Fraud reports usually describe losses and chargebacks. A less visible question is how many genuine customers were prevented from buying by the controls intended to protect the business.
A blocked order can look like a successful intervention in a dashboard. If the customer was legitimate, however, the merchant may have lost revenue and damaged the relationship. Both outcomes deserve attention.
The difficulty is measurement. An unsuccessful payment does not prove that a customer was genuine, and an order that looks ordinary is not necessarily safe. False declines require investigation, rather than a simple count of rejected transactions.
A review should separate merchant-side fraud blocks from issuer declines, authentication failures and technical errors. Those categories may produce the same customer-facing message while requiring very different remedies.
Merchants can examine a sample of blocked orders for evidence such as established customer relationships, verified account activity and subsequent successful purchases. Where appropriate, customer follow-up can provide additional context. The result is an estimate to improve with further evidence, not a definitive label based on appearance alone.
Gordeon‘s emphasis on a merchant’s specific payment requirements provides a useful frame for this review. A business selling high-value B2B goods may need different controls from a consumer retailer, even when both accept cards through similar checkouts.
Differences in behaviour do not justify blanket exemptions. A returning customer can still be compromised, and a corporate order can still be fraudulent. Segmentation should improve decisions while preserving controls against genuine threats.
An intermediate response may be more useful than an immediate block. Depending on the payment method and risk, that could mean authentication, account verification or a manual review before fulfilment. Each step introduces friction, so the merchant should measure whether it improves the overall result.
Changes should be introduced cautiously and evaluated over enough time for disputes to emerge. More approvals alone do not establish success if fraud losses rise later.
A complete performance review considers completed purchases, customer friction, confirmed fraud and chargebacks together. The question is whether the controls protect the business while allowing genuine customers to pay.
Written in partnership with Tom White
