The decision nobody teaches you
Most advice about chargebacks assumes you've already decided to fight. But the first real decision a merchant faces, the moment a dispute lands, is whether to fight at all. Contesting a chargeback costs time and attention, and not every dispute is winnable. Spend your effort indiscriminately and you'll burn hours on hopeless cases while letting winnable ones slip past the deadline. The skill isn't just fighting well — it's knowing which battles are worth fighting.
This is a genuine decision with real trade-offs, and it deserves a framework rather than a gut reaction. Here's how to think it through, dispute by dispute.
First, a thing that isn't a real option
Before the framework, clear up a common confusion: once a chargeback has been filed, refunding the customer is usually not a clean exit. A refund and a chargeback move through separate systems and don't cancel each other, so refunding on top of an open dispute can mean paying twice — the refund plus the chargeback if it resolves against you — and you still eat the dispute fee. So "just refund them to make it go away" doesn't work after the fact. The time to prevent a chargeback with a refund is before it's filed, when a customer is merely unhappy. After it's filed, your real choices are to contest it or to accept the loss. Refunding into an open dispute is the worst of both.
With that settled, the question becomes: contest, or accept?
The first axis: is the case winnable?
The strongest input is the reason code and the evidence you hold against it. Some disputes are genuinely winnable, and some genuinely aren't, and the difference is usually obvious once you look:
Lean toward fighting when the claim is contradicted by records you actually have. A "product not received" dispute where your carrier confirms delivery to the billing address. A "subscription canceled" dispute where your logs show the customer using the service after the date they claim they quit. A "fraudulent" dispute that's really friendly fraud — the genuine account holder, with a history of prior undisputed orders from the same device and address. In these, the facts are on your side and merely need to be presented. These are the cases worth your time.
Lean toward accepting when the customer's claim is true or you can't rebut it. Tracking that shows the package never arrived. A duplicate charge that really was a double-bill. A refund you genuinely owed and didn't process. Fighting a true claim isn't shrewd; it's a waste of effort that you'll lose anyway, and it can sour a customer relationship that might otherwise have been salvageable. Accepting a loss you deserve to take is a legitimate, often correct, decision.
The honest gut check: if I were the neutral reviewer reading both sides, which way would I rule? If you can't articulate why the evidence favors you, it probably doesn't.
There's a middle band worth naming, too: cases that are plausibly winnable but where your evidence is thin — a delivery that probably happened but where the tracking is vague, a subscription dispute where your usage logs are incomplete. These are judgment calls. A useful tie-breaker is to ask what the rest of the file looks like once it's assembled: a thin piece of evidence inside a clean, coherent timeline reads very differently from the same evidence presented alone. Sometimes a marginal case becomes worth fighting simply because the surrounding story is strong enough to carry it.
The second axis: is it worth the effort?
Winnability isn't the only variable. Even a winnable case has a cost to fight, and you should weigh the recovery against that cost. A clearly winnable $400 dispute is an easy yes. A marginal $15 dispute that would take an hour to assemble evidence for may not be worth the hour, even if you'd probably win — your time is the scarce resource.
But be careful with this reasoning, because it has a trap. The cost of fighting drops dramatically when you have a system, and the cost of not fighting includes more than the lost dollars. Every lost dispute counts against your account's dispute ratio with the card networks, and ratios that climb too high pull you into network monitoring programs that carry fines and, eventually, threaten your ability to process cards at all. So a small dispute you'd shrug off in isolation may be worth contesting if your ratio is creeping upward. The dollar amount is the visible cost; the ratio is the invisible one.
The third axis: what is the dispute telling you?
There's a longer-term consideration the per-dispute math misses. Before deciding fight-or-accept on a single case, notice whether it's part of a pattern. Three "unrecognized" disputes in a month aren't three decisions — they're one signal that your billing descriptor doesn't match your brand on customer statements. A run of "subscription canceled" disputes points at a cancellation flow that's too hard to find. In those cases the highest-value move isn't fighting harder; it's fixing the upstream cause so the disputes stop being generated. You still contest the ones already on the table, but the real win is preventing the next ten.
Putting it together
A workable order of operations: First, can I rebut this specific claim with evidence I actually have? If no, accept the loss and, if appropriate, make the customer whole outside the dispute. If yes, second: does the recovery — counting both the dollars and the dispute-ratio cost — justify the effort? With a system that makes contesting cheap, the answer is usually yes for any winnable case. Third, regardless of how I handle this one, what is it telling me to fix upstream?
Notice that the framework collapses toward "fight the winnable ones" as the cost of fighting falls. The merchant for whom contesting a dispute means a 90-minute midnight scramble rationally lets small ones go. The merchant for whom contesting takes a few taps fights nearly everything winnable, because the effort no longer outweighs even modest recoveries — and their dispute ratio stays healthy as a result. The decision framework is real, but a lot of it exists only because fighting is currently expensive. Make it cheap and most of the agonizing disappears.
Where Argeback fits
That's the shift Argeback is designed to create. By tailoring the evidence questions to each reason code, it makes the winnability of a case obvious — you can see quickly whether the facts support you. By drafting the narrative, packaging the evidence, and filing through Stripe in a single tap, it drops the cost of contesting from a 90-minute scramble to a few minutes, which tilts the whole framework toward fighting the winnable cases you'd otherwise have let go. And its honest dashboard, with wins and losses by reason code, surfaces the upstream patterns worth fixing. The issuing bank still decides each outcome. Argeback just makes the decision to fight an easy one to act on.