Arcane Marketing logoArcaneApply
← All posts

You are spending $45,000 a month across Meta and Google, the contractor who has been running it just gave two weeks notice, and two tabs are open on your laptop: a draft job post for a media buyer, and a proposal from an agency. Most of the in house media buyer vs agency decision gets made in exactly that state, under time pressure, and it gets framed as one comparison: salary versus retainer.

That framing costs money twice. It underprices the hire, because salary is not what an employee costs you. It also overstates the choice, because the live question is which parts of the work belong inside the building, and very few companies answer that with all of it or none of it.

Here is the sequence I use to get an owner to a defensible answer: price the hire the way payroll prices it, decide what you are keeping, count the calendar, then look at what can go wrong on each side and which of those failures survive a contract.

Price the hire the way payroll does

The number on the job board is the smallest piece of the commitment. By the time someone is actually doing work, you are carrying employer payroll taxes, benefits, software seats, hardware, recruiting cost, and your own hours spent managing them.

Be careful with published salary data here. Federal wage data has no occupation called media buyer. The figures people quote in this argument describe marketing managers, or advertising and promotions managers, both of which are wider jobs than running your ad accounts. They tell you something about the hiring market and nothing precise about this role, so build your own loaded number.

Run the model on your own numbers

Assumptions first, and every one is a placeholder for yours. Say you spend $45,000 a month in paid media and you want one mid-level buyer owning Meta, Google and weekly reporting. Put the base at $110,000, the offer you would genuinely make in your market. Assume payroll taxes, benefits and equipment add 30% on top, a multiple you should pull off your last offer letter and your broker's quote rather than borrow from me. That gives $143,000 a year, about $11,900 a month. Add $600 a month for the tooling and data seats the buyer will ask for in week two. Add your own supervision at four hours a week, valued at an internal rate of $150 an hour, so $2,400 a month. Running total: roughly $14,900 a month before a dollar of it reaches an auction, which is about a third of the media budget it manages.

Now price the other side the same way. Retainer, plus anything billed separately for creative, plus the hours your team still spends on briefs, approvals and report review, because no ad account runs unattended.

That arithmetic is a model, not a finding, and the winner of the comparison is the least interesting output. At most spend levels the two columns land close enough that this quarter's cheaper option flips next quarter. What the exercise does is drag the hidden lines onto both sides of the page, and it exposes the structural difference between them: a salary is a fixed commitment you carry through a slow quarter, while a retainer usually ends on 30 to 90 days notice. Our breakdown of fully loaded in-house costs goes further on the employment side, and what each fee model rewards does the same for the retainer.

The in house media buyer vs agency split most companies land on

In-housing is the direction of travel and has been for years. The ANA reports that 82 percent of organizations now operate with an in-house agency, driven largely by the need for creative control and greater agility, and in its latest benchmark study on in-housing, jurors are five times more likely to say marketers are in-housing more than ever than to say they are pulling back. Read that first number carefully, though. Operating an in-house team is a long way from doing everything inside it, and The CMO Survey still has companies outsourcing one-third of digital marketing activities. Both findings hold at once, which tells you the real question is which third leaves the building.

The ANA also finds that 53 percent agree the primary role of an in-house agency is that of a strategic partner in upstream strategy and bold, brand-building work. That describes the role in-house teams get given. It says nothing about whether they produce better numbers than an outside team, and none of the data in this post measures output on either side.

For a founder-led business, the division usually falls along ownership of the money. The offer, the price, the follow-up sequence, the CRM, the sales floor, and the reason last month's leads did not close belong to you permanently, whoever is buying the media. Execution capacity is the part it is reasonable to rent: creative production volume, platform work, a second read on a test. If you are unsure which side strategy sits on, fractional CMO vs marketing agency draws that line.

Count the calendar before you post the job

Hiring has a lead time and a supply problem, and both land before anyone touches your account. SHRM puts 2025 median time-to-fill at roughly a month and a half from job requisition to offer acceptance. Add a notice period at their current job, then a ramp while they learn your offer and your tracking. More than half of recruiting executives, 56%, identified talent shortages as a challenge, and only 7% of nonexecutive positions were filled internally in 2025, so the promote-someone-internally plan is rarer in practice than it is in planning meetings.

Scope is the other thing the calendar hides. The most cited capability gap in The CMO Survey is about resourcing rather than talent: 22.3% of marketers say existing capabilities simply lack the people, time, and budget needed to function effectively. Headcount is not arriving to close that gap either, with growth slowing to 2.5% over the past year, down sharply from 5.4%, and a similar 2.6% projected for the year ahead.

So write the job as a list of weekly outputs before you write the job post: campaign builds, creative briefs, new ad volume, landing page tests, tracking fixes, the weekly report, the monthly read. Count the hours honestly. If the list does not fit inside one person's week, your new hire will quietly drop whatever is least visible, and that is almost always creative volume. The ad creative testing framework is a fair yardstick for how much throughput your account really needs.

Risk on each side, and which kind survives a contract

Right now the budget pressure in the market points at the agency line first. In MarTech's write-up of the Gartner 2025 CMO Spend Survey, 39% said they plan to cut back on agency spending by eliminating unproductive agency relationships, streamlining agency rosters and renegotiating contracts and scopes of work, while marketing budgets flat-lined at 7.7% of revenue and paid media held 30.6% of marketing budgets, or 2.4% of company revenue. That panel is large-company CMOs, so read it as a description of how the biggest advertisers behave rather than a target for your business; marketing budget percentage of revenue covers why those published percentages travel so badly. The signal worth taking is that scrutiny of outside spend is now standard, which argues for tighter scopes and written accountability.

One agency risk gets badly overstated: platform lock-in. A Google Ads manager account lets you view and manage multiple Google Ads accounts from a single location, and users with administrative access can terminate the relationship at any time by unlinking their account. The technical door stays open. What actually traps owners is paperwork and process: who holds the ad account and the pixel, whose card is on the billing, who owns the raw creative files, whether anything is documented, what notice period you signed, what happens to historical data on exit. Settle all of it at signature; accountability SLAs and exit terms has the specific language, and when to fire a marketing agency covers the signals that it is time to use it.

The in-house version of the same problem is concentration. One buyer means one person's judgment on every test read, nobody checking when a channel starts flattering itself, and a full stop on paid activity the day they resign. That is manageable if you plan for it: make documentation a weekly deliverable, keep the accounts in your own business manager, and learn to read the numbers yourself. Reading agency reporting like an operator works just as well on reporting from an employee.

Short version, if you only do two things. Build the loaded monthly cost of the hire before you talk to either side, then decide which slice of the work you are willing to send out and write that scope down. An owner who has done both can run a buyer or an agency and get value from either. An owner who skips them tends to swap between the two every nine months and blame the model for it. If you want another pair of eyes on which parts of your growth system should sit inside the building, apply here.

Sources

Stuck above the $10k ceiling?

We diagnose your account in 24 hours and tell you straight whether it's a creative, audience, or sales problem.

Talk to us