The endpoint economy: why the future of payments is not built around one method

The endpoint economy is the payment world we already live in: money no longer moves toward one dominant instrument, but outward across many – bank accounts, wallets, cards, merchants and now digital assets – each one a valid place for a customer to send or receive funds. Most payment strategy still assumes the opposite, that one method will eventually beat the others. It will not.

What is the endpoint economy?

An endpoint is simply anywhere money can land: a bank account reachable by SWIFT or a national payment system, a mobile wallet, a card, a merchant till, a digital asset address. The endpoint economy is what results when customers stop caring which of these a business supports and start expecting all of them. A migrant worker sending money home does not think in terms of remittance rails. She thinks: my sister needs cash by Friday, and her phone has a wallet on it. The method is a means. The outcome is the point.

This sounds obvious once said aloud, which is exactly why it gets ignored. Most payment organisations, ourselves included at various points in our history, were built the other way round.

Why were payments built around one method?

For most of the last fifty years, payment businesses were defined by the instrument they issued or the network they belonged to. A card scheme built its entire commercial model around the card: acquiring, issuing, interchange, loyalty. A bank built its model around the current account. A remittance company built its model around cash pickup counters. Each of these businesses had every incentive to keep customers inside their own instrument, because that instrument was the product.

That model worked when a customer’s financial life was genuinely contained within one instrument. It stops working the moment a customer holds three or four accounts of different kinds across different providers, which describes most of the world’s population today. A model built to defend one endpoint cannot serve a customer who lives across five.

What counts as an endpoint today?

The list is longer and more varied than most payment strategy documents acknowledge:

• Bank accounts, reachable through SWIFT for cross-border transfers or through national payment systems for domestic settlement.
• Mobile wallets, now the primary financial account for a large share of the world’s population, particularly outside the countries that built the twentieth-century banking system.
• Cards, still dominant at physical point of sale in many markets even as their share of e-commerce erodes.
• Merchants, who increasingly need to receive funds directly rather than through a personal account, particularly in QR-based and small-business payment flows.
• Digital assets, including stablecoins, which are becoming a genuine settlement endpoint rather than a speculative instrument (we set out the mechanics of this in our piece on stablecoins as a rail).

None of these is going away. None of them is about to absorb the others. That is the part payment strategy keeps getting wrong.

Why will no single payment method win?

The data settles this more firmly than intuition does. Worldpay’s Global Payments Report 2026 puts digital wallets at around 56% of global e-commerce transaction value and about a third, 33%, of in-store spending, and forecasts both figures will keep rising through 2030. In China, wallets already account for close to 89% of e-commerce value. Read quickly, this looks like the story of wallets winning outright.

Read carefully, it is the opposite. A method that commands 89% of e-commerce value in one market and a much smaller share of point-of-sale spending, even in that same market, is not replacing cards or bank transfers. It is sitting alongside them, dominant in one context and marginal in another. Meanwhile mobile money, a distinct category from e-commerce wallets, processed over 2 trillion dollars in 2025 across 2.3 billion accounts globally, according to GSMA – a scale that rivals many national banking systems, built entirely outside the traditional bank account. Cards remain the default at physical tills in markets where wallets dominate online. Bank transfers remain the default for salary payments almost everywhere. Every method is winning somewhere and losing somewhere else, simultaneously, permanently. That is not a transition period. That is the terminal state.

What is payment orchestration?

If no method wins, the competitive question changes. It stops being “which instrument should we build our business around” and becomes “how many instruments can we reach through one connection.” That second question is payment orchestration: the discipline of routing a single payment intent to whichever endpoint the recipient actually holds, without asking the sender to know or care which one that is.

Customers arrive through any endpoint and leave through any other. The value is the layer in the middle.

Orchestration is not a feature bolted onto a payment product. It is a different business altogether. A company that owns one endpoint competes by making that endpoint better. A company that orchestrates across endpoints competes by making the choice of endpoint disappear for the customer, while making the underlying complexity someone else’s problem to manage well.

What does an orchestration layer actually have to do?

Reaching every endpoint sounds simple stated as a sentence and is not simple to build. An orchestration layer has four jobs, and skipping any one of them turns “we support wallets and banks” into a set of brittle point-to-point integrations rather than a coherent capability.

Routing

The system needs to know, for a given recipient identifier in a given country, which endpoint that identifier resolves to, and which of several available paths to that endpoint is fastest, cheapest or most reliable right now. Routing decisions have to be made per transaction, not per market, because the right path in one corridor at one moment is not the right path an hour later if a partner’s system degrades.

Foreign exchange

Cross-border orchestration is inseparable from foreign exchange. Every endpoint sits in a local currency, and the rate applied between the sender’s currency and that local currency has to be transparent, competitive and consistent regardless of which endpoint the money eventually reaches. Customers should get a comparable deal whether the money lands in a bank account or a wallet, not a worse one because the wallet route is harder to price.

Compliance

Each endpoint type carries its own regulatory obligations, and a bank account, a wallet and a digital asset address are not subject to identical rules even within the same country. An orchestration layer has to run sanctions screening, know-yourcustomer checks and transaction monitoring consistently across all of them, so that reaching more endpoints does not mean carrying more compliance risk per endpoint. We cover the inclusion and policy dimension of this properly in our piece on interoperability and financial inclusion; here the point is narrower: compliance has to scale with reach, not fragment against it.

One contract, many endpoints

The commercial promise underneath all of this is that a partner signs one agreement and one technical integration, and gets access to every endpoint the orchestration layer already reaches, rather than negotiating and integrating separately with each wallet operator, bank and card network in every market it wants to serve. This is the actual product. Everything above it is plumbing that makes the promise true.

We built TerraPay’s network on exactly this logic, and the scale is a useful illustration of what orchestration looks like once it is working: one connection today reaches over 3.7 billion mobile wallets and over 7.5 billion bank accounts across 158 countries, with a 99.9% success rate, alongside SWIFT connectivity and the Xend network for higher-value flows. The point of citing this is not that the number is impressive. It is that none of those endpoints required the sender to know which one the recipient held.

Who wins in the endpoint economy?

Not the business with the best wallet, the best card product or the cheapest bank rail, taken alone. The businesses that win are the ones that stop treating endpoint choice as a battle to be fought and start treating it as a variable to be absorbed. That requires giving up the comfort of owning a single instrument end to end, which is a harder organisational shift than it sounds, because it means the company’s value no longer sits in the instrument at all. It sits in the connective layer above every instrument.

This has a direct implication for how a product organisation allocates its engineering effort. Time spent building a superior proprietary wallet, when threequarters of the addressable customers already hold a perfectly good wallet from someone else, is time not spent building the routing, foreign exchange and compliance capability that would let those customers use their existing wallet through you. The instrument is rarely the differentiator. The reach is.

We think this reframing will keep proving itself out over the next several years, alongside a handful of other structural shifts in how money moves, which we set out separately. The endpoint economy is not a trend among trends. It is the condition the other trends are happening inside.