When to let an account die

In four years of selling infrastructure to people who run fleets, nobody has asked me whether to appeal. They ask how. How to appeal without linking anything, how to word it, which address to send it from.

The whether is the part that decides the outcome, and it gets skipped every time, because by the time anyone is thinking about it the answer has already been chosen by adrenaline.

I run mobile proxy lines on carrier SIMs and a rack of Android phones rented out as cloud phones. I sell the layer underneath other people’s accounts, so I see the aftermath from below, on platforms I never log into. What reaches me is mostly the fortnight afterwards. The ban itself is old news by the time anyone mentions it.

Filing is a linking event

An appeal feels free. You submit a form, somebody says yes or no, and a no leaves you where you already were.

That is not the shape of it.

Every route back into a locked account is an identity check. A code to a number. A mail to an address. A document. A card. A call from a handset. A support conversation typed on a machine that is also running three other accounts in other windows.

Whatever you supply gets attached to that account permanently, and by extension to everything else already sharing it. The platform does not forget an identifier because the appeal failed.

So filing is a deliberate decision to connect identifiers to one another, taken at the single moment when one of them is under active review. That is a strange thing to do by reflex, and almost everybody does it by reflex.

I have covered the ledger of a banned account separately, including why a recovery attempt costs some fraction of a second account. There is a separate guide on running a recovery cleanly as well. This piece is the decision that sits in front of both of them.

Condition one: something inside that cannot be rebuilt

The account is packaging. The question is whether the contents exist anywhere else.

A balance. A payout in flight. An audience that lives on that platform and nowhere else because nobody exported anything. The only copy of a conversation with a buyer who was about to commit.

Those are reasons. Months of work invested belongs somewhere else entirely. Effort already spent goes on having been spent whatever you decide next, and it argues for nothing.

Even a successful appeal fails to return the largest thing you lost. Age is a record and the appeal does not rewind it. What comes back is the contents, attached to an account that now carries a formal review and a note against it.

Condition two: a specific reason you can actually answer

A stated reason with a date, an action, or a named policy is a question. Questions can be answered, with an invoice, a business registration, or a plain account of what happened that matches what the logs already show.

“Your account was found in violation of our policies” is not a question. There is nothing in that sentence to respond to. Filing against it produces a document that is not an argument, only a request to be examined more closely.

If you cannot state, in one sentence, what you are arguing against, you do not have an appeal. You have an application for scrutiny.

Condition three: a route back that asks for nothing shared

This is the condition people skip, and it is the one that determines whether the whole thing stays contained.

If the flow wants a code sent to the number that already belongs to this identity, and a mail to the address that already belongs to it, that is the account proving it is itself with its own things. Nothing new gets welded on.

If it wants a document, a card, a live call, or a chat session from a device, then answering costs an identifier that exists elsewhere in the operation. The card is the ugly one, because you cannot rotate it in an afternoon the way you can swap an exit IP.

All three conditions have to hold. One failing is enough to walk.

The cases that need no weighing at all

An unexplained lock on a young account. Nothing filed, nothing asked, no writing in to find out. It is the cheapest bad outcome available and you take it.

A template refusal on the first attempt. The account history did not clear a threshold, and the history is not the part you are allowed to edit.

And an account under about six months old, in most operations, is not holding anything you could not rebuild in a fortnight. The appeal risks more than the loss it is trying to reverse. People fight hardest for exactly these, because the effort is recent and still stings, which is exactly why that person should not be making the call.

Write the rule while nothing is on fire

The analysis is the easy half. The hard half is that the decision gets made about twenty minutes after the lock, every time, because that is when you find out.

At twenty minutes you are angry, the account was fine yesterday, work is sitting inside it, and there is a button on screen that takes forty seconds and feels like doing something. Nobody in that state has ever decided to walk away. I have not seen it, and I have not managed it myself.

So the decision has to already exist in writing before you get into the state.

One field, alongside the number, the mailbox, the line it sits behind, and what the account is for. Fight or drop, plus the reason, so that the version of you standing there at midnight cannot talk his way around it.

The rule has no more insight than you do. Its only advantage is having been written by somebody who had not just lost anything.

Letting it die is a short list of actions

Walking away is not the same as doing nothing, and the difference is where the second round of damage usually happens.

An abandoned account is still an open one. It holds a number as a live recovery field, names a mailbox as its way back in, and the platform keeps all of it on file. From their side nothing ended. You stopped logging in.

Stop paying for what sits under it first. A dedicated line runs about ten dollars a month for the SIM before anything else, and I have watched people carry one into June under an account they gave up on in January.

Then take the number and the mailbox out of circulation rather than moving them onto something new. The recycling problem has its own guide and I am not rewriting it here. The compressed version is that the identifiers a dead account leaves behind are the most tempting things in your inventory and the most expensive things to reuse.

Then mark it dead in the record, with the date and whatever you know about why. Six months later, provisioning something and hunting for a spare number, that note is the only thing between you and rebuilding the link the account died escaping.

What you do not do is strip it for parts the same afternoon.

The second appeal is where it gets expensive

Here is the position I get argued with about. The second appeal is almost never the right call, and it is the one everybody files.

A first appeal, assuming you had grounds, runs on the clean route. The account’s own number, the account’s own mailbox, the form built for the purpose.

The second one cannot. That route is spent. So it goes somewhere else: a different address in case the first is being ignored, a phone number found on a help page and dialled from whichever handset is in the room, a message to the platform’s social account from a profile that is obviously yours, an email from the business domain because surely a real business gets read properly.

Every one of those is a fresh identifier, volunteered, onto a record that is already open.

That is the actual mechanism behind losing three accounts to one ban. Not the ban. Somebody going through his own drawer looking for something the platform has not seen yet, and handing it over one item at a time over nine days.

If the first attempt comes back as a template, there is no better second attempt to write. The input you are trying to change is the account’s history, and that part is closed.

Where this is wrong

Two places. The first is that I sell proxy lines, and my advice happens to be that you should abandon accounts more readily and keep the infrastructure under them clean. Weigh it knowing that.

The second is real. All of this assumes the account is replaceable. If it is a business’s only sales channel, or it holds four years of customer conversations, or it is the name over the door, then you fight, through the second appeal and the fifth, and you accept the linking as the price. That is the correct decision and I would make it too. What I would still do is fight from one identity, on purpose, with identifiers already written off. The mistake lies in fighting with whatever is nearest to hand.

For years I told people to appeal everything on the grounds that asking costs nothing. I was wrong for a specific reason: I was treating it as a customer service interaction rather than a moment of scrutiny. Asking is not free when the asking is done with your identity.

What changed my mind was the same sequence landing on three separate customers inside a year. Lock, appeal, second appeal from a personal address, then an unrelated account gone about a fortnight later. None of them connected the two events, and all three told me the later loss was coincidence. I cannot prove it was not, and I will not pretend otherwise. I have a pattern watched several times from the outside and no visibility into what the platform did with any of it. Anybody who tells you exactly which field triggered the second ban is guessing.

Guides and the stack we run and test are at Multi Account Ops.

Get new guides and videos first — join the Telegram channel.