What A Bad Media Buyer Actually Costs You (Besides Their Fee)
The fee is the smallest thing a bad media buyer costs you. The real bill is in six other places: ad spend that kept running between their check-ins, creative that took weeks to refresh while performance decayed, audiences nobody rebuilt, a pixel and account history you may not even own, reports designed to make the month look fine, and the quarter you spent finding out. None of those show up on the invoice. All of them come out of your P&L. Here is how to add them up from your own account, without anyone's benchmark.
Why the fee misleads you
When an owner weighs a media buyer, the number in their head is the retainer or the percentage of spend. That number is visible, so it gets all the scrutiny. But the retainer is a cost of having a buyer. The costs below are costs of having a bad one, and they scale with your ad budget rather than with the fee.
The six costs that do not appear on the invoice
1. Spend that ran between check-ins
A human buyer with several clients checks your account when they get to it. Between checks, the losers keep running. Pull your ad set history and look for ad sets whose cost per result was clearly above your threshold for days before they were paused. Add up what they spent in those days. That is money a daily read with a written pause rule would have saved, and on most accounts it is the biggest line on this list. The buyer did not make a bad decision. The right decision was made late, repeatedly.
2. Slow creative
Creative fatigues. When the refresh cycle depends on a buyer who is also a bottleneck for three other accounts, your winners run past their peak while the new batch sits in a queue. Look at how long your best ad sets kept spending after frequency climbed and results fell. Multiply that window by the gap between what they returned at peak and what they returned at the end. That is the cost of a slow creative cycle.
3. Audience decay
Audiences need rebuilding: new buyers added to the exclusion so you stop paying to reach people who already bought, new buyers added to the seed so your lookalikes get sharper, list segments refreshed as the list grows. When this happens "when the buyer remembers," it happens rarely. Check when your exclusion audiences were last updated. If it was months ago, you have been paying to reach your own customers that whole time.
4. Assets you do not own
This is the expensive one and it is binary. If your campaigns run in the agency's ad account, or your pixel lives in their business manager, then when the relationship ends you lose the history, the audiences and the learning. You start the next phase from zero. Every other cost on this list is recoverable. This one is not, and it is the reason the first question in the framework is about who owns the account.
5. Reports that hide the problem
A report built to protect the relationship emphasises the metrics that look good this month and quietly drops the ones that do not. Impressions are up. Reach is up. Cost per result is on page three if it is there at all. The cost is decisions you did not make because you did not have the information: the offer you did not fix, the angle you did not kill, the budget you did not move. Compare the last three reports with the platform's own numbers and see what was left out.
6. The quarter it took to find out
Most owners do not discover a bad buyer in a week. They discover it when a quarter closes and the numbers do not reconcile with the reports. Everything above compounds for that whole period, plus the transition and ramp of whatever comes next. The honest window is not a month. It is from the point things started drifting to the point the new loop was running clean.
A worksheet from your own account
| Cost | Where to find your number |
|---|---|
| Spend between check-ins | Ad set history: spend on clear losers after the threshold was crossed, before pause |
| Slow creative | Days top ad sets ran past peak, times the drop in results per day |
| Audience decay | Spend reaching existing buyers since the exclusion was last updated |
| Assets not owned | If the account is not yours: the cost of restarting from zero |
| Hidden problems | Decisions delayed because the report omitted the metric, at your estimate of what the delay cost |
| Discovery window | All of the above, times the months from drift to clean replacement |
Fill it in. Then put the fee next to it. For most owners that settles the "is the fee reasonable" question, because the fee was never the issue.
What a good buyer costs, for contrast
A good buyer costs the fee and very little else. They check daily, refresh creative before it fatigues, rebuild audiences on a schedule, work inside your account, and raise problems before the numbers do. If that describes yours, this site is about moving the daily loop to agents so their time goes to strategy, not about firing them. The job-by-job comparison shows which parts of their work are which.
What agents change about this list
Look back at the six costs. Four of them, the spend between check-ins, the slow creative, the audience decay and the hidden problems, are costs of inattention. They exist because a human with other clients was not looking at your account every day, and because the report was written by the person being evaluated. Agents running inside written rules on your own account are always looking, refresh on a schedule, rebuild audiences on a schedule, and write the report in the same format every week regardless of how the week went. That is the structural fix, and it is why the replacement sequence in how to replace a media buyer with an AI agent starts with ownership and rules rather than with turning anything on.
The fifth cost, assets you do not own, is fixed by getting your account back, which is step one of that sequence. The sixth, the discovery window, shrinks to a week because someone on your team reads a weekly report that was not written to protect anyone's retainer.
What agents do not fix is the offer. If the offer does not sell, the daily loop runs perfectly and the numbers are still bad. Someone on your team still has to own that, and no buyer, human or otherwise, takes it off your plate. If you have run the worksheet and want to bring the loop in house with someone who already runs it this way, the application on the home page is where to start.
FAQ
How do I tell wasted ad spend from normal testing spend?
Testing spend is money put behind a hypothesis with a threshold for calling it. Wasted spend is money that kept running after the threshold was crossed because nobody was looking. Pull your ad set history and look for losers that ran for days past the point where the numbers were already clear.
Is a media buyer bad if my cost per lead went up?
Not necessarily. Costs rise with competition, seasonality and audience saturation regardless of who is running the account. The question is whether the buyer saw it coming, told you, and had a plan. A bad buyer is one you find out about from the numbers instead of from them.
What is the most expensive thing a bad media buyer can do?
Run your campaigns in an account you do not own or control. When the relationship ends you lose the pixel history, the audiences and the learning, and you start the next phase from zero. Everything else on the list is recoverable; this one is not.
Does replacing a bad buyer with AI agents remove these costs?
It removes the ones caused by inattention and dependency: the gaps between check-ins, the slow creative cycle, the forgotten audience refresh, the account you do not control. It does not remove the need for someone on your team to read the weekly report and own the offer. Budget for that person.