The smallest data point in your dashboard is the loudest witness
When a fraud chargeback lands, most merchants reach for the obvious exhibits: the receipt, the order confirmation, maybe a screenshot of the product page. Those matter. But buried in the transaction details — in a row most people scroll past — sits a pair of two- or three-letter codes that quietly testify to something the receipt cannot: that whoever paid was holding the actual card and knew the actual billing address.
Those codes are the AVS response and the CVV result. They are the closest thing online commerce has to a fingerprint at the scene, and in a dispute framed as "I never made this purchase," they are often the single most persuasive thing you can put in front of the bank.
What AVS and CVV actually verify
AVS stands for Address Verification System. When a card is charged online, the numeric parts of the billing address the customer typed — the street number and the ZIP or postal code — are sent to the card issuer, which compares them against the address on file for that account. The bank sends back a code describing how well they matched.
CVV (also called CVC or CVV2) is the short security number printed on the card itself — three digits on the back of a Visa or Mastercard, four on the front of an American Express. It is deliberately never stored in the magnetic stripe or chip, and merchants are forbidden from storing it after a transaction. So a CVV match means someone, at the moment of payment, was physically looking at the card.
Neither code authorizes the payment on its own. A charge can go through with a mismatched address. But together they answer a question the rest of your evidence only circles: was the person who paid in possession of the genuine card and its private details?
Why these two signals carry weight in a dispute
Most fraud chargebacks fall under a reason that amounts to unauthorized transaction — the cardholder, or someone claiming to be them, says they didn't make the purchase. The issuing bank's analyst is not a detective with time to investigate. They are a person making a quick, defensible judgment from a thin file, looking for corroboration that points one clear direction.
This is where a small, hard data point outperforms a long explanation. Behavioral researchers studying how people weigh evidence describe a preference for diagnostic information — facts that distinguish between two competing stories rather than merely being consistent with one. Your order confirmation is consistent with both "the customer bought this" and "a fraudster bought this." An AVS full match plus a CVV match is diagnostic: it is hard to reconcile with a stranger who stole only a card number from a leaked database, because that stranger typically wouldn't have the billing ZIP and almost never has the printed security code.
That asymmetry is the whole argument. You are not proving the customer is lying. You are showing the bank that the transaction carries the markers of someone who had the real card in hand — markers a typical card-not-present fraudster lacks.
How to read the response codes
Stripe surfaces these results on every charge, and the values are worth knowing by sight rather than guessing at.
For AVS, the meaningful outcomes cluster into three groups. A full match means both the street number and the postal code matched the issuer's records — the strongest version. A partial match means one matched and the other didn't, often the ZIP matching while the street number didn't, which still carries real weight. A no match means neither lined up. There's also a frequent fourth case: unavailable, where the issuer simply didn't return a result, common with some international cards. Unavailable is not the same as a mismatch — it means the test couldn't run, and you should say so rather than let silence look like failure.
For CVV, the result is closer to binary: it either matched, didn't match, or wasn't checked. A clean match is the one you want to quote.
When you write your response, name the specific results plainly: "This transaction returned a full AVS match (street and ZIP) and a CVV match, indicating the payer possessed the physical card and knew the cardholder's billing address." You are translating a code into a sentence the analyst can lift straight into their decision.
Where the signal is strong, and where it isn't
Honesty about the limits of this evidence is what makes it credible. AVS and CVV are powerful against true fraud claims — the stranger-with-a-stolen-number story. They are far less useful against friendly fraud, where the real cardholder made the purchase, used their own address and card, and later disputes it anyway. In that case a full match doesn't refute their claim of "I didn't authorize this"; the genuine cardholder obviously passed their own address check.
For those disputes you need a different kind of proof — evidence of use, delivery, prior undisputed purchases, login records, IP and device history tying the buyer to the account. The verification codes become supporting texture rather than the headline.
The skill is matching the exhibit to the accusation. Read the reason code first. If the claim is that no legitimate cardholder was involved at all, lead with AVS and CVV. If the claim concedes the purchase but denies authorization or satisfaction, lead with the behavioral trail and let the match codes corroborate.
Build the habit before you need it
The practical failure isn't ignorance of these codes — it's that they're easy to forget under the deadline. By the time a dispute arrives, the original authorization is weeks old and the relevant details are scattered across the charge object, the customer record, and your own logs. Merchants who win consistently don't have better cards in hand; they have a habit of always checking what the verification results were and stating them, every time, in plain language.
A simple discipline helps: for any fraud-coded dispute, before you write a word, pull the AVS and CVV results and decide whether they're an asset or a non-factor. If they're an asset, they become your first paragraph. If they're not, you've saved yourself from leaning on a code that doesn't actually answer the question being asked — and you go find the evidence that does.
Letting the evidence assemble itself
This is exactly the kind of detail that gets lost in a busy week — not because it's hard, but because it lives in a field you have to remember to look at within a seven-day window that always seems to open on your worst day. Argeback reads the AVS and CVV results off each disputed charge for you, decides whether they strengthen the specific reason code you're facing, and writes them into an evidence-backed response it can file before the Stripe deadline — from your phone, while the transaction details are still at your fingertips instead of buried three screens deep.
If you'd rather never miss the quiet witness in your own data again, you can see how it works at https://argeback.lumenlabs.works.