not the title

The single most important filter: has this person run operations at or beyond the scale you are heading toward — or merely advised near it? Titles will not tell you; specifics will. Ask what they ran, with what team, measured by what numbers, and what changed. A real operator answers in artefacts and figures: a quality score moved from 95% to 99% across 2,000+ campaigns and 450 clients; a billing cycle cut from roughly two months to fifteen days across 75 entities; a newsroom scaled fourfold to around 400 stories a day. The numbers themselves will differ — what matters is that they exist, that the candidate can explain the mechanism behind them, and that the scale is relevant to yours. Advisors describe projects. Operators describe systems and what the systems did.

The interview: make them show you the operating model

Interviews reward fluency unless you force them toward evidence. Share your one-page mandate in advance, then ask each candidate to walk you through their first ninety days against it: what they would examine, what they would install, in what order, and what artefacts would exist at day thirty, sixty and ninety. The specificity of that answer is the interview. A strong candidate names concrete things — a scorecard on numbers you already trust, a weekly operating cadence, ownership mapped to named roles, playbooks written down — and is willing to say what they would not do yet. A weak one offers methodology-speak and promises a “deep dive.” You are not buying enthusiasm or frameworks; you are buying an installed operating model, so ask to see its shape before you pay for it.

Red flags that should end the conversation

Some signals are disqualifying however impressive the résumé. No numbers: a candidate who cannot quantify a single result from a decade of work has been near outcomes, not accountable for them. No named artefacts: if nothing they built — a scorecard, a cadence, a playbook — survived their departure, nothing was installed. Hourly billing for an operating mandate: the incentive runs against the speed you are paying for. Instant certainty: prescribing solutions before diagnosis is performance, not judgement. Unlimited availability: a serious fractional operator runs a small number of engagements and says so. And vagueness about endings — an advisor who cannot describe how an engagement finishes, with your team holding the model, is describing a dependency, not a service. The stronger the résumé, the more weight these signals deserve — not less.

Reference checks that actually reveal something

References confirm facts poorly and reveal patterns well — if you ask operating questions rather than character ones. Ask: what specifically did they build, and is it still running now that they are gone? What number moved, and did the reference see it move or hear about it? Who did the work day to day — the candidate or the client team under their direction — and is that what you want? How did they behave when something went wrong? What would you not hire them for? That last question is the most useful one on the list: honest referees answer it, and the answer draws the candidate’s real boundary. One reference should ideally be from an engagement that ended — how it ended tells you what the exit will look like.

Structuring the engagement: diagnostic first, then retainer

The structure that protects both sides starts small: a fixed-fee diagnostic — typically two to four weeks — producing a written read on where the operating model strains, what that costs and what to fix first. It converts the hiring decision from a bet on interviews into a decision on evidence: you see the person work before committing, and either side can walk away cleanly. Then a monthly retainer scaled to cadence — one to three days a week — and scope, with the mandate, decision rights, deliverables and measures written down. Month-to-month or with a short notice period, never open-ended hourly. Agree at the start what ending well looks like: the model documented, the team running it, the operator progressively less necessary. Structure agreed early is what keeps goodwill from being tested later.

Measuring the engagement from day one

Agree the measures before the fee starts, or you will be judging the engagement on atmosphere. Take the numbers from your mandate — cycle time, a quality score, decision latency, the billing cycle, whatever the problems were — baseline them in the diagnostic, and put them on a scorecard both sides see every week. Then inspect trajectory, not activity: meetings attended and decks produced are cost, not progress. By day ninety you should be able to point at artefacts that exist — cadence running, scorecard trusted, ownership clear — and at least one number moving the right way. If neither is true, the structure gives you a clean exit; that is precisely why it was set up this way. A fractional engagement you cannot measure is one you cannot manage.

Questions

Common questions.

In five steps. Define the mandate: the two or three operating problems the person must own, written in one page with current numbers. Source through two channels in parallel — trusted networks, platforms, or independents whose published thinking you can inspect. Vet for real scale experience by demanding specifics: what they ran, what it measured, what changed. Structure the engagement as a fixed-fee diagnostic first, then a monthly retainer with defined cadence and scope. And agree the measures before the fee starts, so progress is visible on numbers you already trust.

Three places, each with a trade-off. Referrals from founders and investors who have seen the person operate — highest trust, smallest pool. Fractional-executive platforms — fast and wide, but screening quality varies, so do your own vetting regardless. And independent operators found through their published work, which lets you inspect their thinking before a conversation. Run two channels at once and hold every candidate to the same mandate and the same bar. Where the candidate came from tells you nothing about whether they have run operations at your scale.

Materially less time than a full-time executive search, which is part of the point. Writing the mandate takes days if you are honest about the problems. Sourcing and vetting a shortlist typically takes a few weeks rather than the months a full-time COO search consumes, because the commitment is smaller and reversible. A fixed-fee diagnostic then gives you two to four weeks of evidence before the real decision. From first conversation to an operator embedded and working, weeks — with the option to stop cleanly at each stage if the fit is wrong.

Think in structure rather than a single number: a fixed-fee diagnostic first, then a monthly retainer scaled to cadence — one to three days a week — and scope. Published US benchmarks commonly sit in the low-to-mid five figures per month for experienced operators; the UK and Europe track somewhat below; India has few published benchmarks but prices well below US levels for comparable seniority. The honest comparison is against a third to a half of a full-time COO’s total cost. Be wary of hourly billing and of quotes low enough to signal coordination rather than operating leadership.

Evidence of having run operations at genuine scale, not advised near it. Specific numbers they are accountable for and can explain the mechanism behind. Artefacts that outlived them — scorecards, cadences, playbooks still in use after they left. A first-ninety-days answer specific to your mandate rather than a generic methodology. Comfort with measurement, including of their own engagement. And a clear account of how engagements end, with your team holding the model. Polish, fluency and a familiar logo list are the weakest signals on the table; treat them as decoration, not evidence.

No quantified results anywhere in the track record. No artefacts that survived a previous engagement. Hourly billing proposed for an ongoing operating mandate. Solutions prescribed confidently before any diagnosis. Apparently unlimited availability, which suggests no other clients and no scarcity of judgement. Reluctance to agree measures up front, or to define what ending the engagement well looks like. And references who describe presence — helpful, smart, good in meetings — but cannot name a number that moved or a system still running. Any one of these deserves scrutiny; two or more should end the conversation.

Operating questions, not character ones. What did they actually build, and is it still running now they are gone? Which number moved, and did you watch it move? Who did the day-to-day work — the operator or your team under their direction? How did they respond when something went wrong on their watch? Would you rehire them, and for what specifically? And the most revealing: what would you not hire them for? Ask at least one reference from an engagement that has ended — the manner of the ending is the best preview of your own exit.

Judge on five criteria, weighted toward evidence. Scale fit: they have run operations at or beyond the size you are heading toward. Specificity: their ninety-day answer names artefacts, sequence and measures against your actual mandate. Transfer: their engagements demonstrably end with the client team holding the model. Measurement: they volunteer to be judged on numbers you already trust. Structure: diagnostic first, retainer with defined cadence, never open-ended hourly. The best fractional COO is rarely the most polished interview — it is the candidate whose past you can verify and whose plan you can hold them to.

The next step

A short conversation settles most of this — and a fixed-fee diagnostic settles the rest.