You submitted the appeal, the denial came back faster than a person could have read it, and now the Request Review button is greyed out or gone. Every guide you've found stops at the step you already completed. The second appeal isn't the first one sent again with more feeling. It's a different document, sent through a different door, read by a different reader, and this is what goes in it.
The first appeal is a form. The second is a case file.
The first appeal is triaged at speed. You get a short free text box, you type into it, and a decision comes back fast enough that nobody could have opened your ad account, looked at your ads and loaded your landing page in the time it took. That speed is useful information. It tells you the first pass was scored rather than read, and that whatever you typed was weighed against the same signals that produced the enforcement in the first place.
The second attempt is a different job. You're no longer arguing in the abstract that the system got it wrong. You're handing a person a file that lets them close a ticket safely. Reviewers aren't hunting for a reason to say yes. They need enough documented context that reinstating you is a defensible decision if anyone ever looks at the ticket again. Write for that, not for sympathy.
So the question stops being how to word it better and becomes what someone needs in front of them to sign off. Everything below follows from that. The triage in the first hours after the notice lands is covered in the recovery playbook for a disabled ad account. This piece picks up where that one ends.
Hour 0 to 6: capture the evidence before your access expires
Do this before you write a word of the appeal. When an asset is disabled, your access to it degrades over the following days. Ads Manager may load read only, then it may not load at all. Ad previews stop rendering. Creative you never downloaded lives only inside an account you can't open. The document you're about to write depends on artifacts that are quietly expiring.
Do it from a desktop browser and put everything in 1 folder named with the ad account ID. You're assembling a PDF out of it later, and hunting for a screenshot you took on a phone 3 days ago is how a file ends up with a hole in it.
- Screenshot the enforcement notice in full, including the date, the exact policy it cites and the asset ID it names. That wording is the most important input to everything that follows.
- Screenshot the Account Quality page for the asset while it still renders, including any individually rejected ads listed underneath the account level decision.
- Export the ad account at campaign, ad set and ad level, with ad IDs and the destination URL column switched on. If export has already stopped working, screenshot the ad list instead.
- Download every creative that was live in the 30 days before enforcement. The actual video and image files, not preview links.
- Export the payment activity. Settled charges against a real card in a real business name are the cheapest legitimacy evidence you have, and that export stops being reachable along with everything else.
- Save the landing page as a dated PDF, plus the checkout page, the terms page, the refund policy and the contact page.
- Write down the exact date and time of the enforcement in your own timezone and in UTC. A timeline that doesn't line up reads as reconstructed after the fact.
Work out which object was actually enforced
Most second appeals fail because they're aimed at the wrong thing. Meta enforces on separate objects, each with its own status surface and its own route, and the notice tells you which one it is if you read it as a noun instead of as bad news.
This matters more than the wording does. If a personal profile is restricted, the most polished ad account appeal you can write will still be denied, because the restriction isn't sitting on the ad account. That's also why swapping the admin is sometimes the entire fix and no appeal is needed at all. The same logic decides what you rebuild on afterwards. An asset that's been through enforcement and come back carries a different history to one that was clean from the start, and they behave differently under spend. The difference between a reinstated and a verified business manager is worth getting straight before you commit budget to either.
Here are the objects, and how each one announces itself in the notice.
- The ad account. The notice names an act_ ID and the status sits in Account Quality under that account.
- The business portfolio. The notice names the portfolio, and every ad account, Page and pixel inside it goes dark together. Appealing 1 ad account inside a restricted portfolio does nothing at all.
- The Page. A Page carries its own account status inside its Professional Dashboard, separate from ads entirely, and a restricted Page can stop an otherwise healthy ad account from delivering.
- The personal profile. When an admin's profile loses advertising access, that person carries the restriction into every asset they touch, including new ones. The appeal here is for the person, and it's filed from that profile.
- The payment method. A flagged card follows itself. Attaching the same instrument to a fresh account moves the flag across with it.
Where the second appeal goes once the button is gone
A denied review request usually greys out or removes the button on that asset. That's the end of 1 route, not the end of the routes.
- Account Quality, checked again later. The request option sometimes returns after an unrelated change completes, most often after business verification finishes or a Page problem underneath is cleared. Check it rather than assuming it's frozen.
- Business Support Home, reached from the help menu inside Business Manager. Whether you get a form, an email thread or a live chat depends on the portfolio you contact from, not on the asset that went down. Contact from a healthy portfolio you still control and reference the disabled asset by ID.
- The same surfaces, tried from a different admin. What you're offered depends on the profile and portfolio doing the asking, so an admin who still has full access sometimes sees an option the locked out owner doesn't.
- Live chat where it appears. It's the 1 route where a person reads your case in real time, and they can often tell you the internal reason category, which is frequently worth more than the reinstatement itself.
- Business verification in the Security Centre. If the trigger was identity or business legitimacy rather than creative, completing verification is the appeal in practice.
- The Page's own appeal surface, when the enforced object was a Page rather than an ad account.
What goes in the file, in order
A reviewer forms a view fast about whether this is a real business or another appeal from a shell. Order matters more than length. Put it in 1 PDF, in this sequence.
Don't open 6 tickets in parallel while you build it. Duplicates get merged, and the merge tends to keep the thinnest one. Pick the route that matches the object and send 1 file.
- Identifiers on the first line. Business portfolio ID, ad account ID with the act_ prefix, Page ID, enforcement date, and the notice text quoted exactly. Reviewers search by ID, so make the ID the first thing on the page.
- The policy you were flagged under, quoted verbatim from the notice, followed by 1 sentence naming the ad or asset you believe triggered it.
- The ads themselves. Ad ID, a screenshot of the creative, the primary text and the destination URL. Include the ones you suspect were the problem, not only the flattering ones.
- The landing page as it stood on the enforcement date and as it stands now, both dated. If review saw a page you've since changed, say so plainly and show both versions.
- The correction. What you changed, when you changed it, and the URL where the reviewer can confirm it without hunting for it. This is the line the first appeal never had, and it's the reason the second one gets read.
- Business legitimacy. Incorporation or registration document, the verified domain, a live storefront with working policy pages, a monitored support address, and evidence that orders actually ship.
- 1 sentence on the control you've put in place, written as an operational change rather than a promise. A named person checking every landing page against policy before launch beats a line about being more careful.
The things that get a second appeal closed unread
Before any of this you need a cause you can name. A file that guesses is worse than a file that waits a day, so if you can't point at the trigger, read back through the enforcement triggers and how each one reads in the notice until you can name one you're willing to defend in writing.
Then check the file against this list. These are the patterns that get a second submission closed without being read properly.
- Resending the first appeal's text. That language has already been scored and nothing about it has changed.
- Leading with spend. Spend history isn't a lever a reviewer can pull, and a large number reads as a reason to be careful rather than a reason to agree.
- Legal threats you have no intention of following through. That language routes the ticket away from the team that can actually reinstate you.
- A pile of loose attachments. Send 1 ordered PDF, named with your portfolio ID.
- Filing from a personal profile created 3 days ago, or from a portfolio with no history behind it.
- Emotion. Revenue, staff, rent. None of it is an input to the decision, and it occupies the space your evidence needed.
Day 1 to day 14: the appeal isn't the recovery plan
An appeal is a request with no published timeline attached. Meta doesn't commit to one, and the outcome isn't yours to control. Revenue is a separate problem and it started the moment the account went down, not the moment the appeal resolves. Run both in parallel or you'll spend 2 weeks at zero spend holding a well written PDF.
The line that matters here is between redundancy and circumvention, and it isn't a grey area. Infrastructure you provisioned before an enforcement event, with its own admin, its own payment instrument, its own domain and its own pixel, is ordinary operations. Rebuilding the same account under a new name after a ban, on the same card, the same device and the same domain, is circumvention. It gets caught, and it turns 1 account problem into a portfolio and profile problem. What to move first when an asset goes down mid campaign covers the sequencing.
If you've got nothing standing, the 2 honest options are a second portfolio built properly from scratch, which means weeks of warm up before it carries real budget, or infrastructure that's already aged and already trusted. How handover works and what you actually receive on day 1 is worth reading before you choose, because the whole difference between those 2 options is what exists on the first morning.
When to stop appealing
There's a point where appealing stops being work and becomes avoidance. 2 denials on the same object, the second one with a complete file behind it, is usually that point. A third submission containing nothing new is the same document a third time, and the week you spend waiting on it is another week with nothing running.
The honest trade off with rented infrastructure is that you don't own it. Nobody is selling immunity, because no account on the platform has it, and a rented account can be restricted like any other. What changes is what happens the day after. On a flat monthly rate, a replacement is part of what you already pay for rather than a fresh purchase, and the cost of an outage stops scaling with how much you were spending. A provider taking a percentage of your spend has the opposite incentive on the day you're down.
The questions worth asking before money moves, including what a replacement clause actually commits a provider to, are laid out with the answers that should end the conversation. Ask them of everyone, including whoever wrote the page you're reading.