The agency vs. operator decision.
Most founders default to hiring an agency when what they need is an operator. Same monthly spend, very different outcomes. A framework I’ve walked dozens of founders through.
Twenty thousand dollars a month is the number where this decision gets serious. Below it, you can muddle through with freelancers and a founder doing marketing in the cracks. Above it, you have to pick: agency or operator.
Both can work. Both can also fail in predictable ways. The decision is not a coin flip — it’s a matching problem between what you actually need and what each model is built to deliver.
What you actually get from an agency
An agency sells you scoped deliverables and hours from a team you mostly don’t see. The good ones are excellent at specific things: a paid acquisition agency that runs Meta ads at scale, a creative agency that produces video at volume, an SEO agency that knows how to win for a specific vertical.
What you pay for: scope, output, and risk transfer. The agency carries the headcount, the tools, the bench of specialists. You don’t hire seven people to get all those skills. You write one check.
What you don’t get: someone who owns the marketing P&L. Someone in your numbers meeting. Someone who can hire your next marketer. Someone who reports to your CEO with the same accountability a VP would.
What you actually get from an operator
An operator (full-time, fractional, or interim) is the marketing version of a CFO — they own the function, set the strategy, hire the people, run the rhythm, and report to you with skin in the outcome.
What you pay for: ownership, judgement, accountability, and a leader who is integrated into your business.
What you don’t get: an army. An operator does not have a bench of fifteen designers and a programmatic media buying desk. For specialist execution they will either bring in an agency under their direction or hire it inside.
The honest comparison
The decision tree
Three questions, in order:
1. Do you know exactly what you need executed?
If yes — you have a defined channel, a clear scope, a KPI you can hand over, and you just need execution muscle — an agency is the right answer. Hire one that specializes, write a tight SOW, manage them well.
If no — if you’re still figuring out positioning, the right channel mix, the messaging hierarchy, or which of three growth bets to put weight behind — an agency will fail. Their model is built on knowing the scope going in. They will charge you to discover it, then sell you the execution they already had bench-ready. An operator works the other direction: figure out what to do, then build or hire the execution.
2. Do you need someone on your leadership team?
If your CEO wants a marketing partner in the weekly leadership meeting, in the board deck, in the strategic conversations — an agency will not show up to those. An operator will. This alone tips a lot of decisions.
3. Are you trying to leave a function behind, or a campaign?
Most founders want both, but you have to pick the primary. If you want a marketing function that compounds — a team, a playbook, a tech stack, a rhythm that survives when you change leaders — you need an operator. If you want a campaign to launch a product next quarter, an agency can ship it faster than anyone.
The hybrid model
The smartest move I’ve seen at the $5M to $20M stage is the hybrid: one operator (fractional or full-time) who owns the function, plus one or two specialist agencies they manage for execution depth they can’t hire for.
The operator sets the strategy, owns the P&L, hires the in-house team for the work that compounds (lifecycle, brand, content). They bring in a paid acquisition agency for media buying, because the right Meta or Google ads team is hard to hire full-time at this scale.
This is the model I’ve actually run at multiple companies. It works. The trick is the operator has to be senior enough to manage agencies, not be one.
Hiring an agency to lead your marketing is asking a vendor to do strategy. They will sell you scope, because that’s what they sell. Lead with an operator, and let agencies execute under them.
The common failure modes
I’ve been called in to clean up after both. The agency failure looks like: a polished quarterly report, a deck with no decisions in it, a paid account that hasn’t been touched creatively in six months, and a CEO who can’t answer what the marketing strategy actually is.
The operator failure looks like: a senior person hired too early, with no team or budget around them, who quits in four months because there’s no work for them to lead.
Both failures share a root cause: the founder hired the model instead of the work. The work tells you which model. Start there.
How to know which you need
If you can write down the next 90 days of marketing work in detail, and what you’re missing is execution capacity — hire an agency.
If you cannot write down those 90 days because the strategy itself is unclear, the channels aren’t owned, and the team isn’t built — hire an operator.
The fastest way to figure this out is a twenty-minute call with someone who has done both sides. That’s the call I run free for new clients. If we’re a fit, we move. If you need an agency, I’ll send you to one I trust.