You lost an account, or you are about to launch something a fresh account will not survive, and somebody told you the grown up move is to register a Business Manager in your own name and run everything through it. It sounds right. It is also the answer to a different question than the one you have, because what decides whether your budget delivers tomorrow is not whose name sits on the account, it is what history sits behind it. This page works the trade in both directions, including what an agency account genuinely costs you.
The account does not decide delivery. Its history does.
Meta does not treat 2 accounts the same just because they are structured the same. Every account carries a record: verification status, how long it has been paying, whether those payments cleared, how much it has spent, and what happened the last time something went to review. That record sets the daily spend limit, the billing threshold, how fast a campaign clears review, and how much weight a single rejection carries. Identical creative on identical audiences will not behave identically on an account with years of cleared invoices behind it and one that opened last Tuesday.
This is the part the ownership argument skips. Who holds the keys, and what the platform has already learned about the entity paying the bills, are 2 different questions, and only the second one moves delivery. You can hold every key and still sit capped at a limit that has nothing to do with how good your offer is, because nothing has been proven yet. History is what is being priced, and it is bought with time and cleared payments, which is exactly why an account with a record behaves differently to a fresh one.
Be precise about what carries across and what does not. A pixel that has been firing for 2 years keeps its event history, and a page with real posting history keeps its age. Neither hands a new ad account a spend limit or a billing threshold, because those sit against the account and the entity being billed. Attach a well aged pixel to an account opened this morning and you can still be capped at a daily number that will not carry a campaign through the learning phase.
What registering one this week actually starts you with
None of this is a judgement on how well you would run it. Registering a Business Manager is not hard. The problem is the opening position. Every line the platform reads on a setup created today is empty, and the only thing that fills those lines is time with money moving through them.
That makes it slow rather than wrong, and slow is a cost you pay in the months where the offer was ready and the account was not. If you need to be spending this month on something that gets looked at hard, the choice is between an account that has already proven it pays its bills and one that will start proving it after your window has closed.
Put a real calendar against the steps. Business verification is a document review with a queue, and a mismatch between the name on your filing and the name on the document you upload sends you back to the start of it. The daily limit then steps up on billing cycles rather than on calendar days, so the clock only moves while payments are clearing. Before you forecast on any of it, read how spend limits are set and how they actually get raised, then ask what the account in front of you carries today. On a setup registered this week, it reads like this.
- No payment history, so the account opens on a low daily spend limit and a low billing threshold, and both step up on cleared payments rather than on how well the offer converts.
- No verification record, and that queue sits before your spending rather than after it.
- No review history, so the first rejection arrives with no context. On an account with a long clean record a rejection is 1 data point. On an empty account it is most of the file.
- Nothing standing beside it. If the account goes down in week 2, the appeal is the entire plan, and you are the one writing it and waiting on it.
- A calendar cost that caring more does not shorten, because verification, warmup and threshold steps run on the platform's clock.
What you give up on an agency account, plainly
Now the other direction, because anybody telling you this trade is free is telling you something you can disprove in 20 minutes.
- You are not the account holder. The account sits inside somebody else's portfolio, and the working relationship is part of what keeps you in it.
- On most structures the balance is prefunded, so a slice of your working capital lives somewhere you cannot reach at 10pm on a Sunday.
- Most providers in this market take a percentage of every dollar you spend, quoted anywhere from 0.5% to 20%, and that cost grows with you rather than with the work being done.
- What leaves with you at the end is exactly what was written down. Pixel, page, audiences, historical data: anything not named in the agreement is somebody else's decision later.
- Meta's published terms are written around the account holder being the party that agreed to them. What that means for an account issued to you is a question for your own counsel.
- Shared structure means other advertisers are a variable in your week, in a way they would not be on a setup nobody else touches.
Some of those are negotiable and some are structural
The percentage is negotiable, because flat rate structures exist. The prefunded balance is negotiable, because structures where your own card sits on the account exist. What you keep at the end is negotiable too, and almost nobody negotiates it, because they ask at the exit instead of at the start.
Ask at the start, in writing, and read the answer against what a documented handover actually includes. If a provider cannot say in 1 paragraph what you walk away with, you have your answer without a second meeting. Not being the account holder is the part that is genuinely structural, and no wording fixes it. Price that one honestly, and stop treating the rest as if they were fixed too.
What the portfolio behind the account is doing for you
The reason an agency account behaves differently is not a trick, it is volume. An account issued off a portfolio spending about $50 million a month on Meta sits behind an established billing relationship and a payment record no new entity can manufacture at any price. That record is what gets read when your campaign goes to review and when the system decides what it will let you spend, so the limit conversation has mostly happened before the account reaches you.
It also changes what a bad day costs. On a single setup of your own, an enforcement event is a full stop followed by an appeal on a timeline nobody controls. On a portfolio, a replacement account gets issued and the structure gets rebuilt on it. Nothing about this makes an account immune to enforcement, and anybody promising that is selling something nobody can deliver. What changes is the length of the outage, which is the number that reaches your P&L. Write down a plan for an asset going down mid campaign either way, because both routes need one.
Speed at the front is the same mechanic running forwards. An account issued off an existing portfolio is live in 1 to 2 hours, because the trust it runs on was built before you arrived. A setup created from scratch runs to weeks before it spends at a level worth reading, and data off a throttled account tells you almost nothing about the offer. You end up killing creative that never got a fair test.
The fee structure is a separate argument, and it is the one worth having
Most of the anger pointed at agency accounts is actually pointed at percentage pricing. The 2 get bundled together because the providers charging a cut are the loudest in the market. Separate them. Whose portfolio the account sits in, and how you get billed for using it, are independent decisions, and only 1 of them scales against you.
A percentage is charged on every dollar, every month, and it grows with the exact variable you are working to grow. The service is identical at $30,000 a month and at $300,000 a month. A flat monthly rate does not move at all. The 2 cross at 1 number, the flat rate divided by the rate, and below that spend the percentage is genuinely cheaper and nobody should talk you out of it. Above it the same deal becomes a tax on doing your job well. Work yours out before you argue about it, because the arithmetic on a percentage cut at real spend levels is not close to what the pitch implies.
This matters here because keeping 100% of your budget usually gets filed under ownership. It does not belong there. It is a pricing choice, and you can have it on an account you did not register yourself.
The narrow case where doing it yourself costs you nothing
There is a real case for standing one up, and it is narrower than it sounds, because all 5 of these have to be true at the same time. Note what is not on the list. None of them say the setup will deliver better. They say the delay is free.
- Your offer sits in a category that rarely gets looked at hard, so review is not the thing costing you money.
- You have 2 to 3 months where being slow costs nothing, with no launch date and no inventory sitting inside them.
- Somebody in house owns verification, domain and pixel configuration as their job rather than as their evening.
- You can absorb an outage with nothing standing by, because the appeal will be yours to write and yours to wait on.
- You are running 1 or 2 brands rather than 6, so the setup work does not repeat every time you launch.
What the trade actually comes down to
In DTC those 5 are rarely all true at once. The operator reading this has an offer that gets reviewed hard, payroll that runs whether ads deliver or not, and a date. So take the honest trade off first: you are not the account holder, somebody else's standing with the platform is a dependency in your business, the terms question stays open, and what you keep at the end is exactly what you got in writing. Those are real, and they do not disappear because the accounts perform.
What comes back is the only thing that moves delivery in week 1: a payment record the platform already trusts, live in 1 to 2 hours rather than built across months, with a replacement path for the day something goes wrong. For anybody who needs to be spending now, that trade is not close.
Then price the structure separately, because that is where the money quietly is. A flat monthly rate with no cut of spend runs $299, $599, $995 and $1,995 a month, and the invoice is identical whether you spend $20,000 or $400,000. On the top tier the account sits in a dedicated Business Manager with your own card on it, so Meta bills you directly and no balance of yours is held anywhere else. Put that next to a percentage on your own forecast and the comparison stops being a debate about principle. It turns into arithmetic.