Who should my first marketing hire be after product-market fit?

Hire against whatever is currently blinding you, and for most companies past product-market fit that is measurement. A channel specialist or a generalist head of marketing both come too early at that point. If you cannot trust your own numbers, a paid manager or a content lead just produces confident decisions on top of bad data. Fix the seeing first, then hire into the channel the honest numbers show is already working.

Should I use a contractor or an employee for a new channel?

A contractor to prove it, an employee to compound it. A contractor is the cheapest way to get a truthful answer to one question: does this channel clear our economics at any real volume? That answer comes in weeks and costs little to walk away from. Only once the answer is yes does a full-time seat make sense, because now the work is scaling a bet you already know pays rather than gambling salary on a hypothesis.

In what order should I build a marketing team?

Against whatever bottleneck the machine is hitting, rather than from a template borrowed off a bigger company. The order that has held up for me twice: trustworthy measurement first, then the one channel already returning money, then creative production to feed that channel once volume becomes the limit, then lifecycle to work the customers you have already won. Each hire exists to clear the specific constraint the machine is hitting when you make it.

How big does a growth team need to be?

Smaller than you think, if every person owns a number. A compact team where each seat is accountable for one metric will outrun a larger group organized around activity and hand-offs; ownership does the work headcount is usually hired to do. The moment a role is defined as a pile of tasks rather than a result, you need more people to get less done, because nobody in the room is on the hook for the outcome.

How do I make the team run without me?

Structure it around ownership and a standing weekly review. Give every seat a single number, then hold one meeting each week where each owner faces that number, explains the miss, and commits to a change. Do that long enough and the operator who built the team stops being load-bearing, because the decisions that used to route through them now happen in the room. That handoff is the whole point of building the team in the first place.

How I build a growth team from zero

You are past the part everyone warned you about. The product sells, demand shows up without you chasing it, and nobody owns growing it. So you sit down to hire, and you freeze, because the honest problem is which role comes first.

That ordering question is the whole game, and most founders answer it by copying an org chart from a company ten times their size. A demand-gen lead, a content person, a paid manager, a designer, all brought on in the same quarter. It looks thorough, and it is almost always wrong.

Hire against the machine's next bottleneck, never from an org-chart template. The template solves a constraint that belonged to someone else's business.

I have built two marketing organizations from nothing this way, and the method was the same both times. Find the single thing capping growth right now, hire the one person who clears it, and only then look at what the new constraint has become. Use a contractor to prove a channel before you trust it with a salary. The rest of this piece is that sequence, walked through the two builds it came from.

When I arrived at FX Replay, the product already had genuine pull: real demand landing every week, and no marketing team behind it. The tempting move was to hire a paid manager and start spending against all that traffic. It would also have been the wrong first hire, because the numbers underneath the business could not be trusted. Event tracking had gone unfixed for years, so the ad platforms registered a signup and then lost the trail entirely, with no trial, payment, or revenue attached to it.

A channel hire on top of that would have spent real money optimizing toward a lie. So the first capability I added was not a channel at all. It was measurement: the work of making every number the same across finance, the ad platforms, and one dashboard the whole team could read. Until that existed, any specialist I brought on would have been making sharp decisions from a blurred picture, which is a fast way to lose money with conviction.

Each hire exists to clear the one thing capping growth when you make it. Fix measurement and the next constraint becomes the channel that already pays; feed that and the constraint moves to creative, then to the base you have never worked.

Measurement is the hire that makes every later hire worth making, because it turns the rest of the sequence from guesswork into a series of answerable questions. Once the numbers were honest, the next bottleneck named itself. We could finally see which channel was already returning money, and that told me exactly who to bring on next.

Feed the channel that already works

With trustworthy numbers, the second hire is the operator for the one channel the data shows is already paying. At FX Replay that meant pointing paid acquisition at the events that correlated with revenue instead of the cheap signups it had been chasing, then staffing the channel that responded. You do not diversify into five channels at once when one is working and underfed. You pour into the one that works until it stops rewarding more fuel.

But here is where a lot of money gets wasted: founders decide a channel is worth a full-time hire before anyone has proven the channel clears the economics at real volume. I do it the other way. A contractor goes in first to answer one narrow question, does this channel pay when we push spend through it, and that answer arrives in weeks for very little.

Contractors prove the channel, employees compound it. Flip that order and you pay tuition twice: once for the channel that never worked, and once for the salary that learned it slowly.

The contractor buys a fast, disposable read on a channel. A permanent seat earns its cost only afterward, once the job shifts from testing the bet to scaling one that already pays.

The rule comes down to what each arrangement is good at, and being cheap is the least of it. A contractor is a truth-finding instrument: fast, disposable, pointed at a single question. An employee is a compounding instrument: they carry the memory of every test, get better at the channel quarter over quarter, and own the number long enough to make it move. Spend the contractor to earn certainty, then spend the salary to earn scale.

When the channel starves, hire creative

A working paid channel has a predictable next bottleneck, and it is creative. Once you are spending seriously against an audience, the thing that decays fastest is the set of ads in front of them, and the operator who can run the accounts brilliantly cannot also produce a fresh concept every few days. So the third hire is creative production, brought on precisely when raw volume has become the limit rather than targeting.

Timing is the whole point here. Hire creative before a channel is proven and you have paid a designer to decorate something that does not convert. Hire it at the moment a proven channel is throttled by how fast you can feed it new concepts, and every asset that person makes goes straight into a machine that is already turning spend into revenue. At FX Replay the creative function grew alongside a scoring system for ad concepts, so winners got iterated and losers got cut without me in the room to adjudicate each one.

Then work the customers you already won

Lifecycle comes late in the sequence because it compounds on top of everything before it. Working the base you have already acquired only pays once acquisition is producing customers worth working and the numbers can prove which ones. At FX Replay the customer database was being used for little beyond transactional messages when I arrived, while it sat on a large and completely unworked pool of revenue.

So the lifecycle hire came once the earlier constraints were cleared, and the mandate was to turn that neglected list into a program: segmented campaigns, automated flows, and offers that brought lapsed customers back. Put that hire first, before acquisition and measurement are solid, and they are working a small base with numbers nobody trusts. Put them in at the right point in the sequence and they inherit a growing base and honest reporting, which is the difference between a marginal channel and a major one.

The team is a set of numbers with names on them

By the time I stepped into the CMO seat, the group had grown to ten people, each one hired into a specific constraint as the machine hit it. What held it together was simple: every seat owned exactly one number. The paid operator owned cost to acquire by channel. The lifecycle marketer owned revenue per contact. The creative lead owned the win rate of new concepts. Nobody owned a vague territory called their function.

When a role is a number rather than a pile of tasks, the weekly review has a spine: each owner shows up to defend their metric, explain the miss, and say what they will change. That is what lets the team run the machine once the operator steps out of the room.

That design is what made the team runnable without me. Each week everyone brought their number to one review, defended it against the target, explained what had moved it, and said what they would change. Delegation happened through the structure itself: when a person owns a metric and faces it in front of their peers every week, the operator who built the team slowly stops being the one making the calls. The rhythm that carries this is the weekly growth review.

A team of ten that owns numbers beats a team of twenty that owns tasks. Ownership is the multiplier that headcount alone can never buy.

Over eighteen months that ten-person team more than doubled the company's recurring revenue, reaching 2.6 times its starting point, and by the end it was running the growth machine without me. The full account of that engagement, from the broken data through the pricing lever to the handoff, is the FX Replay case study in full. The team was the delivery mechanism for all of it, and the sequence above is how it got built.

The same sequence across a multi-brand org

The method is not specific to one company. At Leverage Companies I owned growth for several brands at once: Brick City Capital, a lender working in real estate; Deals & Dollars, which teaches property investing; and Leverage Homes, which wholesales houses. Three audiences, three funnels, one growth function underneath them. The temptation with a portfolio like that is to build a full marketing department per brand, which is how you end up with three underpowered teams instead of one that compounds.

So I ran the same sequencing, sized to the constraint each brand was actually hitting. The binding limit across the portfolio was creative throughput: several brands all needed a steady flow of assets and none could produce them fast enough, so designers came in first, shared across the brands rather than siloed inside one. Then an email marketer, because the brands were sitting on lists they were barely working and lifecycle was the fastest revenue nobody was collecting. Then engineers to rebuild the customer data plumbing, because the systems holding the customer records could not yet feed the marketing that depended on them.

Notice that the Leverage order looks different from the FX Replay order, and that is the point. The sequence is a rule for reading the machine: whatever is capping growth this quarter is the role you hire next. At one company that started with measurement; at another it started with creative capacity. The discipline is refusing to hire ahead of the constraint, however tempting the empty box on the org chart looks.

Run this sequence yourself

None of this requires a holding company or a nine-figure funnel. If you have real demand and are staring at a hiring plan you are not sure how to order, the steps are portable:

Name the one thing capping growth this quarter before you name a single role. The bottleneck picks the hire, and never the reverse.
When the numbers themselves cannot be trusted, fix measurement first. Every later hire is only as good as the data they decide with.
Pour into whichever channel is already returning money, and resist spreading thin across channels you have not proven.
Prove any new channel with a contractor first, then convert the winner into a full-time seat once it clears the economics at real volume.
Add creative production when a proven channel is throttled by asset volume, and lifecycle when acquisition is finally worth compounding.
Give every seat one number, and run a weekly review where each owner defends theirs. That is what lets you step out of the room.

Done in this order, the team you build is not a cost center you assemble and hope for. It is a machine that gets more capable each quarter, where the newest hire is always aimed at the thing that matters most today. The org chart fills in on its own, one solved bottleneck at a time, and you end up with the team you actually needed instead of the one a template told you to want. I detail the system this team runs in the growth operating system and the full arc of an engagement in the fractional CMO engagement guide.

If you have real demand and no marketing team yet, and you are trying to figure out who to hire first, this is the kind of team I build. Let's talk.

How I build a growth team from zero

Start where the machine is blind