The Chargeback Defense Kit

The problem

A customer disputes an $80 order. The chargeback notification lands in your Stripe dashboard on a Tuesday. By the time you actually deal with it, you have lost the $80, paid the $20 processor fee, spent ninety minutes assembling tracking screenshots and a defensive paragraph, and watched the case close against you anyway. The disputed transaction line in your accountant’s report says $80. The true cost — card amount + processor fee + COGS on the lost inventory + staff time + the CAC you spent to acquire the customer in the first place — is closer to $218 per dispute. Across a year of forty disputes, that is roughly $8,700 of margin disappearing into a workflow nobody owns.

The industry average win rate is about 20%. Most sellers know that number and assume it is just the cost of doing business online. It is not. Sellers who run a structured evidence workflow — reason-code-specific evidence, six labeled exhibits, a one-page timeline, submission inside 48 hours — win 60%+ of the same disputes. The 40-point gap is not luck and not generous processors. It is evidence quality. And the reason the industry average stays at 20% is that most sellers either do not respond at all or respond with a screenshot of the order page and a sentence saying “the customer received the product.” Processors interpret silence as guilt, and they interpret weak prose as silence with a stamp on it. The seller who folds repeatedly teaches the processor (and the customer) that this seller will keep folding.

What most people get wrong

They submit generic evidence and ignore the reason code. A Visa 13.1 (Merchandise Not Received) wants tracking, signature, AVS match. A Visa 13.7 (Cancelled Recurring) wants a screenshot of the cancellation flow and proof the customer never used it. Submit the wrong evidence pack against the wrong code and you lose by default — not because the case was unwinnable, but because the arbitrator is reading a 13.1 evidence template against a 13.7 dispute and finding no overlap. The reason code is the lens. Every defense begins by looking it up. The book’s Chapter 2 decoder ring walks Visa 10.4/13.1/13.2/13.5/13.7, Mastercard 4837/4853/4863, Amex C04/C05/C32, and Discover DA/NA/RG/UA02 with the exact evidence each one requires.

They only fight the big ones. The intuition is rational: a $400 dispute is worth ninety minutes of work, a $40 dispute is not. The math is wrong. Processors flag merchants with sub-30% win rates for elevated risk monitoring, raise reserve requirements, and in extreme cases close the account. Stripe Radar’s risk model also incorporates your dispute-to-resolution ratio — a merchant who responds to every dispute and wins half signals competence to the network. A merchant who responds to two out of ten signals incompetence even if those two were both wins. The win rate matters as a metric, not just as recovered dollars. Fight all of them, with a template that makes the marginal cost of fighting a $40 dispute close to zero.

They forget the pre-fulfillment fraud screen. The cheapest dispute is the one you prevent. A package that never shipped because Stripe Radar flagged the order at checkout costs you a refund and nothing else. A package shipped to a freight-forwarder address with an AVS-N result and a velocity flag costs you the chargeback plus the inventory plus the fee plus the time. Most sellers spend 90% of their dispute attention on defense (after the chargeback arrives) and 10% on prevention (before the order ships). The ratio should be inverted. A single Stripe Radar custom rule — block when card country ≠ IP country ≠ billing country — typically blocks 0.5% of total transactions and eliminates 70%+ of fraud disputes. The book’s Chapter 8 walks the screening stack: Stripe Radar custom rules, Signifyd for Shopify, Sift and Kount for higher-volume merchants, the manual-review checklist for medium-risk orders, and the velocity thresholds that catch card-testing patterns.

A working approach

The defense logic in the book reduces to a four-step pattern. Every dispute, regardless of processor or amount, runs through the same loop:

Reason code -> Product type -> Evidence pack -> 48-hour submit

Step 1: Reason code lookup
  Visa 13.1, 13.2, 13.5, 13.7, 10.4
  Mastercard 4837, 4853, 4855, 4863
  Amex C04, C05, C08, C32
  Discover NA, RG, RM, UA02
  -> Maps to one of six underlying disputes

Step 2: Product type lens
  Physical / Digital / Subscription / Service / Hybrid
  -> Determines which evidence is even possible

Step 3: Six-exhibit evidence pack
  A. Order confirmation
  B. Customer acknowledgment (checkout checkbox)
  C. Proof of fulfillment (tracking / download log / signup record)
  D. Reason-code-specific artifact
  E. Communication thread (full export, not screenshots)
  F. Policy excerpts (version active at purchase)
  + Optional G. Fraud-screening output
  + Optional H. Carrier investigation outcome

Step 4: 48-hour workflow
  Hour 0-12: Triage (identify code, pull order, locate prior comms)
  Hour 12-24: Evidence collection (artifacts per product type)
  Hour 24-36: Assembly (cover summary + timeline + exhibits)
  Hour 36-48: Review and submit through processor portal

The reason-code decoder ring is the chapter most sellers wish they had read two years earlier. The six underlying disputes (fraud, non-receipt, quality, cancellation, credit not processed, unrecognized) cover roughly thirty active reason codes across the four networks. Once you can map any incoming code to one of the six, the evidence template selects itself. The bonus file reason-code-quickref.md is a one-page lookup that sits next to the response template.

Evidence by product type is where most rushed responses fall apart. Trying to submit a tracking number against a digital course dispute, or a download log against a physical-goods dispute, is the second most common loss pattern after ignoring the reason code. Physical goods need carrier tracking + signature + AVS match. Digital goods need access logs + IP/device fingerprint + pre-sale acknowledgment + course progress data. Subscriptions need the signup terms acknowledgment + renewal notifications + cancellation flow screenshots + usage data during the disputed period. Services need the signed agreement + delivery log + customer sign-off + the communication thread. The chapter is structured so you can pick the product type, see the required exhibit list, and assemble the pack without thinking about it.

Policy pages are the unglamorous part. Most sellers treat refund, shipping, and subscription cancellation pages as legal afterthoughts. They are dispute-prevention tools. A processor evaluating a Visa 13.7 (Cancelled Recurring) is checking whether your auto-renewal disclosure was visible at checkout, whether your cancellation policy was as easy as your signup (FTC click-to-cancel rule), and whether the customer received pre-renewal notification emails. If any one of those three is missing, the dispute is hard to win regardless of evidence. The book’s Chapter 6 covers the seven clauses processors specifically look for, the refund policy template, the shipping policy template, the subscription cancellation policy template, and the placement hierarchy (footer link → checkout link → checkout excerpt → checkout checkbox → checkbox with inline summary). The policy-pages-pack.md bonus file ships three full policy starter templates with the version-archive pattern that defends grandfathered disputes when you update policies later.

Fraud screening before fulfillment is the highest-ROI defensive activity in the entire workflow. Stripe Radar covers most merchants out of the box; the free tier scores each transaction 0–100 and lets you block above a threshold. Radar for Fraud Teams adds custom rules at $0.05 per screened transaction, which pays for itself within the first quarter for any merchant processing more than $100K/year through Stripe. For higher-volume merchants the chapter walks Signifyd (Shopify-focused, with chargeback guarantee on approved transactions), Sift (multi-channel and marketplaces), Kount (enterprise card-not-present), Riskified (cross-border and luxury), and the manual-review workflow for medium-risk orders that should not auto-approve or auto-block. The velocity-trap section in particular — same card + 3 orders in 24 hours, same email + 3 orders in 24 hours, same shipping address + 5 orders in 7 days — catches the card-testing pattern that precedes most large fraud orders.

The 48-hour workflow is the closing chapter and the operational spine of the entire book. Stripe gives you 7 days. PayPal gives you 10. Square gives you 5. The temptation is to fill the available time. The reality is that evidence quality degrades hard after the first 48 hours — memory fades, the email thread gets harder to assemble, the carrier ticket takes longer to come back, and the inevitable surprise (a vacation, a holiday, a sick day) pushes a day-6 response to day 8 and the case closes by default. The four 12-hour blocks (triage → evidence collection → assembly → review and submit) force the rigor that ad-hoc responses lack. Pair the workflow with the chargeback-evidence-template.md bonus (the fill-in-the-blank cover-summary + timeline + exhibit-checklist template) and a response that used to take ninety chaotic minutes on day 6 takes forty-five disciplined minutes across days 1–2.

Where this scales

The article above is the operational pattern. The book covers each layer with the templates, the platform-specific gotchas, and the case studies that show the difference between a 22% win rate and a 58% win rate in one quarter. Chapter 4 on physical proof-of-delivery is where most physical-goods sellers find the highest-leverage fix: USPS Signature Confirmation vs Adult Signature, the UPS My Choice override that silently negates your signature requirement (use Adult Signature Required, not standard Signature Required), the FedEx Direct Signature pattern, the pack-out photo procedure for orders above $300, and the carrier service ticket (USPS Missing Mail Search, UPS Lost Package Claim, FedEx Trace Request) that becomes third-party validation in your evidence pack.

Chapter 5 on digital access proof is where most course creators, software sellers, and download merchants find theirs. The five pillars (download log, IP/device fingerprint, license server pings, video/course progress, account activity log) compound into evidence packs that are nearly impossible to argue against — a customer who disputes a $497 course after watching 47 lessons across 21 days with three submitted assignments has very little defense. Chapter 7 on communication logs and timelines is the structural unlock: the same evidence, presented as a labeled timeline with one-line annotations rather than a stack of raw screenshots, moves win rate roughly 30 percentage points on average. Processors are time-constrained. The clearer the presentation, the better the outcome.

The bonus files make the system usable from day one. chargeback-evidence-template.md is the fill-in-the-blank response document — cover summary, event timeline table, required fields by product type, exhibit checklist, closing statement, pre-submission checklist. reason-code-quickref.md is the one-page lookup that maps every Visa/Mastercard/Amex/Discover code to its required evidence. policy-pages-pack.md is three full policy templates (refund, shipping, subscription cancellation) with the version-archive pattern. Three files. Cover the workflow end to end.

Included with the book

• Chargeback Evidence Response Template (markdown) — fill-in-the-blank cover summary, event timeline format, required fields by product type, exhibit checklist, closing statement, pre-submission checklist. Open a fresh copy per dispute, fill the brackets, attach the exhibits, submit.
• Reason Code Quick Reference (markdown) — one-page lookup mapping every Visa/Mastercard/Amex/Discover code to its required evidence artifacts. Sits next to the response template.
• Policy Pages Pack (markdown) — refund policy, shipping policy, and subscription cancellation policy starter templates with the version-archive pattern that defends grandfathered disputes after policy updates.

Get the full picture

Beyond the dispute portal, the next layer of defense is identity. Most chargeback prevention stops at “did this transaction look fraudulent at checkout?” The harder question — “is this person actually who they say they are, before the high-risk order ships?” — is where cryptographic identity verification becomes the prevention move that fraud-screening alone cannot make. trust.authority issues verifiable credentials that a buyer can present at checkout for high-ticket or high-risk orders, providing third-party-verified identity assertion that becomes evidence in any subsequent dispute. The integration sits upstream of the chargeback workflow, not downstream of it. For merchants who have squeezed everything they can out of post-fact defense, identity verification before fulfillment is the next compounding layer.

Readers of this also chose

Questions readers ask

Does this apply to Shopify, WooCommerce, BigCommerce, or just one platform?

All of them. The reason codes and evidence requirements come from the card networks (Visa, Mastercard, Amex, Discover) and the processors (Stripe, PayPal, Square, Shopify Payments, Adyen), not the storefront platform. Shopify merchants who use Shopify Payments are actually using Stripe under the hood — same dispute portal, same evidence format, same 7-day window. The workflow is portable across any platform that runs through one of the major processors.

What about PayPal disputes specifically? They feel different.

PayPal disputes run through the Resolution Center rather than the Stripe-style dashboard. The 10-day window is more forgiving than Stripe's 7 days. PayPal Seller Protection covers eligible transactions for unauthorized payment and item-not-received disputes if you meet their criteria (eligible items, ship to the address on the transaction, signature for orders above $750). The book's Chapter 9 covers the PayPal-specific submission steps. The evidence pack itself is the same six exhibits.

Is the 60% win rate realistic or aspirational?

Realistic for sellers who use the full system. The 60%+ figure is the documented average across sellers running a structured evidence workflow with reason-code-specific responses and 48-hour submission. Some categories run higher (physical goods with signature on delivery routinely hit 70–80%); some categories run lower (subscription disputes max out around 55–65% even with good evidence, because the cancellation question is genuinely contested). The 60% blended average is the realistic operational target. The 90%+ outliers are the merchants who also nail prevention (Chapter 6 policy pages and Chapter 8 fraud screening), which moves the question from "can I win this dispute" to "can I prevent this dispute from being filed in the first place."

Can a third-party service like Chargehound or Midigator do this for me?

Yes, and the book covers when to consider that move. Chargehound, Midigator, Justt, and similar services automate evidence assembly and submission — they integrate with Stripe, Shopify Payments, and other processors, pull the order data, generate the response, and submit. The economics work for merchants processing more than 50 disputes a month, where the per-dispute service fee (typically $20–$40) is less than the staff cost of an in-house workflow. Below that volume, the templates in this book + a disciplined 48-hour workflow win more disputes than the automation does, because human judgment on which evidence to feature outperforms generic AI-assembled responses for low-volume custom cases. The break-even is somewhere between $1M and $3M in annual revenue depending on dispute rate.

What if I need a refund?

Checkout runs on Lemon Squeezy. The standard refund window applies. You keep the PDF either way.

Does this cover EU disputes (SCA, PSD2, Klarna)?

Partly. The reason codes for Visa and Mastercard are the same in the EU. The procedural layer differs: PSD2 Strong Customer Authentication (3D Secure 2 challenge) shifts fraud liability to the issuer when invoked, which makes 10.4 / 4837 / UA02 fraud disputes much harder for the cardholder to win. The book references SCA but does not deep-dive EU-specific procedure. Klarna, Afterpay, and other BNPL providers run separate dispute processes outside the card networks and are out of scope here. EU merchants reading this should adapt the workflow to their processor's specific dispute portal (Adyen, Mollie, Stripe EU) — the evidence taxonomy is portable.