A percentage of ad spend agency charges you a cut of everything you put through the ad platforms, usually 10 to 20 percent, often on top of a monthly retainer. On paper that sounds fair. You pay more when you spend more, less when you spend less, and the agency has skin in the game. In practice the model hides a stack of costs that never show up on the invoice, and most of them work against you. Once you add them all up, the percentage is rarely the smallest number in the deal.
I have run ad accounts on both sides of this, and I want to walk through where the money actually goes, because the sticker price is the least of it.
The incentive is pointed the wrong way
Start with the obvious one that nobody likes to say out loud. When an agency earns a percentage of what you spend, their income goes up when your spend goes up. Your job is to get the most result for the least spend. Those two goals are not aligned, and over a long enough relationship the misalignment shows.
It rarely looks like fraud. It looks like a nudge to raise the budget going into a slow month. It looks like an extra campaign that "tests a new audience" and quietly adds to the monthly total. It looks like nobody being in a hurry to pause a mediocre performer, because a paused performer is spend that stopped earning the agency its cut. None of this requires anyone to be dishonest. The pay structure just tilts every judgment call in one direction, and small tilts compound over a year.
Compare that to how you would run it yourself. You would kill a weak ad the day it looked weak. You would pull budget out of a campaign that stalled and sit in cash until you found something better. An agency paid on spend has a quiet reason not to, and you are the one funding the difference.
The retainer usually comes first, then the percentage on top
Most percentage shops do not run on the percentage alone. There is a base retainer, often a few thousand dollars a month, and the percentage of spend sits on top of it. So the real question is not "what is 15 percent of my budget," it is "what is the retainer plus 15 percent of my budget, and what does that come to as a share of the money I actually have to work with."
Run the arithmetic on a real budget and it gets uncomfortable fast. Say you are spending $20,000 a month on ads. A $2,500 retainer plus 15 percent of spend is $5,500 a month to manage the account. That is more than a quarter of your ad budget going to management before a single customer is acquired. If you had put that $5,500 into the ads themselves, you would have had 27 percent more media in the market. That is not a rounding error. That is the difference between a channel that scales and one that stalls. We built a simple calculator that shows what agency fees cost you at your spend level if you want to see your own numbers.
Creative is almost always a separate bill
Here is the cost that catches people off guard. The percentage buys you management of the campaigns. It does not usually buy you the ads. Creative, the actual images, videos, and copy that determine whether any of this works, is a separate line item or a separate vendor entirely.
This matters more than the fee math because creative is what moves performance now. Meta and Google have automated most of the knob-turning that media buyers used to charge for. What is left, the part that genuinely changes your results, is the quality and volume of creative you feed the platform. So you end up paying a percentage for the commoditized part of the job and paying again, separately, for the part that actually matters. Two bills, and the expensive skill sits on the second one.
The lock-in you do not notice until you leave
The last hidden cost is the one you feel only when you try to walk away. When an agency runs everything inside their own ad account, their own reporting, and their own creative files, leaving means starting over. The learnings the platform accumulated live in their account, not yours. The creative was made by their team. You do not have clean access to the campaign history that took months to build.
That friction is a cost even if you never leave, because it weakens the one piece of leverage that keeps any vendor honest, which is the ease of firing them. A relationship you cannot exit is a relationship that stops trying to earn the renewal. Before you sign, ask a plain question. If I leave in six months, what do I keep. If the honest answer is "not much," that is a cost, whether or not it ever appears on a bill.
What the model should look like instead
Add the pieces up and the percentage of spend model asks you to pay more as you grow, pay a retainer regardless of results, pay again for the creative that does the real work, and accept an incentive that quietly favors bigger budgets over better ones. For a large brand with a big team and a big budget, an agency can still earn its keep. For most small and mid-sized businesses, the math stops working long before that.
The fix is not to give up on running ads. It is to fix the pricing so the people spending the money keep the upside. A flat fee decouples what you pay to manage from how much you spend, which means growing your budget no longer grows your management bill. Owning your own ad account and creative library removes the lock-in. And having the same system make the creative and run the campaigns collapses the two bills into one. If you want the fuller case for that structure, we laid it out in flat fee versus percentage media buyers, and if you are weighing an agency against software in general, what a media buyer actually costs breaks down the options.
This is the model we built Mebume around. It is a fully automated ads manager that makes your creatives and runs your Meta and Google campaigns 24/7, inside guardrails you set. It moves budget toward what works and pauses what does not, and optimizations can be reversed within 24 hours if you disagree with a call. It costs a flat $499 a month per business, never a percentage of your spend, so the incentive that bends every percentage deal simply is not there. It was built by operators who have spent over $500,000 a day on Meta, which is where the opinions in this post come from.
The percentage looks reasonable because it is framed as fair. Fair would be paying for the work, keeping what you own, and not being charged more for the crime of growing. Do that math on your own account before you renew.
Mebume launches soon. Join the waitlist to be first in when we open access.
