The reader you forgot you were writing for
Most merchants picture a chargeback as a contest of evidence. Gather enough screenshots, the thinking goes, and the truth will speak for itself. So they upload everything — the order record, the login timestamps, the terms of service, the email thread, the IP logs — and wait for the strength of the pile to carry the day.
Then they lose, and it makes no sense. The evidence was right there.
The missing piece isn't evidence. It's the person on the other end. Your rebuttal is read by a human reviewer working through a queue of disputes, each one a stranger's mess to untangle, each one allotted a thin slice of attention. They did not witness your transaction. They have no loyalty to your business. They have a stack of cases and a clock. Whether you win often comes down to a quieter question than "who is right": can this reviewer find your case inside the time they're willing to spend on it?
That is what a chargeback rebuttal letter is for. Not to prove you're honest — to make your honesty findable.
What a tired brain does with a wall of evidence
Cognitive load is a real and well-studied constraint, not a figure of speech. Working memory — the mental scratchpad where we hold and compare pieces of information — is sharply limited. When the material in front of someone exceeds what that scratchpad can hold, comprehension doesn't degrade gracefully. It collapses. The reader stops integrating and starts skimming, and skimming is where nuance dies.
A reviewer facing twelve unlabeled attachments and no explanation is carrying a heavy extraneous load — effort spent figuring out what they're looking at rather than what it means. Every PDF they have to open, orient, and interpret on their own is a tax on the same limited attention you need them to spend on your argument. Pile on enough, and you haven't strengthened your case. You've buried it.
There's a companion idea from decision research: satisficing, a term the economist Herbert Simon coined for how people actually choose under pressure. We rarely optimize. We scan for the first option that clears the bar of "good enough" and we stop. A queue-bound reviewer satisfices constantly. If your letter hands them a clean, sufficient reason to rule in your favor in the first few lines, they can take it and move on. If it doesn't, they satisfice toward the path of least resistance — and the path of least resistance is finding for the cardholder and closing the case.
You are not writing to overwhelm. You are writing to lower the cost of agreeing with you.
Lead with the conclusion, then earn it
Journalists call it the inverted pyramid: most important thing first, supporting detail after, background last. It exists because readers leave early, and the writer who front-loads the point survives the reader who quits halfway.
A chargeback rebuttal should open the same way. Before any narrative, state plainly what the dispute is and why the evidence resolves it. Something a reviewer can absorb in one breath:
This is a 'product not received' dispute. The customer downloaded the file twice from their account — on March 3 and again on March 9, six days after the charge they're now disputing. Access logs and login records are attached.
That's the whole case in three sentences. Everything after it is proof, arranged for someone who has already grasped the shape of what they're proving. A reviewer who reads only that opening can still rule correctly. That is the bar. Write so that the person who quits after sentence two still lands on your side.
Speak the reason code's language
Every chargeback arrives with a reason code — the bank's classification of why the cardholder disputed. "Product not received" and "fraudulent transaction" are not the same argument, and evidence that wins one can be irrelevant to the other. Proof that the customer logged in and used the product is decisive against a fraud claim; it does almost nothing against "the item arrived damaged."
The reviewer is, in effect, looking for an answer to one specific question the reason code defines. Your job is to answer that question, in the order they'll ask it, using the words the dispute itself uses. When your narrative mirrors the structure of the claim it rebuts, the reviewer doesn't have to translate — and translation, again, is load. Match the frame and you've done their hardest work for them.
Make every claim point to a document
A narrative without exhibits is an assertion. Exhibits without a narrative are a puzzle. The letter is what fuses them.
The reliable pattern is to make a factual claim and immediately name the evidence that backs it, in the same line: The billing and shipping addresses matched (see AVS result, Exhibit B). The customer's IP at purchase resolves to the same city as prior undisputed orders (Exhibit C). Now each attachment has a job and a label. The reviewer never opens a file wondering what they're meant to see in it, because you already told them. You've converted a pile into a guided tour.
This is also where restraint earns its keep. Three exhibits that each prove a necessary point beat ten that blur together. Extra documents don't read as thoroughness; they read as noise, and noise is exactly the extraneous load you're trying to spare. If a screenshot doesn't advance the specific argument the reason code demands, it's working against you by being there.
Plain, calm, and short
The instinct under threat is to argue — to convey, in tone, how wrong the customer is. Resist it. The reviewer has no stake in your indignation, and emotion costs them processing effort to wade through. Neutral, concrete language reads faster and lands as more credible precisely because it isn't trying to persuade by force.
Write the way you'd explain the situation to a reasonable stranger who is on nobody's side: dates, facts, what happened, what the attached record shows. Short sentences. One idea per paragraph. No jargon you can replace with a plain word. The goal isn't eloquence. It's a document someone can move through at speed without snagging — because the reviewer who finishes your letter is the only one who can be persuaded by it.
Why this is so hard to do on deadline
Everything above assumes you have the clarity to write it. In practice, a chargeback notice lands mid-week, the response window in Stripe is short — often around a week — and the moment competes with everything else a business owner is carrying. So the letter gets written at 11 p.m. in a hurry, or not at all, and the dispute is lost by default. Not because the merchant was wrong, but because the work of translating evidence into a reviewer-shaped argument never got done.
That translation is the entire job. It's also the part that doesn't scale by effort alone — it scales by knowing what a reviewer needs and giving them exactly that, on time.
That's the gap Argeback is built to close. It ingests your Stripe disputes, drafts an evidence-backed rebuttal organized around the reason code and the reader who has to act on it, and files it before the deadline closes — all from your phone, in the minutes you actually have. You stay the one who knows your customer; it handles turning that knowledge into a letter a busy reviewer can say yes to without working for it. If chargebacks keep slipping away while you mean to get to them, that's worth fixing: argeback.lumenlabs.works.