What Does a Head of Growth Actually Do?

What a Head of Growth actually is

A Head of Growth owns the number that the whole company is trying to move, and the cross-functional machine that moves it. Not a channel. Not a campaign. The number. At most companies that means acquisition, conversion, lifecycle, and retention all report up through one person whose job is to make the business bigger in a measurable, repeatable way.

The confusion starts because the role does not have a fixed shape. It changes with company size, and most articles describe one stage as if it were the whole job. When I have run growth, the scope has looked completely different at different stages:

• 1Early (pre-scale). You are an individual contributor with a title. You run the experiments yourself, you write the ads, you build the landing pages. The "team" is you and maybe a contractor.
• 2Scaling. The role becomes a manager of a cross-functional team, usually a handful of people across acquisition, lifecycle, and analytics. Andrew Chen puts the emergence point at a growth team above ten people inside a company of fifty to a hundred, which matches what I have seen. You stop doing the work and start setting the direction and defending the budget.
• 3At scale. You are running a portfolio: multiple paid channels, a content engine, a lifecycle program, an affiliate or partnerships channel, and the analytics to tell you which of them is actually working. At this stage the job is less about any single tactic and more about allocation, sequencing, and knowing which lever to pull when.

So the honest one-line answer is: a Head of Growth is the person accountable for growth as a system, at whatever altitude the company currently needs. The rest of this guide is what that system actually contains.

The real responsibilities

Job descriptions list "drive growth" and "own KPIs" and tell you nothing. Here is what the work actually is, from the inside, split into the four areas I have owned.

1. Set the goal, then work backward to the constraint

I do not set a revenue goal and hope. I solve backward. You pick the revenue number, divide by the return on ad spend you are willing to run at, and that hands you the exact spend and the exact acquisition cost you are allowed to pay. Now the goal is a constraint, not a wish, and every channel decision either fits inside that cost or it does not.

The trap in growth reporting is only watching lagging metrics like revenue and ROAS, which grade a decision you made six weeks ago. When I ran the weekly growth review at Superpower it was one fixed dashboard, same rows every week: a top-line KPI, then per-channel spend with a target, actual, and variance column, then an initiative tracker where every row had a name next to it. I split the scorecard into output metrics and input metrics and paired them one to one, so we were steering on the activity we could still change this week, not just the result. If a number was blank on a Friday, that was the meeting.

And I learned to plan the stack, not the silver bullet. On paper we would sometimes have needed to cut acquisition cost by more than a third, or lift revenue-per-customer by more than half, to hit a target. Neither is realistic alone. So I ran it as a combined model: small, believable gains across conversion, revenue-per-customer, referral, and cost of goods that compound to the number. I would rather bank five fifteen-percent wins than bet a quarter on one fifty-percent miracle.

2. Run paid as a portfolio, and build the creative supply chain

Anyone who tells you a Head of Growth spends the day in the campaign manager tuning bids is describing a job that stopped existing around 2019. At the scale I ran, roughly eighty percent of the work was creative operations and twenty percent was media buying. The algorithm allocates the budget now. Your leverage is the quality and volume of concepts you feed it.

Meta's retrieval system evaluates orders of magnitude more ad variants in parallel than a human ever could, which means it rewards genuine conceptual diversity, not five versions of the same ad with a new hook. So we scaled concepts, not ads: when a pattern-break format hit, we produced ten more in that family instead of re-cutting the winner. That turns the growth org into a production line, internal producers plus agencies mapped to funnel stage and persona, with a kill-and-refresh clock measured in days.

I did not optimize a channel, I ran a portfolio with a phase. In a scaling phase I would weight the primary paid engine heavily and protect brand search; in an efficiency phase I would cut the experiments and pour into retargeting. Same account, opposite instructions, depending on what the business needed that quarter.

3. Build the organic and lifecycle flywheel, and defend it

Scaling SEO is not "write more posts." At Superpower I took organic from around three thousand to two hundred thousand monthly visitors, worth roughly two million US dollars a year in revenue, across editorial, programmatic, and technical work. The programmatic side was a CMS of more than a thousand pages across roughly twenty topic clusters.

Then a Google core update cut the templated clusters overnight, and the actual Head of Growth job showed up: not producing content, diagnosing why. I pulled our own Search Console data and a cohort of competitors, and found a clean natural experiment inside a rival's own domain, where their templated blog fell while their structured, doctor-reviewed guide cluster rose. Same brand, same writers. The only variable was page structure. That told me the fix was a content spec, not a link campaign, and I wrote the spec.

The other half of that flywheel is lifecycle and referral, which is almost always the underfunded side. About one in four of our members arrived by word of mouth, yet the referral program did almost none of the work. The cheapest growth I could touch was not a new ad channel; it was capturing the demand we were already generating and giving away for free.

4. Own attribution, and be willing to disbelieve it

Two attribution ledgers that disagree is not a data problem, it is the job. We ran a self-reported "how did you hear about us" survey alongside last-click tracking, and they told different stories. Last-click said podcasts drove zero conversions; the survey said listeners were Googling us. Word of mouth was a quarter of our discovery and zero percent of our tracked links. If you manage to last-click, you defund the exact channels that feed the one it over-credits.

That is the connective tissue across everything above. The Head of Growth is the only person positioned to see that the channels the tracking undervalues, podcast, word of mouth, brand, are precisely what make the trackable channels work.

Head of Growth vs CMO vs Head of Marketing

This is the question the search results are full of, and most answers blur it. Here is the distinction I would draw, having held more than one of these titles.

The cleanest way I have heard it put, and I agree, is that a Head of Marketing builds the brand and generates demand, while a Head of Growth is responsible for turning attention into measurable, compounding revenue across the entire customer lifecycle. A CMO sits above both as the executive who owns the function. The reason the titles collide is that at a smaller company one person is often doing all three, and the title just signals which part the founder cares about most.

I break the vs-CMO and vs-Head-of-Marketing distinction down properly, with the interview-scoring angle most people miss, here.

The skills that actually matter

Every list says "analytical" and "creative" and "leadership." True and useless. Here are the three that separate a real Head of Growth from a senior marketer with a bigger title, drawn from what the job actually demanded of me.

• 01Reading a number and naming the cause. The valuable skill on organic was not publishing pages, it was looking at a forty-percent traffic drop and saying "this is a page-structure problem, not a links problem, and here is the exact spec to fix it." Diagnosis over production, everywhere.
• 02Distrusting your own attribution. The competent move is not picking the winning channel, it is holding two contradictory ledgers and knowing which one is lying about which channel. Most people optimize toward the number that is easiest to measure. That is how you starve your own demand engine.
• 03Refusing to scale a broken unit economic. The single most valuable paid decision is often "not yet." Scaling spend on top of an acquisition-cost gap does not cure the gap, it multiplies it. Every DTC cautionary tale, Blue Apron, Peloton, HelloFresh, scaled spend before fixing the math. Knowing when not to press the pedal is a skill.

In a regulated vertical there is a fourth: treating compliance as a growth function, not a blocker. In a health company, every keyword ran through claims and medical review before a word was written. Volume you cannot publish is worth zero.

Salary and seniority

Compensation tracks the stage of the role more than the title. Public ranges cluster roughly like this in the US market:

Two things the number ranges hide. First, equity often matters more than base at the stage where this role has the most impact, because you are joining to build the growth engine, not to run a mature one. Second, the title is heavily inflated and deflated across companies, so a "Head of Growth" at a fifteen-person startup and one at a five-hundred-person company are barely the same job, and the pay reflects that spread more than the title does.

I put together a fuller breakdown of Head of Growth salary by stage and city here.

Fractional or full-time: how to hire one

Most companies reach for a Head of Growth about a year too early or a year too late. Too early, and you are paying a senior salary for work that is really one channel that a specialist could own. Too late, and you have a founder still running acquisition personally while the rest of the business waits.

The honest test is whether you have a growth system that needs an owner, or a growth task that needs a doer. If it is a system, multiple channels, a funnel, a lifecycle, and nobody holding the whole thing, that is a Head of Growth. If it is one channel that is not working, hire a specialist in that channel first.

This is exactly where a fractional Head of Growth fits. You get someone who has built the system before, at the fraction of the cost and time you need before the role justifies a full-time executive salary. It is what I do now: come in, build the growth engine and the team to run it, and hand it over. If that is the stage you are at, that is the work I do.

I wrote up the fractional Head of Growth model, what it covers and when it beats a full-time hire, here.

Who to actually learn from

The best way to understand this role is to watch the people who do it in public and write about it honestly. I keep a running set of profiles on the sharpest Heads of Growth and growth leaders, updated as they publish, so you can see how each of them actually thinks about the job right now rather than a static bio.

New profiles are added as I find growth leaders worth following. If you run growth and want in on a small, invite-only room of people who actually do this well, tell me.

Frequently asked questions