Chargeback

A chargeback is a forced reversal of a card payment, initiated by the cardholder’s issuing bank rather than by the merchant. It exists to protect cardholders from unauthorized charges and undelivered goods, and it has become the mechanism through which a great deal of card fraud is discovered — usually weeks after it happened.

What the process looks like

The cardholder disputes a charge with their bank. The issuer assigns a reason code, reverses the funds provisionally, and notifies the acquirer, who notifies the merchant. The merchant can accept the loss or contest it — representment — by supplying evidence that the transaction was valid and the goods delivered. If the issuer rejects that evidence the case can escalate to pre-arbitration and then arbitration, where the network decides and the losing party pays the fees.

Two features shape merchant behavior more than the rules themselves.

The cost is not only the transaction. The merchant loses the payment, the goods, and a chargeback fee that applies regardless of outcome. Contesting costs staff time. For low-value transactions the rational choice is frequently to accept a loss the merchant would win, which is a fact attackers rely on.

Ratios matter more than amounts. Card networks operate monitoring programs that trigger when a merchant’s dispute rate crosses defined thresholds, bringing fees, remediation requirements and ultimately the risk of losing card acceptance. A merchant can be profitable and still be in serious difficulty on this measure, which is why chargeback ratio is managed as carefully as fraud loss.

Friendly fraud, and why the category is contested

Not every fraud-coded chargeback involves a criminal. First-party misuse — commonly called friendly fraud — is a genuine cardholder disputing a transaction they actually made: because they forgot it, because a family member used the card, because the return process looked harder than the dispute, or because they simply preferred not to pay.

The merchant experiences this identically to criminal fraud, and the reason code frequently says fraud in both cases. That makes chargeback data a poor measure of actual fraud, and it explains why a merchant’s fraud figures and an issuer’s can differ substantially while both are accurate.

It also puts a limit on what identity controls can achieve here. Verifying the cardholder does not prevent a verified cardholder from later disputing their own purchase.

Why this matters for identity verification

Chargebacks are how criminal card fraud usually surfaces, and the timing is the problem. A dispute arrives weeks or months after the transaction, so a merchant reading chargeback data is measuring decisions made long ago against attack methods that have since moved on. The feedback loop is real and far too slow to tune against.

That argues for pushing the decision earlier. Preventing a card-not-present fraud at authorization avoids the chargeback entirely, along with the goods, the fee and the ratio impact. Payment card capture with fraud signals examines the card at the point of entry rather than waiting for the dispute.

Where the exposure justifies it, verifying the person — not the card and not the billing data, both of which appear in every breach — changes what a dispute can claim. A transaction backed by a verified identity is materially harder to contest as unauthorized, which is a defense against both criminal fraud and first-party misuse. That is the argument for identity document verification on high-value or high-risk transactions, and why payment fraud controls should be judged on what never reaches a chargeback.

What chargebacks can’t do

They are not a fraud metric. First-party misuse and criminal fraud carry the same reason codes, so the number measures disputes rather than fraud.

They arrive far too late to tune controls. Weeks or months of lag means the data describes attack methods that have already changed.

Winning does not make a merchant whole. Fees apply regardless, staff time is spent, and the dispute still counts toward monitoring ratios in most schemes.

They do not cover every payment type. Bank transfers and instant payment rails have no equivalent mechanism, which is exactly why fraud migrated to them.

Frequently asked questions

What is the difference between a chargeback and a refund?

A refund is the merchant voluntarily returning money to the customer. A chargeback is the issuing bank forcibly reversing the payment following a cardholder dispute, and it carries a fee and a ratio impact for the merchant whether or not the dispute is ultimately won.

What is friendly fraud?

First-party misuse — a genuine cardholder disputing a transaction they actually made, whether because they forgot it, a family member used the card, or disputing looked easier than returning. Merchants experience it identically to criminal fraud and the reason codes often match, which makes chargeback data a poor measure of real fraud.

Can a merchant fight a chargeback?

Yes, through representment — supplying evidence that the transaction was valid and the goods were delivered. Cases can escalate to pre-arbitration and arbitration decided by the card network. For low-value transactions the cost of contesting frequently exceeds the amount, so winnable disputes are often conceded.

Why do chargeback ratios matter more than chargeback costs?

Because card networks run monitoring programs triggered by dispute rates rather than amounts. Crossing a threshold brings fees, remediation requirements and ultimately the risk of losing card acceptance, so a profitable merchant can still be in serious difficulty on this measure.

Related reading

• Card-not-present fraud — the category most fraud chargebacks originate from
• Credit card fraud — the two families of card fraud, and which one a chargeback reveals
• Address Verification Service — a control that runs at authorization, and what it does not prove
• First-party fraud — the category that looks identical and has a different cause