Who should lead growth depends on who can be held to the number. A full-time CMO owns it with their job attached. A fractional CMO owns the plan and the order the fixes get worked.
A head of growth owns throughput inside a funnel that already exists. An agency owns the work it produces. A consultant owns the recommendation and then leaves.
I watched a founder take all five of those pitches inside a single month. Her growth number had been flat for two quarters and the board had started asking in writing.
An agency proposed a full-funnel program. A fractional CMO proposed senior leadership for part of each week. A recruiter had two CMO candidates warmed up.
A head of growth she admired was quietly available. A consultant offered a six-week diagnostic. Five conversations, one question, five different things being sold.
All five landed on one spreadsheet inside four weeks, sorted by monthly cost, because cost was the only attribute they shared. A price column quietly implies five sizes of the same thing.
These five get compared on price. Accountability is what separates them: who answers when the number misses.
Choosing by title is how a company ends up paying an agency to run traffic into a funnel nobody has fixed. The ads are good and the money leaves anyway.
Who is accountable when the growth number misses?
Accountability stays a soft word until you make it specific. Pick a date ninety days out, imagine the number has missed, and ask what each of the five is holding.
The agency is holding the work. Ads shipped, emails sent, cost per lead inside the agreed range. All of it can be true while revenue sits where it began, because the scope came from you.
The consultant is absent. That engagement closed weeks ago with a document that may well be correct.
The head of growth is holding an experiment log and a funnel that converts better than it did. If the miss came from a weak offer, that is a fair defense.
The fractional operator is holding the sequence. What got worked, what got deferred, what they would change now.
The full-time CMO is holding all of that plus the one thing none of the others bring. A personal stake in the answer.
None of those five answers is dishonest. Each is what the relationship was built to produce, and its shape decides more than its price does.
One row in that table deserves a warning label. A head of growth can clear every goal on their scorecard while the company misses its own. Conversion up, activation up, revenue flat, because the funnel they were hired to improve pointed at a customer who was never going to stay.
That is what you buy when someone is accountable for throughput and the trouble is living in direction.
Does paying more buy more accountability?
Not in any dependable way. The two options everybody files as the cheap one and the expensive one are priced within a rounding error.
A core fractional engagement starts from $20,000 a month. A full-time CMO, once salary, bonus, equity and benefits are loaded in, costs $250,000 to $450,000 a year. On a monthly line that is $21,000 to $37,500.
Scope takes the fractional number up from there. The decision was never about buying the same leadership for less.
What separates the two is the shape of the commitment. A retainer stops when you stop it. A salary has a search in front of it and an exit cost behind it.
Both get paid in months as much as in dollars. That is what decides how expensive it is to be wrong.
An agency reads as the cheapest of the five and often ends up the largest, because the media runs on top of the retainer.
Look at what the contract promises. Creatives a month, a cadence of sends, a channel managed toward a target. Good agencies keep those promises.
An agency is excellent at producing work and structurally unable to tell you to stop spending.
That is what puts an agency and a fractional operator on opposite sides in most founders' heads.
Where each of the five actually wins
An agency wins when you already know what needs doing and the work needs more volume than one person can produce. If paid social needs eight fresh concepts a week, a studio that ships eight a week beats any strategist alive.
The test is blunt. Could you write the one-page scope of work today, on your own, with nobody helping?
If yes, buy production. If no, the missing piece sits upstream, and production will not supply it.
A fractional operator wins when the business already sells something people keep buying and the missing piece is judgment about order. That was FX Replay.
Real demand, a real product, and a measurement layer that had quietly stopped telling the truth. Every downstream argument was over numbers finance did not believe.
Deciding what to repair first mattered more than any tactic. A diagnosis, then a build, then a handoff is what a fractional engagement looks like, run on the same operating system I use everywhere.
A full-time CMO wins when growth is already predictable and the binding constraint is the next three years: category position, a brand customers can describe, a team of twelve, a board narrative.
None of that pays back inside a quarter. Recruiting a good one takes about three months, and their judgment needs another three to show up.
A head of growth wins when the offer is settled, the funnel converts, and someone has to run the loop for years. At Rocket Mortgage that was a test a week on the application flow.
No individual test was clever. The rhythm was the entire asset, and years of it added up to roughly two billion dollars of incremental loan volume.
A consultant wins on a bounded question with a deadline attached. Should we enter this market. Why did this channel stop working in March.
A consultant sells judgment and leaves. An operator sells judgment and stays to be wrong in public.
When should I hire nobody senior at all?
There is a sixth answer, and nobody pitches it, because nobody bills for it. Sometimes a flat number is flat for a reason no growth leader can reach from where they sit.
The first case is a retention curve you have never verified. If you cannot show me a cohort alive in month four, every dollar above the funnel is a faster way to lose it.
A senior hire here buys a more sophisticated version of the same leak, delivered with better slides.
The second is an offer nobody has pressure-tested. Healthy traffic, a weak conversion rate and a founder who has stopped talking to buyers is a positioning problem no channel can outrun.
Fixing it takes twenty customer conversations, and no executive alive can have those better than the person who built the thing.
The third is a company too small to feed anyone senior. This decision goes live past product-market fit, at $15M in revenue and up.
Below that, the best return on the next dollar is the founder's own hours in the ad account. Leadership needs something to lead.
Hire it early and you get a beautifully run version of an experiment you could have finished yourself in a week.
All three share a cheap escape hatch. Spend thirty days grading what you already have, because a growth audit tells you which column you are shopping in.
Start at the first row that describes you
Most founders can find themselves in a single line. Read top down and stop at the first honest match, rather than the row you would prefer to be in.
The pitches keep coming, priced like one product in five sizes. Ask every one of them the same thing before you look at the last page.
Ninety days from now, when this has missed, who is standing in the room, and what are they holding?
If you are weighing these five and want a straight read on which column you belong in, that is a conversation I am glad to have. Let's talk.