How to hire a fractional CMO.

← Back to all articles

How to hire a fractional CMO without hiring the wrong shape

How to hire a fractional CMO starts before you talk to a single candidate. The most common mistake founders make is to scope the role from a generic full-time CMO job description, trim it to two days a week, and hope the right person reads it. That gets you a smaller version of the wrong hire.

Begin by naming the actual gap, because the gap decides the person. There are four common ones. You have no marketing function and need someone to build it. You have a team but no senior leadership over it. You have leadership and a team but demand has gone flat. Or the product moved and the positioning stopped landing. Each of those needs a different operator. If you are not yet sure which gap you have, what a fractional CMO actually does and the timing triggers for hiring one are the place to start before you shortlist anyone. You do not need a precise diagnosis first, though. Recognising that you need someone to step in, own the function, and do both the strategy and the execution is already enough to justify the hire. The exact shape of the gap can get named with the person, not before them.

Write the gap down in one sentence with a number attached. "Pipeline has been flat for three quarters and we have no one who owns demand." That sentence is your brief. Every candidate, every question, every line of the eventual contract gets measured against it.

The slide-maker and the builder

This is the distinction that decides the hire, so it gets its own section. A slide-maker delivers a strategy: a positioning deck, a funnel diagram, a 12-month plan, a set of recommendations. Then they hand it to you and leave you to execute. The work is real and sometimes good. It is also the easy half. A builder writes the plan and builds the thing: the campaigns ship, the hires get made, the dashboards get wired, the function runs. A builder is judged on what is working when they leave, not on what they presented when they arrived.

In 2026 the builder line has moved. Building a modern marketing function now means building AI into it: the workflows that draft, score, route, and report without a human babysitting them. A fractional CMO who only recommends a tool list is a slide-maker wearing a newer deck. The one worth hiring embeds the systems. This is the difference between a marketer who knows AI exists and one who ships it. What that looks like in practice is covered in AI marketing systems for B2B. You want a CMO who codes the systems, not just the slides.

What to compare	The slide-maker	The builder
Main deliverable	Strategy deck and recommendations	A running function: campaigns, hires, systems
What they leave behind	A plan you still have to execute	Work that keeps running after they go
Role of AI	A tool list to evaluate later	Workflows already built into the function
How you measure them	Did they present well	Did the number move
Your risk	Paying for advice you cannot act on	Lower: progress is visible inside 90 days

Selection criteria: what to screen for

Four criteria separate a strong shortlist from a flattering one. Score every candidate on all four before you fall for the one who interviews best.

Stage and model fit. Someone who ran marketing for a 500-person enterprise is not automatically right for a 20-person B2B SaaS company, and often is not. The motions, the budgets, and the speed are different. Look for an operator who has run marketing at roughly your size and your model, product-led or sales-led, B2B or consumer. Relevant beats senior.

Operator track record. The single sharpest filter is whether they have built the thing, not advised on it. Ask for a function they built or rebuilt and what they personally owned inside it. A real operator answers in specifics. A career advisor answers in frameworks.

References who speak to outcomes. A reference who says the person was lovely to work with tells you nothing. A reference who says "she rebuilt our demand engine and pipeline doubled in two quarters" tells you everything. Ask for references who can describe a result, and ask them what actually changed.

AI literacy that ships. Marketing depth and AI literacy is the combination that is genuinely scarce. Plenty of marketers can talk about AI. Far fewer have built it into a live function. Ask for one concrete system they put into production and what it replaced. The wedge is the intersection, not pure AI depth and not pure marketing pedigree.

The questions to ask in the interview

Ownership and proof are what you are testing for, so ask questions that are hard to answer in the abstract. Strong candidates reach for specifics without prompting. Weak ones retreat to theory.

• Walk me through the last marketing function you built or rebuilt. What did you personally own, and what did you hand to someone else?
• Which number did you move, by how much, and over what period? How did you know it was you and not the market?
• How many clients do you carry at once, and concretely, how much of your week do I get?
• What would the first 90 days here look like? What is done by day 30?
• Show me one AI system you built into a marketing function. What did it replace, and is it still running?
• Tell me about something you tried that did not work, and what you changed.

The last question matters more than it looks. An operator who has actually built things has a stack of things that did not work and talks about them plainly. Honesty about the hard parts is a signal, not a weakness. A candidate with only clean wins has either never shipped or is editing.

Red flags that should slow you down

Some answers should make you pause, and a cluster of them should end the conversation. Watch for these.

• Cannot name a number. No metric they moved, no before and after. Strategy with no scoreboard.
• Only frameworks. Every answer is a model or a matrix and never a built thing. That is the slide-maker tell.
• Too many clients. A roster so long that you will get scraps of attention. Fractional means part of your team, not absent from it.
• Long retainer, no diagnostic. Anyone who wants a 12-month commitment before they have looked at your situation is selling capacity, not judgment. The work should start with a short, paid diagnostic that earns the rest.
• Recommends tools they have never run. Especially with AI, the gap between naming a tool and shipping it is the whole job.
• Treats you as their prospect. If the conversation is all pitch and no reading of your business, you are watching how the engagement will feel. A good operator reads the situation before offering the solution.

How to scope the engagement so it actually works

Scope the engagement around outcomes, not hours, because hours measure presence and outcomes measure progress. Three rules keep it honest. Start with a diagnostic: a short paid audit that maps the gaps and names the two or three highest-return moves before anyone commits to months. The audit is the product, not a free pitch. Set a real 90-day milestone, written down, so progress is visible early and either side can call it. And define what "running" looks like at handoff, so the engagement has a finish line and the systems outlive the contract. Build the system once, let it run.

On commercials, hire on a clear day rate rather than a vague monthly headcount, so you are buying defined senior output instead of an ambiguous slice of someone's calendar. If you want the full picture on what the model costs and how it compares to a full-time hire, the cost of a fractional CMO breaks down the day-rate math, and the services overview shows how a Focus4ward engagement is structured from audit to running function.

Hire the person who will be building the week after they start, not the one who will still be presenting. The slides are the easy half. The system that keeps running after they leave is the job.

Keep reading: When to hire a fractional CMO · Hiring direct vs through a company · Why a senior day rate costs you less

Frequently asked questions

How do you hire a fractional CMO?

Hire a fractional CMO in four steps. First, name the specific gap: a missing function, a leadership gap over an existing team, flat demand, or broken positioning. Second, screen for operators who have built the thing, not advisors who only recommend. Third, interview for ownership by asking what each candidate personally built and which number moved. Fourth, scope the engagement around outcomes and a 90-day first milestone, not a block of hours. The order matters. Scoping the gap first is the step most founders skip, and it is the one that makes the other three work.

What should you look for in a fractional CMO?

Look for stage and model fit, an operator track record, references who can speak to outcomes, and AI literacy. Stage fit means they have run marketing at your size and model, not two stages above. Operator track record means they built or rebuilt a function and can name what they personally owned. References should describe results, not just say the person was pleasant. AI literacy in 2026 means they build AI into the marketing function rather than recommending a tool list. The single sharpest filter is the builder test: will this person be building the week after they start, or presenting?

What questions should you ask when hiring a fractional CMO?

Ask questions that surface ownership and proof. Walk me through the last marketing function you built or rebuilt, and what you personally owned. Which number did you move, by how much, over what period. How many clients do you carry at once, and how much time do I actually get. What would the first 90 days look like here. How do you build AI into a marketing function, with one concrete example. Show me a system you built that kept running after you left. Good answers are specific and numeric. Vague, framework-only answers are the warning.

What are the red flags when hiring a fractional CMO?

The red flags are a candidate who cannot name a number they moved, one who talks only in frameworks and never in built things, one carrying so many clients that you will get scraps of attention, and one who wants a long retainer with no diagnostic first. Add two more: recommending tools they have never implemented, and treating you as their prospect rather than reading your situation. Any one of these is a reason to slow down. Two together is a reason to stop.