Agency or In-House Media Buyer? Here Is the Real Math

Start Here

What I Learned Carrying 40 Accounts at Once

At my peak as a media buyer, working across a couple of agencies at the same time, I was carrying around 40 accounts. That sounds extreme, but it is common: a lot of agency work is done by contract media buyers stacking clients from multiple shops, and the contractor prioritizes whoever pays them the most.

Here is what actually happens at that load. Accounts get weekly touches, not daily ones. Budgets overspend and campaigns break or get ads rejected, and nobody notices until the next report gets pulled. The day of the client call, you go in, pull a synopsis, write up a story, present it, and often do not touch the account again until the next call. Action steps from the last call get done the day before the next one, because every day in between is spent putting out fires.

And still, here is the uncomfortable part: those few hours were sometimes worth more than a full-time hire. Not because I worked harder, but because I saw dozens of accounts at once. You learn seasonality: which holidays deflate results, which months get expensive. You learn what messaging works across different markets, how to craft a message instead of posting ads and hoping. A solo in-house buyer stares at one account and cannot tell a platform problem from a campaign problem.

That is the honest core of this decision. You are not choosing between cheap and expensive. You are choosing between dedicated hours and pattern recognition. Everything else is math, so let's do the math.

The Numbers

What Each Option Really Costs

Real salary numbers, from someone who has earned them and hired them. When I started in 2019, buyers came in around $40K. Today a starting buyer is more like $60K, senior buyers run $70K to $80K, and directors are six figures plus. It moves with experience, market, and how the agency or brand structures responsibilities, but those are honest ranges. For calibration against official data: the Bureau of Labor Statistics puts the national median for marketing specialists, the closest government category to a media buyer, at $78,760 a year as of 2025, with a mean of $89,490. That blend includes corporate marketing roles in expensive metros, which is why hands-on ecommerce buyers tend to come in under it. Add taxes, benefits, software seats, and management time, and a good hire is realistically $6K to $9K a month all-in. Then add ramp: months before a new hire performs at full speed in your account, and the full cost of a bad hire is paying that ramp twice.

Agencies price in the ways we broke down in our agency pricing guide: retainers from $2K to $8K a month for one or two channels, percentage-of-spend deals at 10% to 20% of budget, and performance models that take a percentage of sales.

Notice the fourth row, because it has gotten dramatically stronger. On Meta specifically, the platform now does most of what a junior buyer used to do. A brand can set up a broad-targeted Advantage+ campaign optimized for purchase volume, drop in the post IDs of its top organic posts, use creator partnership posts where it has them, and honestly crush it with just that. If your best content already earns attention organically, the machine knows what to do with it.

The Failure Modes

How Each Path Goes Wrong

The hiring mistake I see most is not timing, it is misalignment. A brand brings in an agency or a buyer and says "run my ads," and the hire starts doing what worked for their other brands with no understanding of this brand's market, ICP, or North Star. The brand ends up overpaying for media buying and underpaying for the market research and strategy it actually needed first, the stuff that points the whole company in one direction.

Agency-side failure looks different. Clients of big agencies would be shocked at how infrequently their account gets looked at, and how much of "performance" is a story crafted the morning of the call. When I was Director of Paid Social, the pushback I got on reporting was explicit: do not bring up anything negative, lead with the good, have an answer ready if they ask. Add agency churn, where your buyer quits and someone junior inherits the account and the methodology changes overnight, and your results get volatile for reasons that have nothing to do with your business. The clients who do not push back get the bare minimum, because the incentives allow it.

Whoever you hire, demand the same three things:

• Results reported as contribution profit, not platform ROAS. If the scoreboard stops at the ad account, you will eventually pay for growth that loses money. We wrote up how good ROAS hides falling profit and the five leaks to check.
• Proactive communication. The best founders we work with are always moving the ball forward, asking what we need from them. The best partners do the same. An agency that waits for you to reach out is an agency doing weekly touches.
• A teacher's heart. Whoever you hire should want to show you how things work. We have built P&L tracking for clients when it was not part of the service, and taught founders how to vet future hires and how to tell if they are getting ripped off. Teaching makes us less necessary, and it makes clients trust us more. Someone who keeps the knowledge locked up keeps you as a hostage, not a client.

The Elephant in the Room

Where AI Leaves All of This

AI is already flooding this industry with agencies that have never operated anything, doing simple prompting and undercutting on price. Here is the problem with that: if you do not have the context and experience to begin with, you cannot tell whether what AI hands you is good enough or what the brand actually needs.

Experience does not get replaced here. It gets amplified. AI makes good operators dramatically faster and lets a founder with good taste do more than a junior media buyer could a few years ago. That raises the bar for what you should accept from any hire, in-house or agency: if they are just prompting and passing it through, you are paying a markup on something you could do yourself.