What goes in a fractional CMO job description? Seven sections: role summary tied to a business outcome, explicit scope, 90-day outcomes, day-one systems access, decision rights, hours and compensation band, and reporting line. The template is below. Write it after you have answered one question the req itself cannot answer: whether your problem is actually a marketing problem.
The job description is where you commit, in writing, to a definition of your problem. Most founders write it in an afternoon, borrow the shape from a full-time CMO req, cut the hours, and post it. The document reads fine. The problem definition inside it is the thing to slow down for, because a req scoped to marketing will be worked as a marketing req, whatever your actual breakdown turns out to be.
I see the cost of that mismatch in a specific population: companies that changed their story. I audited 47 B2B SaaS companies that renamed, rebranded, or changed category. A median of 21 months after announcing, 87% still had at least one major public surface telling the old story, and the lag ran through targeting filters, sales decks, and dashboards, systems no marketing job description reaches. Several of those companies hired marketing leaders during that window. The hires worked the req they were given.
Five decisions before you write anything
1. Is this a marketing problem?
The one decision the req cannot make for you. Marketing problem: positioning is sound, targeting and sales motion are healthy, and the marketing function itself underperforms on volume, quality, or cost. Revenue-system problem: the company changed its story or its market, and message, targeting, motion, and measurement have not caught up together.
The 30-minute diagnostic in my repositioning-lag piece scores this with eight items. If your “old story” answers cluster inside marketing, write the req. If they spread across systems, a fractional CMO will inherit a problem the role cannot reach, and you should scope a diagnostic before a hire.
The decision in table form, including the option most reqs never consider:
| Fractional CMO | Full-time CMO | Fractional GTM operator | |
|---|---|---|---|
| Scope | The marketing function: strategy, team, budget | Marketing plus a permanent executive seat | All four systems: message, targeting, motion, measurement |
| Time | 10 to 15 hours/week | Full time | Embedded, part time |
| Cost | $5,000 to $25,000/month | $300,000+ base before benefits and equity | $20,000 to $25,000/month |
| Fits when | Marketing underperforms inside an otherwise healthy revenue system | The workload justifies a full-time executive | The company changed its story and the system did not follow |
2. Are you buying judgment or a build?
Judgment: which segment, which message, which bets. That fits an advisory-weight engagement, 5 to 10 hours a week. A build: new positioning operationalized, a demand engine stood up, a team restructured. That needs 15 or more hours and a leader with operator habits. The two profiles interview differently and price differently. Deciding which you are buying before you write the req prevents the most common failure: advisory hours billed against a build-sized problem.
This tradeoff has a 2026 update, because build work no longer takes the headcount it used to. Companies with high AI adoption run GTM with 13 full-time people where their peers need 21 (ICONIQ, 2025). A fractional leader who operates the AI layer personally, meaning they build the signal monitoring, run the content pipelines against a documented voice, and instrument the funnel themselves, delivers build-scale output inside 10 to 15 hours a week. One who cannot will spend the same hours writing requirements for other people’s sprints. Strategy pedigree still matters, and it is no longer sufficient. I run my own practice this way; the gap between a leader who has evaluated AI tools and one who ships with them shows up in your pipeline within a quarter.
So write the req for a builder. The unit of progress is the shipped artifact: the sequence live in the CRM, the page published, the scoring pass that ran this week. Strategy that has not shipped is a document, and you can buy documents cheaper. Fractional hours make this stricter, because a leader who ships within them has no room for handoffs to a team you may not have.
3. What access do they get on day one?
CRM, analytics, ad platforms, call recordings, and the financial model. Decide this before hiring, because a fractional leader who spends their first month negotiating access is burning your retainer on your own org chart. If you are unwilling to grant data access, you want a consultant, and you should price accordingly.
4. What can they decide without you?
The section almost every fractional req omits. Can they kill a campaign? Reassign an agency? Change the website hero? Spend within a budget without a founder sign-off? Ten hours a week only works when those hours are spent deciding rather than queueing for your calendar. Write the decision rights down; the argument you have now is cheaper than the one you will have in month three.
5. What is the exit?
Fractional engagements end in one of three ways: a full-time hire the fractional leader helps recruit, an internal promotion the fractional leader develops, or a system handed over to the team that runs it. Pick the intended ending before you start and put it in the req. It changes who applies and how they build.
What the templates you would copy leave out
Before writing my own template, I scored nine fractional CMO job description templates and hiring guides, the pages that rank for this query and get copied into reqs, against the seven sections a working req needs (SandsDX audit, July 2026). Five state weekly hours. Three state compensation. Two commit to outcomes at 90 days. Two specify systems access. Zero specify decision rights, and zero write the reporting line into the template as a commitment rather than a passing mention.
The pattern: the market documents the meter and omits the mandate. Hours and rates make the invoice auditable. Decision rights, access, and 90-day outcomes make the engagement governable, and they are precisely the sections nobody writes down. The template below exists because I could not find one to recommend.
The template
Copy, then edit against the five decisions above.
Role. Fractional CMO, reporting to the CEO. [X] hours per week. Own marketing strategy and execution against [the one business outcome: pipeline coverage, category re-entry, launch]. Engagement reviewed at 90 days against the outcomes below.
Scope. In: positioning and messaging, demand generation, marketing team leadership, marketing budget, and the AI operating layer (signal monitoring, content pipelines, what AI assistants say about the company). Out: [name what is out: sales process, pricing, product]. If repositioning work is in scope, say so explicitly and name the surfaces.
Outcomes at 90 days. Day 30: written funnel assessment with numbers. Day 60: priority bets selected and defended to the leadership team. Day 90: first bet shipped with an early signal; team scored with a development plan.
Systems access, day one. CRM, analytics, ad accounts, call recordings, financial model.
Decision rights. Decides without approval: [campaign changes, agency management, content]. Recommends for approval: [budget over $X, headcount, pricing-adjacent messaging].
Hours and compensation. [X] hours weekly at $[X] monthly retainer. Market rates and models are covered in what a fractional CMO costs.
Reporting. Weekly written update against the 90-day outcomes. Monthly metrics review with the leadership team.
Want the scored version?
The seven sections as a weighted checklist: score a draft req, or a candidate's proposal, in five minutes. I send it personally, usually with a note on the one section I would tighten first.
Reading the responses
The req will surface two kinds of candidates, and one question separates them: “walk me through your first two weeks.” An operator answers with a work plan: which data they pull, which calls they listen to, what they will have written down by Friday of week two. An advisor answers with a meeting cadence. Both are legitimate hires. Only one matches a build-shaped req, and by decision 2 you already know which shape yours is.
Ask the second question too: “what would make you tell me this is not a marketing problem?” The candidates worth senior rates have an answer, because they have seen reqs written for the wrong problem before. The ones who promise marketing can fix anything are telling you the scope of what they sell, and it is the scope of what they will look at.
The third question earns its place if you scoped the AI layer in: “show me something you built and still run.” Some candidates answer with a list of tools they have used. The ones who can do build work inside fractional hours open a laptop and show you a system that ran this morning, and can tell you what it costs and what broke last month.
When not to post the req
If you repositioned in the last two years and pipeline has not caught up, run the read before the hire. The odds say the old story is still live somewhere in your revenue system, and a marketing req scopes the search to one system of four. The Gap Map is my version: two weeks, fixed fee, every finding cited to source, and it ends with a build order you can hire against, fractional or otherwise. A 30-minute conversation is enough to tell you whether you are writing the right job description, and I will say so if you are.