Step 1: Decide what the seat is actually for
Most failed fractional CMO engagements start with a fuzzy mandate. "We need marketing leadership" is not a mandate; it's a feeling. Before you write a brief, force the leadership team to answer four questions:
| Question | Why it matters |
|---|---|
| What decision is currently taking too long, or being made with weak evidence? | This is the actual job. The fractional CMO's first value is faster credible decisions. |
| Whose job is the marketing plan today, and why is that not working? | Defines what the seat replaces, and what it should not. |
| What does the team look like under the new senior — agencies, in-house, hybrid? | Sets the scope and the cadence. |
| What does success look like in 90 days that we'd recognise without a dashboard? | Forces a real outcome, not a vanity metric. |
What decision is currently taking too long, or being made with weak evidence?
- Why it matters
- This is the actual job. The fractional CMO's first value is faster credible decisions.
Whose job is the marketing plan today, and why is that not working?
- Why it matters
- Defines what the seat replaces, and what it should not.
What does the team look like under the new senior — agencies, in-house, hybrid?
- Why it matters
- Sets the scope and the cadence.
What does success look like in 90 days that we'd recognise without a dashboard?
- Why it matters
- Forces a real outcome, not a vanity metric.
If you can't answer these together, the fractional CMO will spend the first month answering them for you — at executive rates. Get there first.
Step 2: Decide between a firm and an independent
Both can work. The trade-offs:
- An independent fractional CMO gives you exactly one senior, often with deep specialism. Lower friction; smaller bench if the seat outgrows the operator.
- A firm gives you a named operator backed by a team and a methodology. Higher continuity if the engagement scales or the operator's load shifts; some risk of the firm assigning a less senior person than you saw in the pitch.
The single best filter for a firm is "who specifically would lead this engagement on day one, and can I meet them?" If the answer is vague, walk.
For more on this trade-off, see how to pick a fractional CMO firm.
Step 3: Build a shortlist on substance, not category
The category is full of category-trained operators who can describe a fractional CMO engagement in three articulate paragraphs without ever having owned a P&L. The signals that actually predict a useful hire:
- Operating history at your stage. Series A SaaS does not need someone whose last decade was at Fortune 500s. Pre-IPO does not need someone whose last role was at a six-person seed startup. Match the stage.
- Specificity in the first conversation. Ask "if you walked in tomorrow, what would the first two weeks look like?" A good answer is concrete and tied to your actual situation. A bad answer is generic and could apply to any company in your sector.
- A real point of view on AI in marketing. As of 2026 this is non-negotiable. The candidate should be able to name where AI improves the marketing operating system, where it doesn't, and what they have personally shipped using it.
- References from operators, not investors. Anyone can collect VC references. Ask for two CEOs and one founder/CMO who reported to (or alongside) the candidate.
- A clean view of what they will not do. A senior who can't name what they will turn down is selling availability rather than judgment.
Step 4: Pressure-test in the interview
Three questions to ask in the second conversation:
- "Walk me through a previous engagement where the original mandate was wrong." Senior operators will have one. Junior ones will pretend the mandate was right and the execution went smoothly. Look for someone who reset the scope.
- "Where would you push back on the plan we just shared with you?" A candidate who agrees with everything is a bad hire. The job is partly to be the person who says no on behalf of the team.
- "What does the exit look like, and what should make you fire yourself?" If the candidate cannot describe the conditions under which they should step down, they are selling permanence, not value.
Step 5: Structure the engagement to compound
The engagement shape that holds up:
- Monthly retainer, not hourly. Hourly billing produces consultants; retainers produce embedded operators.
- Two- to three-day-a-week cadence for most engagements. Less than two days is advisory; more than three is approaching full-time.
- A 30-day diagnostic kickoff with a clear deliverable: ranked bottleneck list, 90-day plan, named owners. If the firm cannot describe this without a deck, that's the answer.
- Quarterly review against the original mandate. Not a dashboard review — a "is this still the right job?" review. Mandates drift; the cadence catches it.
- A planned succession path. Either toward a full-time CMO hire, or toward an internal promotion. A fractional engagement that runs forever is rarely working.
For pricing context, see how much a fractional CMO costs.
Where teams get this wrong
- Hiring on resume velocity. The candidate who was at four hot companies in eight years is not necessarily the one who fixed anything. Tenure with attached outcomes beats brand-name surfing.
- Hiring on chemistry alone. "We just clicked" is not a strategy. The seat needs friction; chemistry-only hires often turn into agreeable consultants.
- Hiring before the leadership team is aligned on the problem. A fractional CMO cannot succeed if the CEO and the board want different jobs done. Get internal alignment before the search.
- Hiring an execution lead and calling them a CMO. A great paid-media operator is not a CMO. Match the seat to the actual scope.
- Underwriting the budget for what the plan will require. A senior plan will name spend. A senior who is not given budget authority becomes a slide-deck author. Decide upfront whether the seat owns the marketing budget.
If the search is going badly, the problem is usually upstream: the brief is fuzzy, the team is divided, or the seat is shaped wrong. Fix that before hiring anyone.