You Won the Chargeback. Then the Pre-Arbitration Hit.

You submitted evidence on time. The issuer reviewed it, reversed the chargeback, and the money landed back in your settlement account. Two weeks later a new entry shows up in your disputes queue labeled pre-arbitration. Same cardholder, same transaction, round two. Most merchants do not know this stage exists until it costs them money. The mechanics are simple, the fees are not, and the decision deserves actual math.

How a dispute actually ends (or does not)

Most merchants treat a won chargeback as a closed file. The lifecycle has one more exit. A cardholder files. You respond with representment. The issuer either sustains the chargeback or reverses it. When they reverse it, they still hold a final move: pre-arbitration. Filing it means the issuer reviewed your evidence and still believes their cardholder. The dispute lands back with your acquirer and you get a fixed window to either accept the loss or push the case to formal arbitration, where the network itself rules and the fees get serious.

The fee stack nobody quotes upfront

The network schedules are public, but acquirers rarely surface them until the invoice arrives. Current published amounts:

Run a $250 dispute through that stack. Accept the pre-arb or lose arbitration and you are out the $250 in reversed revenue, plus whatever handling your processor charges, plus the chargeback fee you already ate. Escalate to arbitration and lose, and the $400 to $500 filing fee lands on top. Defending a $250 sale can quietly cost four times the sale, with the goods already gone.

Price the bet before you take it

Arbitration is a wager with your own money. Expected value equals win probability times recoverable amount, minus loss probability times total fees at stake. Say it out loud with real numbers before you click escalate.

A $400 dispute where you honestly rate your odds at 35 percent: 0.35 x 400 minus 0.65 x 700 in filing and handling gives you a negative $315 expectation. Fold. The same dispute at 60 percent odds still lands around negative $40 once handling is counted, before you value the two hours your ops person spends on it. Below roughly $500 in recoverable revenue, arbitration almost never pays unless you are making a point to an issuer about a repeat attacker. Above $2,000, merchants with real documentation win often enough that the math flips.

Your first submission is your whole case

Here is the part that stings. At arbitration most network flows restrict new evidence. The arbitrator reviews the record that already exists, which is your original representment package. The courier proof you dug up after the deadline does not come in. Merchants who submit a thin first response because they assume they can strengthen the file later are playing a hand they cannot improve.

What holds up in review: signature capture that matches the cardholder billing address, device fingerprints and IP logs tied to the account, timestamped terms of service acceptance, the refund policy as displayed at checkout, subscription usage logs dated after the billing date, and any email where the cardholder acknowledges the order. What does not: unlabeled screenshots, order confirmations with no delivery linkage, and paragraphs explaining that the customer is obviously lying. Evidence beats argument at every stage of this process.

Which disputes come back for round two

Escalation is not random. Pre-arb concentrates on reason codes where the issuer keeps facing an insistent cardholder: Mastercard 4837 and 4853, Visa 10.4, and the Visa 13.x consumer dispute family. Digital goods, subscription boxes, electronics, and ticket resale see the heaviest second rounds because proof of delivery is weaker and the products are easy to resell. Travel disputes escalate less because the cardholder usually consumed a documented service.

Practical move: log every pre-arb in one sheet for 90 days, with issuer, reason code, amount, and outcome. Most merchants discover that two or three issuers drive the bulk of their escalations, and behind them the same handful of cardholders. That sheet is your first-party fraud file, and it shifts your strategy from case-by-case arguments to pattern-based pushes, which issuers respect far more.

When to fold, when to fight

Fold immediately on merchant error: duplicate billing, wrong amounts, orders you cannot prove were delivered. Escalating a case you lost on the merits burns several hundred dollars to avoid admitting a $150 mistake.

Fight when you hold signature delivery to the billing address, when a repeat attacker shows up in your pattern file, or when subscription usage logs prove the cardholder kept using the service after the billing date. That last one is close to unbeatable evidence.

Everything else, run the expected value math, cap the hours you will spend, and default to folding when the number is close. Prevention compounds month over month. A grudge match with an issuer does not.

Stop round two from existing

The cheapest arbitration is the one that never opens, and the decisions that protect you get made weeks before the pre-arb lands. Two levers matter. Alert networks like Verifi RDR and Ethoca resolve qualifying disputes before they become chargebacks at all; we broke down the cost math in an earlier post. Visa Compelling Evidence 3.0 lets you return qualifying first-party fraud disputes using prior undisputed order data, which kills many would-be escalations at the first stage instead of the third. Neither lever helps once the pre-arb exists, which is the entire point.

Treat pre-arbitration as a pricing question, not a moral one. The issuer is not calling you a liar; they are spending their cardholder's credibility against your documentation. Price the bet honestly, check whether your first submission was actually your best one, and either fight it clean or fold it fast.