Justt 2026 Benchmark: Service Disputes Surge for Sellers

The shift, in numbers What counts as a service dispute Industry spotlights Evidence discipline, not fraud tools What this data does not show FAQ

The shift, in numbers

For over a decade, a merchant's chargeback strategy was mostly a fraud strategy. Justt's 2026 Chargeback Benchmark Report, The Fraud to Service Shuffle, says that era is ending — on its own platform's data. The headline finding, announced October 27, 2026: disputes over undelivered orders and subscriptions customers thought they had cancelled nearly doubled in a year, while fraud claims fell to a two-year low. Every figure in this article is Justt platform data — contested chargebacks processed through Justt's platform from July 2024 to June 2026, with trends checked against a fixed cohort of 165 merchants — never an industry census. Read it as a yardstick from one large platform, not as a count of all chargebacks everywhere.

Justt platform measure	Q2 2025	Q2 2026
Service disputes (Product Not Received + Cancelled Recurring), share of contested chargebacks	13.0%	24.1%
Fraud-coded claims, share of contested chargebacks	74.6%	60.4% (two-year low)
Product Not Received alone, share of all chargebacks	7.1%	13.3%
Cancelled-subscription disputes, share of disputes	5.9%	10.8%
Median merchant's contested chargeback volume, year over year	—	+13% (top quarter +68% or more)

Two cautions belong in the same breath as the headline. First, these are shares of contested chargebacks — disputes whose merchants chose to fight them on Justt's platform — so they describe the mix of fought cases, not every dispute filed. Second, the median merchant's volume rose 13% at the same time, so fraud's falling share does not prove fewer fraud disputes in absolute terms. Justt's chief executive put the merchant implication bluntly: cardholders increasingly treat disputes as a customer service channel — why chase a merchant about a late package when a call to the bank claws the money back?

What counts as a service dispute

Justt groups two claim types under "service disputes": Product Not Received, where the customer says the order never arrived, and Cancelled Recurring, where the customer says a subscription kept billing after they cancelled it. Together they moved from 13.0% to 24.1% of contested chargebacks in a year. Separately, Product Not Received grew from 7.1% to 13.3% of all chargebacks, and cancelled-subscription disputes grew from 5.9% to 10.8% of disputes. The practical point for a seller is what these claims are not: they are not claims that the payment was unauthorised. A merchant who answers a not-received claim with proof that the cardholder authorised the purchase has answered the wrong question. The question the issuer is actually weighing is whether the goods arrived, or whether the billing stopped when it should have.

The UK and Ireland are the one region where service disputes already outnumber fraud in Justt's data — fraud accounts for just 45% of chargebacks there — which Justt links to Strong Customer Authentication shifting fraud liability to card issuers. Sellers operating in that region should expect the service-dispute mix to arrive first, not last. Everywhere else the direction of travel is the same; only the starting point differs.

Industry spotlights

Software subscriptions: +222%. Same-merchant chargeback volume in software subscriptions more than tripled, and the median merchant in the category still grew 51%, so this is not a few outliers. Cancellation disputes are the category's defining problem — nearly one in five software-subscription disputes is a cancelled-subscription claim — and Justt calls the category the hardest major category to win. Its conclusion is blunt and worth repeating: clear billing descriptors and painless cancellation flows beat fighting the disputes.

Ticketing: the most expensive disputes. Ticketing has the highest average disputed purchase in the report at $593, up 33% year over year, with a median chargeback of $292 — premium events dominate. Ticketing is also the least fraud-driven industry in the report, at just 34% fraud share. A ticketing seller's dispute problem is therefore mostly a service problem: what was promised, what was delivered, and what the refund terms said.

Delivery: the sharpest mix change. In delivery, fraud fell from 79% to 66% of disputes, and Product Not Received nearly tripled to 15% of delivery disputes. Customers increasingly dispute whether the order arrived, not whether they authorised the charge. For a delivery seller, the tracking record is the case.

Fintech and financial services: volume doubled, and service disputes cost more. Same-merchant volume more than doubled (+107%), with the median merchant up 71%. And in this industry's data, service disputes are the costlier kind: non-fraud chargebacks average $128, compared with $85 for fraud chargebacks. That $128-versus-$85 comparison is a fintech-section figure in Justt's report, not a whole-platform average — do not quote it as one. Its lesson generalises carefully: a service dispute is not a cheaper dispute. It carries the same fee and lost sale, on a claim your fraud stack cannot rebut.

Evidence discipline, not fraud tools

If the dispute mix has moved, the evidence file has to move with it. Fraud tools — device checks, velocity rules, 3-D Secure — prove authorisation. Service disputes need a different file, kept as a matter of routine rather than assembled after the dispute lands:

For not-received claims: the carrier tracking record with delivery status and address, proof of delivery where you have it, the fulfilment timestamp, and the customer's own delivery messages. If you cannot show delivery, say so early and refund — a fought not-received case with no tracking is a fee paid to lose.

For cancelled-subscription claims: the cancellation record itself — when the customer cancelled, through which channel, and what confirmation was sent — plus the billing descriptor the customer actually saw on their statement and the cancellation flow as it stood on that date. Screenshots of today's flow do not prove last quarter's flow; keep dated records of flow changes.

For every service dispute: a reachable customer-service contact before the bank becomes the contact. Justt's thesis is that customers dispute because disputing is easier than complaining. A seller whose support replies in hours, whose descriptor names the business plainly, and whose cancellation takes two clicks instead of a phone call will simply receive fewer of these claims. That is prevention, not representment — and in a mix where nearly a quarter of fought cases are service claims, prevention is where the volume is.

Two companion tools put numbers on the fight-or-refund choice for your own cases: the Fight-or-Refund Decision Calculator and the Chargeback Cost Calculator. For subscription evidence specifically, see the subscription evidence checklist, and for the general decision, the fight-or-refund guide.

What this data does not show

Honest limits, stated plainly. Justt's report covers contested chargebacks on its own platform across seven industries; it does not disclose absolute win rates or total dispute counts beyond "millions," and its processor, issuer and network names are not published — only the gaps between them (fraud win rates varying by more than 30 percentage points between processors and more than 35 points between major US card issuers for the same claim type and quarter). Do not use this report to pick a processor: the report does not name one. Do not use it as an industry census: a platform cohort of 165 merchants, however carefully fixed, is not the whole market. And do not read a share as a count: with median volume up 13%, fraud's share can fall while fraud cases still rise. What the report does establish, within its stated scope, is a direction that matters for how sellers spend their next dispute dollar: toward delivery proof, cancellation records and customer service that answers before the bank does.

FAQ

What does Justt count as a service dispute?

Two claim types grouped together: Product Not Received and Cancelled Recurring (subscriptions the customer says they cancelled). In Justt's platform data those two types rose from 13.0% to 24.1% of contested chargebacks between Q2 2025 and Q2 2026. They are not fraud-coded claims, so fraud-screening tools do not answer them.

Does a falling fraud share mean there are fewer fraud disputes?

Not necessarily, and Justt's own numbers warn against reading it that way. These are shares of contested chargebacks, not counts: the median merchant in the same 165-merchant cohort handled 13% more contested chargebacks year over year, so a smaller fraud share can still sit on top of a larger total. Do not cut fraud controls on the strength of a share alone.

What evidence decides a service dispute for a seller?

Proof of the service itself: delivery scans and proof of delivery for not-received claims, and the cancellation record — the date, the channel and the confirmation sent — plus the billing descriptor and the cancellation flow for subscription claims. A fraud argument (this customer authorised the payment) does not answer a service claim (the package never arrived, or the subscription kept billing after cancellation).