You're 2 quotes deep and both fees look small sitting next to your budget. What decides your real cost is not the rate printed on the page, it's whether the bill is tied to your spend or fixed while your spend moves. This page prices both structures across 12 months of growth, names the costs that never appear on the fee line, and says plainly which one wins at your number.
A percentage is a bill indexed to how well you do
Nothing about the service changes between a $20,000 month and a $200,000 month. Same account, same support, and whatever setup work existed happened once, in week 1. The only variable that moved is how good you got at your job, and on a percentage the invoice is wired to it. That's the part that slides past you on the call, because 2% sounds like a rounding error while a monthly figure sounds like a commitment.
Rates in this market run anywhere from 0.5% to 20%, so borrow nobody's worked example but your own. At 2% on delivered spend, with nothing charged on top, a steady year looks like this.
- $20,000 a month costs $400 a month, so $4,800 across the year.
- $50,000 a month costs $1,000 a month, so $12,000 across the year.
- $150,000 a month costs $3,000 a month, so $36,000 across the year.
- $400,000 a month costs $8,000 a month, so $96,000 across the year.
What the percentage is charged on matters more than the rate
2 providers can both quote 2% and send invoices that differ by thousands in the same month, because the rate gets multiplied by a base that almost nobody writes down. Delivered spend means you're charged on ads that actually ran. Funded amount means you're charged on every dollar you move onto the account, including the balance that sits idle for 3 weeks while creative is stuck in review, and including money that eventually comes back to you.
There is also a version where the fee comes off the top of your transfer, so the quoted rate and the rate you pay on usable balance are 2 different numbers. Get the base named in writing before anything moves. The full arithmetic on each base, and the 9 questions that settle it takes about 10 minutes to work through, and it's the highest paid 10 minutes in this whole decision.
Price the year you're growing into, not the month you're in
A comparison run at today's spend flatters the percentage, because today is the smallest month you plan to have. Price the ramp instead. Take an account that opens at $20,000 in month 1 and reaches $150,000 by month 12, stepping up the way a working offer actually does: $20,000, $25,000, $30,000, $40,000, $50,000, $65,000, $80,000, $95,000, $110,000, $125,000, $140,000, $150,000. That is $930,000 of spend across the year.
At 2% on delivered spend, the fee for that year is $18,600. Against a $995 flat monthly rate, the same year costs $11,940. The gap is $6,660, and the shape of it matters more than the total. Across months 1 to 6 the percentage bills $4,600 against $5,970 flat, so the percentage is winning and you feel clever for taking it. Across months 7 to 12 it bills $14,000 against the same $5,970.
Put your own forecast in place of that ramp and the pattern holds, because the pattern is arithmetic rather than opinion. The decision gets made in month 2 on the strength of the month 2 invoice, and it gets paid for in month 11, in the exact season you were working toward. That is the whole complaint about percentage pricing, stated without drama: it takes its raise on your results, at the point where your margin is already committed to inventory, creative and headcount.
The crossover is 1 division, and the row you use has to fit your spend
A flat monthly rate and a percentage cross at exactly 1 number: the flat rate divided by the rate. Below your crossover the percentage is genuinely cheaper and you should take it. Above it, the identical service starts charging you for growing. Against the flat monthly rates you're most likely to be shown, at a quoted 2%, the crossover sits here.
- $299 a month is matched by 2% at $14,950 of monthly spend.
- $599 a month is matched by 2% at $29,950 of monthly spend.
- $995 a month is matched by 2% at $49,750 of monthly spend.
- $1,995 a month is matched by 2% at $99,750 of monthly spend.
Compare the tier that fits your spend, not the cheapest row
Flat tiers carry spend ceilings, and a cheap monthly rate you cannot spend through is not a comparison, it's a trap you set for yourself. On the flat structure those 4 numbers come from, $299 a month carries a $10,000 monthly spend ceiling and $599 carries $35,000, while the $995 and $1,995 tiers carry none. So if you're running $60,000 a month, the honest row to price against is $995, and your crossover is $49,750 rather than the $14,950 the cheapest row would have told you. Any comparison that prints a crossover without naming the ceiling attached to it is quoting you a number you cannot use.
Do the same check on the percentage side. A rate with a monthly minimum attached behaves like a flat fee in every month you spend under the minimum, which is exactly the month you were relying on it not to. And check whether the rate is quoted per ad account, because a structure billed per account prices very differently the moment you run more than 1. What providers in this market actually charge and how the structures differ is the sheet to build before you take another call.
The costs that never show up on the fee line
The rate is the part you remember from the call. These are the parts you pay, and the first 1 is routinely larger than the fee itself.
- Float. On a prefunded account, your buffer sits inside somebody else's balance. Spending $100,000 a month is about $3,300 a day, so a 7 day buffer parks roughly $23,000 of working capital where you cannot reach it on a Sunday night, and it stays parked for as long as you keep running.
- Per account charges. If the fee or the minimum is billed per ad account and you're running 3 to spread risk, you pay the structure 3 times to solve a problem the structure created.
- Dead months. A monthly minimum turns a paused month into a bill for nothing. Ask what you owe in a month where you spend $0, and get the number rather than a reassurance.
- The warm up. A fresh account carries a low daily spend limit and the fee runs for the whole time it steps up. How spend limits are set and raised decides how long you pay for capacity you don't have yet, so ask for the current daily limit on the specific account you're being handed, not the tier average.
- FX. If you fund in a currency the account does not bill in, the spread is a fee with nobody's name on it. Ask what it is against the mid market rate on the day, and treat a refusal to name it as the answer.
Where a percentage is honestly the cheaper deal
A percentage isn't a scam, and anyone telling you it is has something to sell. Under roughly $15,000 a month, a low single digit rate charged on delivered spend costs less than a flat monthly rate, and it moves the cost of a slow month off your books entirely. If you're testing a new offer, running seasonal, or genuinely unsure whether you'll spend anything in March, that's the structure that fits your business and nobody should talk you out of it.
There's a softer argument for it too. A provider taking a cut of your spend has a live financial reason to keep you spending, and that's worth something at 9pm on a Friday when you need somebody to pick up. The structure stops being fair at the point your spend crosses the flat rate divided by the rate, and it gets worse every month after, because the thing inflating the invoice is the thing you're working to grow.
What you give up on an agency account, either way
Both structures on this page are agency accounts, so these tradeoffs apply to a flat rate exactly as much as to a percentage. A comparison that skips them is selling, not comparing.
- The account isn't yours. If the arrangement ends, access ends with it, on somebody else's timetable rather than yours.
- You sit inside a portfolio. Other advertisers on the same structure are not your call, and their behaviour is part of your environment.
- The pixel, the page and the audiences need a written answer before you fund anything. Ask what specifically leaves with you if you stop paying next month, and get it as text you can reread.
- A prefunded balance sits behind the same review that disables an account. Decide what you do on that day before it arrives, the same way you would plan for an asset going down mid campaign.
- No structure makes an account immune to enforcement. What a structure changes is what an enforcement event costs you and how quickly a replacement is running.
Going it alone removes the fee and hands you a delivery problem
The instinctive answer to a fee you resent is to open a fresh Business Manager in your own name and cut the provider out. It works, in the narrow sense that the fee disappears. What arrives instead is a brand new account with no payment history, no delivered spend behind it, and a daily limit low enough that your first campaign cannot clear the learning phase at the budget it needs.
Delivery on Meta is bought with history, and history is the 1 thing a new setup cannot have on day 1. You can build it, and plenty of operators do, but that's a project measured in months at reduced volume and reduced margin, not a way to save $2,000 next month. An account issued off a portfolio spending about $50 million a month across Meta already carries the payment record and the trust tier a new one is trying to earn, and it's live in 1 to 2 hours.
So the real question is never whether you could build history yourself. It's what the months of running under a new account's limits cost you against a monthly fee, and for most offers worth running that arithmetic is not close. Price the gap before you decide the fee is the expensive part of this.
Price 12 months before you sign 1
The honest part first. A flat rate does not make the account yours, it does not make it immune to enforcement, and it does nothing for you at all while your spend sits below your crossover. If you're running under about $15,000 a month with no plan to move, take the percentage, keep the invoice small, and revisit it the month you cross your own number.
Past that point the shape of the bill is the whole decision. Build 12 rows for 12 months, put the percentage quote in 1 block and a flat monthly figure in the block underneath, at the tier that actually covers your spend rather than the cheapest one. The percentage block is a slope. The flat block is a straight line: $299, $599, $995 or $1,995 a month with 0% of spend taken, identical whether you spend $20,000 or $400,000, on an account issued in 1 to 2 hours instead of warmed from zero. On the top tier your own card sits on the account, so Meta bills you directly and no balance of yours is held anywhere.
That's the trade, stated both ways. You accept that the account is not yours and that you sit inside somebody else's portfolio. In exchange you get spend history you didn't have to build, a replacement when an account goes down, and an invoice that stops moving the month your revenue starts to. A 7 day trial before anything is billed is the cheapest way to test whether an account behaves the way the pitch says, and it's worth doing before the sheet rather than after, because the daily spend limit sitting on the specific account you're handed will shape your next 30 days more than the fee structure printed above it.