Academy › Build › Payments

The reason the bank gives decides whether you can win

Knows how to read the reason for a chargeback within the first minute and decide whether there's anything worth fighting, before gathering any proof.

By Yodai · Published September 6, 2026 · Updated October 12, 2026

You get an alert from your payment processor: "Your $68 charge has been disputed and the funds have been pulled from your balance. You have 7 days to respond or you lose the dispute." Nobody called you. The customer didn't text you. The money's already gone from your account.

The normal reaction is to start gathering everything you have about that sale: the receipt, the confirmation email, the WhatsApp conversation where you set the time. That's the normal reaction, and it's the wrong one. Before you gather anything, there's just one thing to decide: whether it's worth fighting. That decision doesn't depend on how good your evidence is. It depends on the reason the dispute comes in with.

You don't write the reason: it comes from the cardholder and their bank

When someone disputes a charge with their bank, you don't get a letter explaining what happened. What the cardholder says, often by picking an option on their bank's form, turns into a reason from a fixed list. That reason is pretty much all you get as a merchant: you don't see the full complaint, you see a label.

Every card network has hundreds of reason codes, and your payment processor groups them into a handful of categories. To decide whether to fight back, it's enough to sort them into four families, and each one plays by different rules. The first is I don't recognize this charge: the cardholder says they never bought anything, that someone used their card without permission. The second is I never got it: they paid, and the order, appointment, or access never showed up. The third is it wasn't what I ordered: they got it, but say it doesn't match what you advertised. And the fourth is I canceled and you kept charging me: they canceled or asked to stop, and the charge is still there, or it comes in twice.

Think about what you sell: a haircut appointment, a candle order, access to an online course. The same reason, I never got it, means different things depending on the case. For an appointment, it means the customer says the service was never provided. For an order, it means the package never showed up. For access, it means they never got the link or password. The reason is the same, but what you need to show changes, and that deserves its own look.

The first reason is the scariest, and the one you can do least about. If a stranger really did pay with someone else's card, no receipt, photo, or conversation changes the outcome: the law protects the actual cardholder, not you. The card network isn't asking you to prove the service was done well, it's asking you to prove that it was that person, with their consent, who paid. That's almost never something you can prove on your own. What does change things is how the payment was made: according to Stripe's documentation, if the payment was confirmed through the customer's bank verification (3D Secure, a code, or an alert in their app), liability for a fraud dispute usually shifts to the card-issuing bank, and chances are that dispute never even reaches you. Check that detail in your payment processor before writing it off as lost.

There's one exception, and it's worth knowing because it changes the decision: if the customer is someone who's bought from you before, with the same phone number, the same delivery address, the same appointment on their calendar, then there is something to push back with. Signs that this wasn't random fraud, a repeat history, matching contact details, are what shift the odds, even though they never fully prove that this person was the one who paid. Outside of that exception, the best move with I don't recognize this charge is not to fight it: you almost always lose, and the time you'd spend doesn't change the outcome.

With it wasn't what I ordered, something different happens, and it's worth knowing before you decide if it's worth the effort. There's no single document here that settles the argument, the way a delivery notice does. It's the customer's opinion against the listing you published, and the person deciding is someone at the bank who has never seen your product and doesn't know your trade. If your listing was vague, with a generic photo and a one-line description, you've already lost before you start. If it had measurements, materials, photos of the actual product, and the return policy spelled out, you have something to show, even though the outcome doesn't depend on that alone.

The cancellation family is the most different of the four, because here you write the evidence before the sale, not after. If your policy states how many days in advance an appointment can be canceled without a penalty, and the customer agreed to it when booking, a charge that follows a late cancellation is defensible with that text. If that policy doesn't exist, or exists but you never showed it before charging, there's nothing to push back with, because the bank isn't going to invent a rule you never wrote down.

The deadline to respond is usually 7 to 21 days from when the notice arrives, depending on the card network, and your payment processor shows you the exact date. That clock doesn't wait for you to decide at your own pace. That's why order matters: first the reason, then, if it applies, the evidence.

Now you can do something you couldn't before: read the notice and know, within the same minute, whether you're going to build a response or let it go. What you gather to respond, and how that changes depending on whether you sold an appointment, an order, or access to something, is the next decision.