Average chargeback rate by vertical: 2026 benchmarks

If you sell physical product online, your average chargeback rate should sit between 0.6% and 1.0%, with the all-industry average near 0.65%. High-AOV and subscription categories run hotter: electronics and CBD hit 1.0% to 2.0%, subscriptions 0.9% to 1.2%, and supplements 0.86% to 1.0%. Cross 1.5% and the card networks flag you as Excessive.

Most operators we talk to learn what a chargeback rate even is from their processor or their accountant, not from a dashboard they built. The number shows up as a surprise, usually attached to a fee line that jumped, and the first question is always the same: is this normal for my category, or is this a problem? A chargeback (a card dispute the issuing bank forces back onto you) is two things at once. It is a direct margin leak, and above roughly 1% it is a threat to the processing relationship that keeps your store online. This page sets the benchmark by vertical and shows you where the network red lines are in 2026.

What counts as a normal chargeback rate in 2026

The chargeback rate is disputes divided by transactions, times 100. A brand doing 10,000 orders a month with 80 disputes is running 0.8%. That is the number every monitoring program watches, and the denominator is transaction count, not dollar volume. That detail matters for high-AOV brands: you can hit a rate threshold on a surprisingly small number of disputes because each one is a large slice of a smaller order count.

Where your category sits drives most of the answer. General card-not-present (CNP, meaning the card was never physically swiped) retail averages 0.6% to 1.0%. Card-present retail runs closer to 0.5%. Then the categories spread out based on average order value, billing model, and how much buyer regret the product invites. Electronics and CBD sit at the top of the mainstream range because they pair high AOV with strong fraud interest. Subscriptions run hot because recurring billing generates disputes that one-time purchases never do. Online info products are the outlier, with published rates from 1.0% to 4.8%, reflecting aggressive direct-response funnels rather than traditional education.

The table below carries the full benchmark with the typical range, a risk tier, and the main thing driving disputes in each category. Use the range, not just the midpoint: within-vertical variance is wide, and an offer type (free trial versus full price, subscription versus one-time) can move you a full band.

Supplement brands running near the top of the 0.86% to 1.0% health and wellness band should cross-reference the supplements financial benchmark for the gross margin and unit economics context that shows how much chargeback cost actually compresses the P&L at different AOVs.

One more thing the benchmark hides: the trend is up. Sift's 2025 index shows the average network-wide rate climbing from 0.17% in Q1 2025 to 0.26% in Q3 2025, a 53% jump in nine months. That figure is an all-rails network average across all merchant types including card-present retail, which is why it sits well below the 0.6% to 1.0% CNP benchmarks in the table above; the signal is the direction of travel, not the absolute level. A brand that held a comfortable 0.7% through 2024 can drift toward 1% without changing anything about how it operates. The benchmark is a moving target, so re-check yours every quarter rather than treating last year's number as fixed.

The card network red lines: Visa VAMP and Mastercard ECP

Above your category benchmark sits a harder set of numbers: the thresholds where Visa and Mastercard stop treating you as a normal merchant. This is where a margin issue turns into a processing issue.

Visa replaced its old VDMP and VCMP programs with VAMP, the Visa Acquirer Monitoring Program. From June 2025, the Excessive Merchant threshold was 2.2% for US, Canada, EU, and Asia-Pacific merchants (CEMEA remains at 2.2% through 2026). On April 1, 2026, that drops to 1.5% for US, Canada, EU, and Asia-Pacific. VAMP also changes the math: instead of fraud and non-fraud disputes being tracked separately, it counts combined fraud plus non-fraud disputes against your settled transaction count. Mastercard's Excessive Chargeback Merchant (ECM) tier flags any merchant at 1.5% (150 basis points) or higher with 100 to 299 monthly disputes, with fines starting in the second month. Cross 3.0% with 300+ disputes and you are in High Excessive territory, where termination is on the table.

Here is the nuance most under-$30M brands miss. The 1,500-dispute count threshold for Visa's Excessive designation is high enough that you may never hit it by volume alone. That does not make you safe. The rate thresholds apply at the acquirer level regardless of count, and acquirers make relationship calls based on rate. When we talk to founders who got "the call," it almost never came at 1.5%. It came when their rate sat near 0.9% to 1.0% for a couple of months and the acquirer's risk team decided to get ahead of it. Treat 1% as your real ceiling, not the published network number.

The real cost of a chargeback is not the $15 fee

The dispute fee on your statement is about $15 at Stripe, Shopify Payments, and Braintree. That number is a trap, because it makes a chargeback look like a minor processing cost. The all-in number is roughly $128 per chargeback, per Mastercard's 2025 estimate.

The gap is everything you already spent before the dispute landed. The chart below shows an illustrative industry-blend component breakdown -- product cost, marketing and CAC, ops overhead, and dispute and network fees. The individual line items are drawn from a blend of published cost research; Mastercard's own published split is $82 in internal costs and $46 in third-party fees, while Chargebackgurus' breakdown totals $103 with a $25 dispute fee. Every version of the math reaches the same conclusion: the $15 dispute fee on your statement is less than 15% of the real number. Signifyd frames the total impact as roughly $3.75 lost for every $1 of chargeback face value.

This is where the rate connects to the P&L. When we sit down with founders in the $5M to $30M range as their fractional CFO, the actual payment-processing line almost always lands at 2.8% to 3.5% of net revenue, well above the headline Shopify or processor rate of 1.5% to 2%. Chargebacks, fees, and reserves are a big part of that gap. A sudden, unexplained jump in the processing line is itself a dispute signal worth a same-week review, not a quarter-end one. If you are running 0.9% on a $20M brand at a $60 AOV, that is roughly 250 disputes a month at $128 each, about $32,000 a month, or close to $384,000 a year walking out the door. That is real money, and most of it is invisible until you add back the CAC. If unexplained jumps in your processor deposits brought you here, the processor payout timing and cash flow guide covers why what hits your bank account rarely matches the revenue line you expect.

The $15 dispute fee is the part you can see. The other $113 is the order you already paid to win and fulfill. Price a chargeback at $128, not $15, and the case for prevention stops being a debate.

Friendly fraud: the dispute you can't underwrite away

Roughly 43.8% of chargebacks are friendly fraud by merchant estimate (Chargebacks911 2024 Field Report). Visa's own data puts the global figure at 20% to 30%, Riskified says about 50%, and first-party fraud grew from 15% to 36% of all reported fraud in a single year. However you count it, a large share of your disputes come from your actual customers, not strangers with stolen cards.

That changes the playbook, because friendly fraud does not respond to the tools you bought to stop true fraud. A 3DS challenge or a fraud-scoring filter will not stop a real customer who forgot a subscription renewal, regretted a purchase, or decided a dispute is a faster refund than your support queue. The categories most exposed are exactly the hot ones from the benchmark: subscription digital, high-AOV electronics, and supplements. Digital subscriptions alone account for about 20% of first-party misuse incidents, second only to groceries.

For subscription operators this is the core exposure, and it is usually self-inflicted. The pattern we see again and again is a recurring-billing flow that makes it genuinely hard to skip, pause, or cancel. One operator running a supplement subscription described the cancel path as "really difficult to crack" by design, which felt like retention strategy until the disputes showed up on the renewal charges. When customers cannot find the off switch, they call their bank instead, and a friendly-fraud chargeback costs you far more than the saved subscription was worth.

Prevention and representment levers that move the number

Two separate plays, because the two kinds of dispute have different root causes.

On prevention, the highest-yield moves are unglamorous. Make the billing descriptor obviously match your brand name so customers recognize the charge. For subscriptions, build a self-service skip, pause, and cancel flow that takes fewer clicks than a phone call to the bank. Send a renewal reminder before you bill, not after. Use 3DS authentication to cut true (stranger) fraud, knowing it does nothing for friendly fraud. And set a clear refund-versus-dispute policy: a proactive refund costs you COGS, while a chargeback costs you $128, so winning the race to refund the genuinely unhappy customer is almost always the cheaper outcome.

On representment (fighting the dispute after it lands), your win rate depends heavily on what you sell. Physical-goods merchants win roughly 40% to 50% of representments because delivery confirmation and signature are strong evidence. Digital goods win only 20% to 30%, because "I delivered access" is harder to prove than "I delivered a box." Build your evidence package around that reality: tracking and delivery proof for physical, and login timestamps, usage logs, and accepted terms for digital.

One operator habit worth copying: treat chargeback monitoring as a daily practice at launch, not a monthly reconciliation. As one founder put it about a new launch, the goal was to "set up daily monitoring for this when it launches so we can react quick." The networks measure you monthly by rate, but by the time a monthly number is bad, you have already booked a month of disputes you cannot undo. Daily or weekly alerting on dispute volume buys you the time to find the offer, the funnel, or the billing bug driving it before your rate crosses a line.

Sources and methodology

Vertical benchmark ranges come from public chargeback-ratio content, not a single paywalled report. The category figures (retail 0.52% to 1%, travel 0.89% to 2%, electronics 1% to 2%, digital goods 0.66% to 1.5%, subscription 0.9% to 1.2%, health and wellness 0.86% to 1%, food and beverage 0.5% or less) are compiled from the published benchmark pages at Chargebacks911, ClearlyPayments, and 2Accept. No single unified vendor benchmark table with full methodology was publicly available, so ranges should be read as directional category guides rather than audited statistics.

Card network thresholds are taken from the primary network documentation where possible. Visa VAMP merchant and acquirer thresholds and effective dates come from the Visa Acquirer Monitoring Program Fact Sheet 2025, with the April 2026 reduction to 1.5% (applicable to US, Canada, EU, and Asia-Pacific merchants) corroborated by Signifyd's program explainer. Mastercard ECM and HECM tiers are documented by Chargebackgurus.

The cost-per-chargeback figure is Mastercard's own estimate. The $128 all-in number is from Mastercard's 2025 State of Chargebacks; Mastercard publishes the split as $82 in internal costs and $46 in third-party fees. The component chart uses an illustrative industry-blend breakdown drawn from Chargebackgurus' published cost breakdown ($103 total with a $25 fee) and Mastercard's residual estimate; it is not Mastercard's own line-item table and should be read as an estimate, not an audited split. The per-dispute fees of about $15 are the published US rates for Stripe, Shopify Payments, and Braintree.

Friendly-fraud and trend figures come from network and industry index data. The 43.8% merchant estimate is from the Chargebacks911 2024 Field Report; Visa's 20% to 30% range is from its friendly-fraud insights; and the rising rate trend (0.17% to 0.26%, Q1 to Q3 2025) is from the Sift Digital Trust and Safety Index, Q4 2025 -- an all-rails network average across all merchant types, not a CNP-only vertical figure. Representment win rates by goods type are from Solidgate's chargeback win-rate data.

Limitations. Supplement-specific rates are filed under "health and wellness" by most sources, so the 0.86% to 1.0% range is a proxy; free-trial and continuity supplement offers likely run higher. Government bodies (the Federal Reserve, CFPB) do not publish merchant-side chargeback rates by vertical. The 4.79% online-education figure reflects aggressive direct-response funnels, not traditional education, and should be read with that context.

Frequently asked questions