SituationsA big spender charges back

On three creators it looks like bad luck. On one roster it is a pattern.

A spender who has been good for two months disputes everything at once. If you look at each creator separately you see three unrelated refunds, none big enough to chase. Looked at across the roster it is one person, one afternoon, and $2,340.

The expensive part is rarely the money itself. It is that commission on it was computed days earlier, and if the period closes before the reversal is noticed, you have paid a team a share of revenue that left the account.

Book a demoPayroll and reversals

Reversed in one afternoon
$2,340

Across three creators on the same roster, by the same fan, on the 27th.

Commission already computed on it
$163.80

Owed to three different chatters, on money that is no longer in the account.

Days before the period locked
6

Which is the entire difference between an adjustment and a clawback conversation.

The same fan, across four accounts.

Each of these creators is managed by a different chatter, and on any single creator’s page this is a modest refund on a quiet Tuesday. The roster view is the only place it is one event.

Creator Spend this period Reversed on the 27th Commission computed Status
Luna Vex $1,240 −$1,240 −$86.80 Caught before the lock
Nova Reign $680 −$680 −$47.60 Caught before the lock
Aria Blake $420 −$420 −$29.40 Caught before the lock
Mia Storm $310 Still active, and now flagged
The fan, across the roster $2,650 −$2,340 −$163.80 One person, one afternoon, three chatters affected

The fourth row is the reason this view is worth having. Mia Storm’s account was never touched, and the chatter working it had no way of knowing that the fan they are mid-conversation with reversed everything on three other accounts this afternoon. Whether you keep selling to that fan is your call; making it without the information is not a call, it is a coin toss.

If it surfaces after payout

You are chasing money you already paid out.

The period closed with the sale in it. Commission went out on the 1st. The reversal appears in the following month’s figures as an unexplained dip, and by the time anybody traces it back, three chatters have been paid a share of revenue that left the account a week before their check cleared.

Your options from there are all bad. Absorb it, and the same thing happens next month at a slightly larger number. Deduct it from the next check, and you are taking money back from people who did nothing wrong, on a line they will not recognise, in a conversation that starts with them being told their pay is short.

If it surfaces before the lock

It is an adjustment, and it is itemised.

The reversal lands against the period it belongs to, the commission comes off before the pay run computes, and each affected chatter sees a line on their own statement saying what was reversed and when. Nobody is told after the fact that their pay was wrong.

The same event, six days earlier in the month, is administrative rather than political. That is the whole argument for detecting reversals against the transaction feed continuously instead of reconciling at month end.

How a reversal is noticed at all.

A chargeback is not announced. It is an absence: money that was in one record and is not in the other. Finding it means holding both records and comparing them, which is the only reason a gap can be spotted before the month ends.

The transaction feed says the sale happened
It arrived on the 19th, landed inside a shift, and had commission computed against it that night. Nothing about the feed changes when the money is later taken back — the original event is still there.
feed
The Statements page says the money is not there
The page the platform actually pays from, reconciled daily rather than at month end. On the 27th the two records stop agreeing by exactly $2,340.
statement
The difference is the reversal
Attributed to the period the original sale belongs to rather than the month it was noticed in, tied to the fan rather than to the creator, and surfaced against every creator that fan has touched.
reconciled

This is the same two-ledger comparison that produces the month-end reconciliation, run continuously instead of once. A reversal is simply the first place the two ledgers disagree, and the reason it is worth watching daily is that the value of finding it falls off a cliff the moment a period locks.

The repeat, and the judgement call

What the roster view can tell you, and the line where it stops and you start.

This fan disputed a $290 purchase three months ago, on a creator who is no longer on the roster. On a per-creator view that history left with her. Held against the fan, it is the second event, and it changes the question from “was this bad luck” to “do we want this customer”.

What the record will not do is answer that for you. It will not block anybody, it will not refuse a sale, and it does not score a fan as risky — a score would be a guess wearing a number, and two disputes is not a pattern you should let software act on. What it does is put both events on the same page, with dates and amounts, at the moment a chatter is about to send to that person.

What reversal detection does not do.

The limits worth knowing
  • It does not prevent chargebacks. The dispute happens between a fan and their bank, and nothing in an analytics tool sits in that conversation.
  • It cannot see a reversal the platform has not processed yet. Detection is as fast as the daily reconciliation, which means same-day rather than instant, and a dispute raised on the last day of a period may still land after the lock.
And two boundaries we will not cross
  • It does not score or rank fans by risk. Counting two disputes is a fact; turning that into a probability is an invention, and it would be acted on as if it were evidence.
  • It never moves money. The adjustment is computed and itemised; taking it off a payment is something you do on your own rail, for the reasons set out on the disclosure page.

Catch it on the 27th, not the 3rd.

A 20 minute walkthrough, then 14 days on your real roster. Bring last month’s refunds and see how many of them were the same person.

Book a demoAll eight situations

FanOptik

Analytics and chatting for OnlyFans agencies and independent creators. Measured across the whole roster, with every figure graded by where it came from.

© 2026 FanOptik. All rights reserved. · Built for agencies that would rather measure the roster than estimate it.