Should an edtech startup hire in-house, an agency, or a fractional CMO?

These three options solve different problems, and the mistake is choosing between them on cost. The question that actually decides it is whether you already know what works.

If you do, you need volume and consistency, which favours hiring. If you do not, you need range and the ability to change your mind, which favours an agency. If you know what works but not what to do next, that is a strategy gap, which is what a fractional CMO is for.

Hire in-house when the channel is proven

A full-time marketer is the best option once you can name the channel that produces users and the job is to do more of it, better, every week.

In-house wins on context. Someone in your standups hears the support complaints, knows which feature is shipping, and can rewrite a page the afternoon it becomes wrong. No external team will match that, and on a proven channel that daily proximity is most of the value.

The cost is not the salary. It is the commitment to a guess. A content hire assumes content is the answer. A performance hire assumes paid is. If you are wrong, the correction takes a resignation, a new search and two quarters, and early-stage companies are frequently wrong about this.

Hire in-house when most of these are true:

• One channel is clearly producing users and the constraint is volume.
• The work needs daily attention rather than weekly.
• You can describe the role in one sentence without the word "and".
• You can afford to be wrong about the choice for two quarters.

Use an agency when you need range, not headcount

Early edtech marketing needs several disciplines in unpredictable amounts. One month is a site rebuild. The next is a cohort launch. The next is fixing the enrolment process that three people are running manually in a spreadsheet.

Hiring for that mix means hiring three people, or hiring one generalist who will be excellent at one of the three. An agency exists to cover the range without the headcount, and the honest reason it works at this stage is that you can change what you are buying next month without changing anyone's job.

What you give up is proximity. An external team is not in your standups and will always know your product less well than you do. Good agencies narrow that gap. None of them close it.

Use an agency when:

• The work spans disciplines that no single hire covers well.
• Monthly needs vary, and locking into one specialism would be premature.
• You need to start in weeks rather than months.
• You want the option to stop without a redundancy process.

Use a fractional CMO when the problem is the plan

A fractional CMO is senior judgement bought by the day: what to do, in what order, what to stop, what to measure, and whether the team you have can deliver it.

It is the right call when you have execution capacity that is busy but directionless, or when a founder has been running marketing by instinct and can no longer tell which of five active projects matters.

It goes wrong in one specific way, which is hiring strategy without execution. A fractional CMO produces a plan. If nobody has the hours to build what the plan describes, you have bought a document. This is the most common failure of the three and it is entirely predictable at the point of hiring.

Compare total cost, not rates

The comparison people run is salary against retainer, which flatters the hire. A fairer version includes everything the hire actually costs.

A full-time marketer carries salary, employment costs, equipment, tooling, recruitment, and the management attention of whoever they report to, which at a small company is usually a founder. It also carries a hiring timeline that is realistically two to four months before anyone starts, plus a ramp.

An agency costs more per hour and typically less per year, starts inside a month, and stops when you stop. That flexibility is the product. You are paying a premium for optionality, and the premium is only worth it while the optionality has value, which is to say before you know what works.

Once a channel is proven and needs a full week of attention every week, the arithmetic reverses and you should hire. Agencies who do not tell you this are not worth keeping.

The combination that usually works

Most edtech companies past seed end up with one in-house marketer and one external team, and the split that works is by cadence rather than by discipline.

The in-house person owns anything that needs daily attention: the channel that works, the community, the calendar, the priorities, and the product context that only comes from being in the building. The agency covers what is needed in bursts: design, site and landing page work, launch campaigns, automation builds, and the periodic strategic reset.

This fails when it is split by discipline instead of cadence, because two parties owning "marketing" jointly means neither owns the number.

A short decision test

• Can you name the channel that produces users, with numbers? If yes, hire for it. If no, do not hire yet.
• Does the next six months need three disciplines in varying amounts? If yes, an agency is the cheaper way to buy that.
• Do you have people who can execute but no agreed plan? That is the fractional CMO case, and only that.
• Is one function at capacity every week, permanently? Hire, regardless of what anything above says.

We are an agency, so read the second point with that in mind. The cases where we tell companies not to hire us are the first and last ones, and they come up often enough to be worth saying in public.

Common questions