Before you hire a marketer, write down the job

Most marketing job descriptions get written before anyone has described the job.

I've watched it happen more times than I can count. Something slips. Maybe the newsletter goes quiet for a quarter. Maybe the website hasn't changed since the last rebrand. Someone in the leadership meeting says, "We need someone for this." A week later there's a draft job post on the table.

That feels decisive. It's actually a guess.

The guess skips the one question that decides whether the money works. What exactly needs to get done, and what kind of help fits it?

Why "hire someone" is the first answer in the room

Hiring is the fix everyone knows. You can draw it on an org chart. You know how to budget for it, how to interview for it, and who it reports to. Agencies and freelancers come next. That's why most advice on this topic is a two-way fight: in-house or agency, agency or freelancer.

The need itself is much harder to see. Take "our content isn't going out." You might be missing a writer. You might be missing someone who knows what to write. Or the writer you have might be buried, and the blog is always the thing that gets bumped. Each of those calls for a different fix. In a meeting, they all sound the same.

I know this default from the inside. I spent about three years as a founder building alone. I coded our web app and both mobile apps myself. When you build alone, every gap gets the same answer: you'll handle it. That's a resourcing decision too. You just never make it on purpose.

The stakes are real. Gartner surveyed 402 marketing leaders for its 2025 CMO Spend Survey. In-house labor came to about 22% of their marketing budgets, and agencies to about 21%, according to Marketing Brew. And 39% of those CMOs planned to cut each one.

That survey leans toward big companies. The lesson travels anyway. People and outside partners are two of the biggest lines in any marketing budget, and both are under a microscope.

So the call deserves more than a reflex.

Decide one need at a time

Most advice treats resourcing as a company-wide model. Pick in-house, or an agency, or fractional, and build around it.

Real companies don't work that way. One company has several needs at once, and they rarely look alike. A mid-sized firm might need someone to own email every week. It might need a specialist to rebuild its paid search account once. And it might need senior judgment on positioning for a few hours a month. That's three different kinds of help inside one marketing function.

So decide one need at a time. Name it in one sentence, in plain words: what has to get done, and why it matters now.

"We need marketing help" is too broad to act on. "Our sales team has no current case studies for our two biggest service lines" is specific enough to work with.

Once the need is on paper, you can start to see what kind of help it calls for.

Four questions that size up any need

Run these against the one need you just wrote down. Answer them before anyone starts thinking about who to call.

1. How often does this need doing?

Is it a one-time project with an end date? A recurring task that takes part of someone's week? Or an always-on job that never finishes?

A one-time project points away from a hire. An always-on job that needs daily attention and deep knowledge of the business points toward one. The middle ground is recurring, part-time work. That's where freelancers and agencies usually earn their keep.

2. How deep is the skill?

Does the work need a good generalist? One deep skill? Several skills working together? Or senior strategic judgment?

A good internal marketer brings general competence. One deep skill, needed for a stretch, is classic freelancer work. Several skills at once is where an agency's bench pays off. Think creative, media buying, and analytics all working on the same campaign.

Sometimes the real gap is direction. Nobody has decided which markets, which message, or which channels come first. That calls for a senior leader. Few mid-sized companies need that judgment 40 hours a week, which is why a fractional CMO often fits.

3. Who's going to steer it?

Every option needs managing. A freelancer needs someone who can brief them and say what good looks like. An agency needs a point person who holds it accountable. A new hire needs a manager who understands the work well enough to coach it.

If nobody on your team has that capacity, a resource that needs direction will drift. That's usually a sign the missing piece is leadership. More hands won't fix it.

4. What budget do you actually have?

Use the real number set aside for this need. Do it before you look at what anything costs. That number filters the options faster than any comparison chart.

If the budget can't carry a full-time salary plus overhead, the hire is off the table for now. If it's well below an agency's minimums, so is the agency.

Notice that all four questions are about the need. None of them is about the options. By the time you've answered them, one or two options usually stand out.

Sometimes the answer is a better process

There's one option that rarely comes up when people debate a hire versus an agency. It's also the cheapest one to test.

Here's a pattern I see a lot. A company keeps missing content deadlines. The team knows what to write. Nobody owns getting it out the door. So the blog post waits for whoever has time that week. Most weeks, that's no one.

Hiring a content marketer into that situation adds a person to a broken loop. An agency is more firepower than the job needs. What fits is a process: one named owner, a calendar with real dates, and a 15-minute check-in each week.

It's like a car that won't start because the battery cable is loose. You could buy a new car. Or you could tighten the cable.

So before you resource anything, ask one more question: do we already have the people and skills, and is the work stalling anyway?

If the answer is yes, start there. A process fix costs some management attention. You can try it this month without signing anything. And if it doesn't work, you'll know exactly where the gap is. That makes whatever you buy next a better bet.

The limit is just as clear. A process can't create skills you don't have. It can't create hours that don't exist either.

Why this is harder than it looks

The four questions and the process check will get you most of the way. In practice, three things trip people up.

The first is that everyone at the table has a favorite. The CEO who got burned by an agency once leans toward a hire. The marketing manager who's drowning leans toward anything that adds hands. Being honest means rating all five options against the need. That includes the ones you already like and the ones you've already ruled out.

The second is the close call. Sometimes a freelancer and an agency both fit. The temptation is to force a winner so the meeting can end. The honest answer is "these two are close, and here's what separates them."

The third is that the answer depends on facts that change. The one-time website project turns into an ongoing content program. Now the right kind of help is different too. Few teams write down which fact would flip their decision. So when that fact changes, the decision quietly goes stale.

What you want at that point is a second opinion. It should take the need as you describe it and rate all five options against the same four questions, every time. And it should tell you plainly what would change its mind.

That's the job of Should We Hire Marketing?, a free, open-source Claude skill we built for this exact decision. You describe one specific need in your own words. Then you answer how ongoing it is and how specialized it is. Budget and timing are optional, and they sharpen the answer.

It rates an internal hire, an agency, a freelancer, a fractional marketing leader, and a better internal process against your need. Then it names the best fit. When two options are close, it names a real runner-up. And it tells you the one fact that would tip the answer somewhere else.

It recommends a type of help. It rates fit against the budget you give it and doesn't estimate salaries or fees. Picking the actual vendor or candidate is still your call, and you'll make it better once the type is settled.

Write the need down before the job post

The most expensive marketing help is the right help for the wrong need. A great hire assigned to a process problem tends to get frustrated and leave. A great agency hired for a one-person job burns a chunk of its retainer on overhead.

So the next time someone says "we need to hire someone for this," ask for one sentence first. What exactly needs to get done, and why now? Then run the four questions. Then check whether it's really a process problem.

You might still post the job. That's fine. You'll just know why, and you'll know what would change your mind.