The notification arrives at 9:40 on a Tuesday night. A customer you remember, one who emailed twice and whom you answered warmly both times, has told their bank you stole from them. Your thumb is already moving. Block the email. Block the card. Never again.
That reflex is human, and it's often right. It's also the point where a lot of merchants quietly make a second mistake on top of the first, because the person who filed the dispute is not always the person you think they are.
A block list is a verdict, so treat it like one
When you add someone to a block list, you're deciding something about them: that they are a bad actor, that they will stay one, and that the cost of ever serving them again exceeds the revenue. Those are three separate claims, and a chargeback only gives you evidence for the first one on a good day.
Most disputes don't come from villains. They come from a handful of very different situations that look identical in your dashboard:
- Real fraud. A thief used a stolen card. The person whose name is on it never bought from you and is also a victim.
- Confusion. The charge appeared under a descriptor they didn't recognize, or a renewal surprised them.
- Frustration. They tried to reach you, felt ignored, and found the bank's button faster than your support page.
- Opportunism. They got the thing, then claimed they didn't. This is what's usually called friendly fraud, and it's the only category where a permanent ban is an easy call.
Blocking all four the same way means banning a fraud victim's identity, a confused loyal customer, and a thief with one blunt tool.
Why the card and the email are not the same thing
Stripe's Radar lets you maintain lists of values to block, such as card fingerprints, email addresses, IP addresses, and others, and you can write rules that reference them. What matters is which identifier you put on the list, because each one says something different.
A card fingerprint identifies a specific physical or virtual card. If that card was used fraudulently, blocking it is sensible: the genuine owner has probably already had it cancelled and replaced, so you lose almost nothing.
An email address identifies a person, or at least a login. In a stolen-card case, the email is often the thief's throwaway, which is fine to block. In a confusion case, the email belongs to a real customer who may be one clear explanation away from being your best advocate.
An IP address identifies a network, often shared. Block an office, a university, or a mobile carrier's gateway and you may cut off dozens of unrelated people. I'd treat IP blocks as the last resort and the shortest-lived.
So the first useful question isn't "should I block them?" It's "which thing am I actually certain about?"
The sorting question: what does your own record say?
Before you touch any list, open the customer's history. You're looking for the story the data tells, not the one your anger tells.
Some patterns lean toward a permanent block:
- Several accounts or cards from the same device or address, each with a dispute behind it.
- A pattern of receiving the goods or using the service, then disputing, then showing up again under a new email.
- A dispute that directly contradicts a long, friendly email thread in which they confirmed receipt.
Others lean the other way:
- A customer with a long clean purchase history and one dispute, especially one near an unexpected renewal.
- A first-time buyer who never contacted you at all. They may have gone straight to the bank without ever realizing there was a conversation to have.
- Anything that looks like true card theft, where the account holder may have no idea you exist.
The question underneath is whether this is a behavior or an incident. Behaviors repeat. Incidents usually don't. Your block list should be for behaviors.
Blocking does nothing for the dispute itself
This is the part people miss. Adding someone to a block list changes who can pay you in the future. It does not touch the dispute already open. That case still has a deadline, still needs your evidence, and is still decided by the issuing bank's review of what you submit.
Which leads to a trap: spending your Tuesday-night energy on the ban and then letting the response slide. A block feels like action, and so it can pass for a response when it isn't one. Respond first. Decide about the customer after the case is filed, ideally after you know how it's resolved.
There's a practical argument for waiting too. If the dispute is a misunderstanding and you win, you may want that customer back. If you lose and the cause turns out to be fraud, the card-level block is worth adding then. Either way, nothing is lost by sleeping on it, since the dispute can't stop a well-informed decision made a day or two later.
A middle path: friction instead of a wall
Binary thinking, banned or welcome, leaves out the most useful option. For a customer who disputed once under ambiguous circumstances, you can add friction rather than a wall:
- Require a manual review before their next order ships.
- Ask for 3D Secure on their next payment, so that if they dispute again, liability is harder for them to push onto you.
- Send a plain, human email explaining what the charge was for and how to reach you directly next time.
None of that punishes them. It simply makes the next dispute less likely or less successful, and it keeps the door open for the version of this story where they were just confused.
Your next moves
Do these in order, the same day the dispute lands:
- Respond to the dispute first. Put the evidence together and file it before you do anything involving the customer's account. The deadline is the only thing that can't be recovered.
- Pull the customer's full history. Count prior orders, support messages, refunds, and any other accounts sharing their card, device, or address. Write one sentence: "This looks like a behavior" or "This looks like an incident."
- Block the narrowest identifier you're sure of. If the evidence points to stolen-card fraud, add the card fingerprint. Leave email and IP alone unless you can see a pattern across several disputes.
- Set a review date. Put a note on the customer record to revisit the decision after the dispute closes. Most of your doubt will have resolved by then.
- Write down your own rule. Something like "Two disputes from the same person, or one dispute plus a contradicting email thread, equals a permanent block." A written rule stops tomorrow's mood from running your policy.
What a decent system gives you
The reason the angry ban is so tempting is that dispute handling usually feels like a scramble, and a block is the one thing that feels finished. Argeback is built for the part that matters more: it pulls your disputes in, drafts an evidence-backed response, and files it before the 7-day deadline, from your phone, so the Tuesday-night anger has somewhere better to go. With the case handled, the question of who stays on your customer list becomes something you can decide clearly rather than something you decide to feel better.
If you'd like to see how that works, you can look at Argeback. Either way, I'd wait a day before banning anyone.