Methodology: How We Calculate Chargeback Cost
Sources, calculation framework, in-scope and out-of-scope coverage, refresh cadence, and corrections process. Every number on the site is re-derivable from the primary sources below.
Primary sources
The sources below are the canonical references the site cites. Card-network rulebooks and processor documentation are treated as primary. Vendor reports (Chargebacks911, LexisNexis, MRC, Riskified) are treated as industry composites and labelled inline where used.
In scope
- •Published per-dispute fees for the top 8 processors (Stripe, PayPal, Square, Shopify Payments, Adyen, Braintree, Authorize.net, Worldpay).
- •Card-network monitoring program thresholds, fine schedules, and exit timelines (Visa VAMP, Mastercard ECM).
- •Reason codes from Visa, Mastercard, American Express, and Discover with evidence requirements and representment timeframes.
- •Friendly-fraud share figures by sector (subscription, physical goods, digital goods, travel, gaming, food delivery).
- •Industry-average chargeback rates by merchant category with primary dispute drivers.
- •Prevention tool list-price ranges (3DS, AVS, fraud screening, chargeback alerts, recovery services).
Out of scope
- •Negotiated enterprise processor pricing (Adyen, Worldpay, and Braintree volume tiers below the published rates).
- •Internal acquirer surcharges and risk-tier upcharges that vary per merchant agreement.
- •Regional regulatory variations beyond North America, EU, and APAC headline thresholds.
- •Custom indemnification clauses and reserve requirements at the individual MID level.
- •Cryptocurrency-specific dispute mechanics (BitPay, Coinbase Commerce) outside the card-network rule set.
- •Bank-to-bank dispute resolution timelines outside the Visa, Mastercard, Amex, and Discover networks.
Calculation framework
VAMP Ratio = (Count of TC40 fraud reports + Count of TC15 disputes) / Count of TC05 settled transactions. This is count-based, not dollar-based: one $5 dispute counts the same as one $500 dispute. The April 2026 North America excessive threshold is 1.5%. Source: Visa Acquirer Monitoring Program documentation.
Total cost = Bank/processor fee + Lost product value (COGS x order value) + Shipping (physical goods) + Fulfillment labor + Dispute response time. Default assumptions: $8 labor, $25 dispute response time. The numerator on each line is the published or median industry figure; the formula is shown on every page that uses it.
Default COGS bands used in the calculator: physical goods 40%, electronics 60%, fashion 50%, digital products 10%, SaaS 15%, travel 60%, food delivery 50%, gaming 30%. These are industry averages from MRC and category-specific public filings; your actual COGS may differ. Merchants should override the default with their own number where it is known.
Prevention ROI = (Chargebacks prevented x Cost per chargeback) - (Prevention tool cost per transaction x Total transactions). Default assumption: 60% chargeback reduction at $0.07 per transaction for combined screening. 3D Secure alone delivers 70-80% reduction for fraud-related chargebacks per Verifi and Kount benchmarks; we use the lower bound.
60-80% of chargebacks are not genuine third-party fraud. The range reflects published figures from LexisNexis True Cost of Fraud (around 60%), Chargebacks911 Field Report (around 70%), and Juniper Research (up to 75%). Subscription businesses sit at the upper end; physical goods sit at the lower end. We show the range, not a point estimate.
Pricing, thresholds, and reason codes are re-verified against primary sources on the first business week of each month. The verification date is held in one LAST_VERIFIED_DATE constant imported by every page. Footer text, schema dateModified, and visible Updated stamps all read from that single source.
Refresh cadence
Monthly first-business-week pass: re-verify processor fees on each processor's own documentation, re-check VAMP and ECM thresholds, scan for new card-network bulletins. The verification date is single-sourced through one constant (LAST_VERIFIED_DATE in src/lib/schema.ts) imported by every page. Footer text, schema dateModified, and visible "Updated" stamps all read from that single source.
Out-of-cycle refreshes are triggered by:
- •Visa or Mastercard threshold-change announcements (most recent: April 2026 North America VAMP tightening).
- •Processor fee changes (most recent: Stripe Smart Disputes June 2025 launch, the $15 counter-fee introduction).
- •New reason-code additions or retirements in the Visa Core Rules or Mastercard Chargeback Guide quarterly bulletins.
- •Publication of the annual LexisNexis True Cost of Fraud Study, MRC Global Fraud Survey, or Chargebacks911 Field Report.
- •Reader corrections that flag a stale fee, threshold, or reason code (typically actioned within five business days).
Limitations
- •Per-processor fees are the published figures; enterprise-negotiated rates can be 20-50% lower at high volume.
- •Friendly-fraud share figures are survey-based and category-dependent; the 60-80% range is a portfolio average, not a precise per-merchant predictor.
- •Win-rate estimates per reason code are industry composites; merchant-specific win rates depend heavily on evidence quality and acquirer relationship.
- •VAMP and ECM thresholds are the announced positions; in practice acquirers may impose tighter internal thresholds before formal monitoring begins.
- •Prevention-tool effectiveness ranges are vendor-claimed at the upper bound; we use the lower bound in calculator output. Real-world results depend on integration quality and merchant category.
Corrections
Found a stale fee, a missing reason code, a threshold that changed, or a calculation step you would compute differently? Email [email protected] with the page URL, the figure you are flagging, and the primary source you would like cited. Substantive corrections are typically actioned within five business days; the LAST_VERIFIED_DATE moves with each substantive review.