Bawar,
This is Q3. You asked how I'd build the marketing team. I've kept it narrow on purpose: headcount and capability, not budget. Spend is a separate conversation and I'd rather have it cleanly once we agree on shape.
One premise up front. I'm assuming the marketing team is zero today. I'm building this as if I'm the first person in. Everything below flows from that.
Five rules drive every choice below:
One operator who owns paid, lifecycle and measurement together. Not three specialists. A T-shaped generalist who has run paid across Meta, LinkedIn and Google, built a lifecycle motion in Klaviyo or similar, and stood up attribution and a measurement stack from scratch. AI-ready or AI-hungry, not AI-curious. A manager in posture, a hands-on operator in reality.
I want a manager title because I want someone who has actually run a small team before. I want them hands-on because at team of two, nobody has reports and everyone ships.
Content is what I can do well myself in the first six months. With a good AI stack I can produce the founder narrative, the LinkedIn motion, and the baseline editorial. I own the brand narrative personally. What I cannot credibly produce myself at the quality Ecomma needs is a clean growth engine: proper attribution, clean UTMs, structured testing on paid, a working lifecycle motion, the first principled view on CAC and payback. That's the first seat that stops being me plus AI and starts being a peer.
Rough size at end of Now: 2 FT. No fractional. No agencies. One person plus the CMO, an opinionated AI stack, and a short list of per-project contractors the CMO calls in by name only for truly one-off work.
There's more than one defensible shape. Below are three options, each shown across the first two horizons (Now and Next). AI tools are drawn as members of each function, next to the human who owns them, not as a separate layer. The FT count reflects humans only.
My recommendation is Option B. One growth generalist who owns paid, lifecycle and measurement together, with content following in the second phase once the numbers are legible. Options A and C exist because Ecomma may have reasons I don't yet see to weight the first hires differently, and I'd rather show the tradeoff than hide it.
Performance first. Content follows. Scale follows evidence.
One senior T-shaped operator covers paid, lifecycle and measurement. Content follows. Thinner bench, higher single-person risk.
Content first. Performance arrives once there is a story worth spending against. Pushes paid learning back by roughly six months.
| Pick Option A if | Pick Option B if | Pick Option C if |
|---|---|---|
| The first six months are a proof-of-traction story for investors or the board. Buyer-side intent is already there and needs clean measurement before scaling. | You have a candidate who is truly T-shaped across paid, lifecycle and measurement, and you accept the thinner bench and higher single-person risk that follows. | The brand is the asset. Exit buyers trust narrative more than paid reach, and content has to be credible before any performance number is worth reporting. |
Across all three options, the stack leans on roughly the same tools. Capability-by-capability, what a full-time hire would traditionally own, and how the stack covers it until a full-time hire is justified.
| Capability | Stack (tool type) | Owner |
|---|---|---|
| Writing, editorial, long-form | Large-language-model writing assistant | CMO + content manager |
| Website and landing pages | AI-assisted website builder | Content manager or growth generalist |
| Image and creative | Image-generation model + design tool | Content manager |
| Video and podcast editing | AI video editor | Content manager |
| SEO audit and keyword work | SEO audit tool + LLM | CMO (audit), content manager (ongoing) |
| Outbound and prospecting | Enrichment + sending infrastructure | Growth generalist |
| PR pitching | Media database + LLM drafts | CMO |
| Analytics and measurement | Product analytics + SQL assistant + Notion dashboard | Growth generalist |
| Deck production | AI deck tool + LLM | CMO + team |
| Paid + lifecycle | Native ad platforms + performance AI + Klaviyo AI | Growth generalist |
One principle: the stack does the work the work doesn't yet justify a person for. The moment a capability produces enough evidence to deserve a full-time owner, we hire against it. Not before.
By month six there is a brand, a measurement layer, a list of tests, a clear read on which channels are working. The second hire is a content manager.
Why content second, not growth: because the AI stack is strongest at content production and weakest at owning growth end-to-end. A human-plus-stack content manager will compound fast, because the tooling is there. A human-plus-stack growth generalist was the necessary first seat because the stack is not yet where it needs to be on attribution, lifecycle orchestration and CRO.
The content manager owns editorial, landing page copy, the LinkedIn motion, and the seller-origination content surface. I continue to own the brand narrative and founder voice personally. They are an operator, not a team lead. They ship work.
Rough size at end of Next: 3 FT. CMO, growth generalist, content manager. No fractional, no retainer agencies. Per-project specialists only for genuinely one-off work.
I don't want to prescribe the 24-month shape right now, and I'd push back on any plan that did. The rate of tooling change between April 2026 and April 2028 is going to reshape what a marketing function even looks like. Roles that exist today may be compressed into workflows. Roles that don't exist yet (agent orchestrator, model-ops for growth, something else) may become obvious by month nine. Committing to a 24-month org chart in month zero is a guess dressed up as a plan.
My preference: no 24-month shape committed in advance. We pick the next hire in month nine against the evidence we have at month nine, and the hire after that against month fifteen. Every seat triggered by proof, not by plan.
I'd oblige, and I'd caveat it heavily. Something like the following, with the understanding that it will almost certainly be wrong in its specifics:
This sketch is a contingency for a conversation, not a plan I'd execute against. The honest answer is: I'll know in month nine what the right month-fifteen hire is, and I don't want to pretend otherwise today.
Under five FT means every hire carries real weight. At 2 FT in Now, the one hire I make is the team. So the profile matters more than the title.
Most of what breaks small marketing teams is not bad execution. It's bad taste decisions made with confidence. So the hiring bar is highest on judgment, then on ownership, then on craft.
I'd rather hire a former founder who ran performance themselves than a VP of Performance Marketing from a 500-person company. I'd rather hire an operator who built a startup's measurement stack from zero than a director who managed a team that did. Scale inside someone else's shape is cheap evidence. Building from zero is expensive evidence.
| Horizon | Hire | Trigger |
|---|---|---|
| Month 0 | CMO starts | Offer accepted |
| Month 1 to 3 | Senior growth generalist (paid, lifecycle, measurement) | After the first 30-day plan is signed off |
| Month 7 to 9 | Content manager | Evidence from the first two quarters of performance work |
| Month 10 onwards | Next seat, if justified | Decided against month-nine evidence. Not pre-committed. |
Every hire after the first is triggered by proof from the horizon before it. No speculative seats.
These would change the shape, so I'd rather surface them now than guess.
This is a shape, not a staffing plan. The method is the important part: hire where it compounds, let the stack cover the rest, and don't pretend to predict past the horizon the evidence covers.
Under five people at 24 months, AI-ready from day one, no manager layer. Performance first, content second, the rest against evidence. If that shape holds up to pressure, I'm excited to build it.
If it's useful, the next step is turning the Now horizon into a two-page hire brief for the first seat. That's a thirty-day piece of work and a reasonable gating item before we agree on anything bigger.
Happy to pressure-test any of this.
Wassim