Embedded payments are maturing into optionality models

A lot of software companies used to talk about embedded payments like a light switch: either you become the payments company inside your product, or you stay a referral partner forever. The newer pitch is a dimmer switch: start by sending merchants to a processor, then add more payment features inside the product, and only become a full payment facilitator if the volume is big enough to pay for compliance, fraud, and support teams. (stripe.com) (worldpay.com) That shift starts with a basic fact about how platform payments work. A software platform can let a seller sign up with a processor directly, or it can sit in the middle and manage onboarding, money movement, and reporting across many sellers. (docs.stripe.com) (docs.adyen.com) Sitting in the middle buys control, but it also buys chores. Adyen’s platform documentation says sellers must complete Know Your Customer verification before they can be paid out, which means the platform has to care about identity checks, sanctions screening, and payout holds instead of just product features. (docs.adyen.com) That is why “full payment facilitator” is no longer the default end state in every sales deck. Stripe’s guide to payment facilitators frames the model as bringing payments in-house, and its migration guide warns that changing charge types affects refunds, chargebacks, fee responsibility, and reporting data. (stripe.com) (docs.stripe.com) So the market is settling on staged migration. A software company can begin with referrals, add embedded onboarding widgets and dashboards, then move into acceptance, payouts, and revenue share once customers are already transacting inside the product. (docs.stripe.com) (docs.adyen.com) (stripe.com) Vendors are selling that path as optionality instead of destiny. Worldpay now markets embedded payments “from integrations to registered PayFac solutions,” which is a way of telling software companies they do not need to swallow the hardest operating model on day one. (worldpay.com) The reason this framing is landing now is that the product surface has widened. Stripe Connect is no longer just card acceptance; its platform pitch includes instant payouts, tax collection, financing, and expense cards, so a software company can deepen monetization before it ever decides to own the full payments stack. (stripe.com) (docs.stripe.com) At the same time, newer payment experiences are pushing the same lesson from the crypto side. 0x says its Gasless API lets developers offer swaps and token approvals without asking users to hold network fees first, which turns a two-step wallet ritual into something closer to tapping “buy” in a normal app. (0x.org) Fireblocks made a similar bet when it launched gas-free transactions on Ethereum Virtual Machine chains in 2025. The company described the feature as a way to streamline stablecoin payments and other consumer-facing digital experiences, which is the same embedded-payments logic in a different wrapper: remove the extra payment step, keep the user inside the product, and hide the operational complexity behind infrastructure. (fireblocks.com) eCash shows where this can go at the low-value end of the market. Recent coverage of its eToken standard describes token transfers and micropayments designed to avoid the usual smart-contract approvals and gas friction, which makes tiny transactions more plausible in apps where a $1 action cannot carry a $5 setup cost. (changenow.io) Put together, the story is less “every platform will become a bank” than “every platform wants a migration path.” The winning sales pitch in 2026 is not all-in embedded finance on day one; it is a ladder where each rung adds conversion, reporting, or revenue, and the platform stops climbing when the next rung costs more in risk operations than it returns in margin. (docs.stripe.com 1) (docs.stripe.com 2) (worldpay.com)