Skip to content

    How Non-Processor Endpoints Are Reshaping Payment Flows

    Non-Processor Endpoints

    Payments within the e-commerce industry are in a constant state of flux. Online sellers are continuously aligning their payment methods with the preferences of buyers around the world; credit card networks continue to be confronted by the threat of online fraud, and Payment Service Providers (PSPs) are bundling all the products and services they can under a single roof to protect their margins.

    But in the second before a transaction reaches the PSP, a lot needs to happen—the customer’s identity needs to be confirmed, behavior needs to be screened, inventory needs checked, and the transaction needs to be routed to the right processor.

    These types of checks can be handled by what the industry is calling non-processor endpoints. And the growth of non-processor endpoints has been swift and meaningful, with the recent evolution of Walmart Payments API a particular sign of things to come.

    A non-processor endpoint is an API or service connection that completes some element of the payment process that does not include the actual submission of a payment transaction. In today’s complex payment ecosystem, there is a need to do much more than simply collect and submit a credit card number to a PSP.

    Examples of non-processor endpoints include:

    • Payment Orchestration: A decisioning engine that assesses the proposed transaction and decides which PSP, from the merchant’s group of contracted partners, should be contacted for processing.
    • Customer Identity and Authorization: A system that attempts to confirm the identity of the customer trying to close a purchase, based on provided, stored, and inferred information.
    • Subscriptions and Billing Management: Systems to ensure that a programmed or scheduled transaction is properly calculated, charged, and submitted
    • Payment Method Lifecycle Management: A system to ensure that the payment details are up-to-date, valid, and updated if necessary; critical to reducing the charges for payments made with out of date card details.

    Marketplaces such as Amazon and Walmart offer increasingly complex non-processor endpoints, as they aim to provide capabilities that merchants can use to manage their business, from shipping options to inventory management. These capabilities entice merchants to do business through the marketplace, earning the marketplace itself royalty revenues.

    Non-Processor Endpoints Won’t Replace PSPs 

    Non-processor endpoints are not a substitute for PSPs, because the actual transactions still must be validated and closed by card networks, issuing banks, and acquiring banks. What non-processor endpoints do is optimize everything around that core transaction process.

    They guard against fraud, reduce friction, and give merchants the intelligence to route transactions to the right processor—at the right time. For example, a full-service PSP like Stripe may charge a fixed per-transaction fee in the 3% range, while another, less-comprehensive provider like Stax might charge just a small percentage over the core interchange fee after an annual fee. Knowing which processor offers the best combination of approval likelihood and cost, and chargeback risk—on a per-transaction basis—is what enables teams to execute a profitable payments operation.

    Non-processor endpoints make this kind of decision-making possible.

    Return to Top

    Real Benefits of Non-Processor Endpoints 

    To understand how non-processor endpoints work in practice, it helps to walk through two of the most common use cases: fraud detection before authorization, and retailer API routing.

    Services like Signifyd, Kount, or Sardine analyze dozens of signals in real time to produce a fraud score before a transaction ever reaches Visa or Mastercard's networks. In a vault-based model, the flow looks like this:

    • The customer's card details are tokenized at entry—the raw PAN never touches the merchant's application servers.
    • The token is routed to the fraud engine, which accesses only the data it needs via a secure proxy
    • The fraud engine returns a score or decision.
    • If the transaction passes, the token is forwarded to the PSP for processing. If not, it is declined before reaching the network.

    Walmart Payments is a particularly significant example of how marketplace APIs are evolving. For merchants operating in the Walmart Marketplace, it simplifies payments for habitual Walmart customers and allows them to switch between using bank accounts or credit cards. The transaction is tied directly to Walmart’s fulfillment and inventory system.

    In practice:

    • The tokenized card reference is routed to Walmart's API, which applies merchant-specific eligibility and fulfillment logic.
    • Once approved, the token is forwarded to the PSP, which retrieves the raw PAN from the vault to process the transaction.
    • Raw card data is accessed only at the point of processing—not stored or exposed anywhere else in the flow.

    Digital wallets like Apple Pay and Google Pay are following a similar model. Apple Pay receives a fee from the card networks for its fraud reduction effect, while Google Pay has a somewhat less privacy-focused model, and delivers more flexibility for merchants, while also still providing a service that has no incremental cost.

    Return to Top

    Building a Robust Checkout Process 

    When used well, non-processor endpoints create a smooth checkout process that delivers a robustness that customers crave. These services can ensure that:

    • The desired product is in stock (or provide a delivery date if it is on backorder.)
    • Customers are properly identified, avoiding the risk of fraudulent orders.
    • The correct payment method is prioritized in the user interface, meeting the customer’s preference, and the processing fee goals of the merchant.
    • Information is collected and securely stored for subsequent transactions, reducing purchase friction.

    Central to the equation is the ability to interoperate with multiple PSPs, which offer different payment methods, currency conversion options, and processing fee schedules. For merchants committed to a single full-service PSP like Stripe or Adyen, non-processor endpoints will tend to be centered on the same provider, and have no meaningful impact on the cost of processing.

    The trick to making a multi-processor strategy work is to add a programmable payments vault, like Basis Theory, to collect and securely store payment data. The vault sits at the center of the payments stack, creating a foundation for payment infrastructure that isn’t limited by any single provider.

    The merchants who understand this are already building. If you want to go deeper on what’s actually happening inside your PSP—and why owning your credential layer can be a game changer—read about the hidden architecture behind your PSP.

    Return to Top

    Stay Connected

    Receive the latest updates straight to your inbox