Account-to-Account Payments at Checkout
Why A2A is winning in some markets and struggling in others, and what needs to change for it to become the default payment method for e-commerce
The Fundamental Experience Gap
When you buy something online today, the checkout experience is still optimised for card payments. You enter card details, expiry date, CVV. You approve a transaction. The payment processes. This is a 30-second flow that has been perfected over two decades. The card network has trained consumers to expect this experience. They arrive at checkout, they expect cards to be available, and they expect the transaction to complete in seconds.
Account-to-Account payments at checkout break this expectation. The experience is fundamentally different. Instead of entering card details, you select your bank. The checkout redirects you to your bank's authentication page. You log in (or approve the payment through your banking app). You return to the merchant. The transaction completes. The time required is longer. The number of steps is higher. The context switching is jarring. For a consumer habituated to card checkout, A2A is friction.
That friction is the core problem A2A faces at the checkout layer. It is not a technical problem. It is not a cost problem. It is a user experience problem. Every extra second in checkout increases abandonment. Every redirect increases dropout. Every time a customer must switch context from the merchant to their bank and back, some percentage of them will leave the checkout and never complete the purchase. This is why A2A remains a marginal payment method at e-commerce checkout despite having massive adoption in other use cases.
A2A at checkout loses because the user experience is fundamentally worse than cards. It is not a feature gap or a market education gap. It is a structural friction problem that cannot be solved by education alone.
Where A2A Is Already Winning
Despite the checkout friction, A2A payments are massive in specific use cases. Understanding where A2A dominates helps explain where it will go next and what barriers exist.
Bill Payments and Recurring Charges
Utility companies, subscription services, and insurance providers process massive volumes through A2A. Stripe reports that recurring A2A payments have conversion rates comparable to cards and substantially lower failure rates. The reason is straightforward: customers do not interact with A2A for recurring bills. They set it up once. They forget about it. The payment happens in the background.
For companies collecting recurring payments, A2A is superior to cards because it does not carry chargeback risk and does not require the payer to update payment methods when a card expires. A cancelled card means the recurring payment fails. A closed bank account means the payment fails, but the failure mode is cleaner: the payment bounces, and both parties are aware of it immediately. For subscription businesses processing thousands of recurring charges monthly, the operational advantage of A2A is enormous.
Person-to-Person Transfers
A2A powers the backend of peer-to-peer payment apps. When you send money to a friend through Venmo, PayPal, or Cash App, the money moves through A2A rails if both parties have bank accounts with integrated institutions. The P2P app sits on top of the A2A network and handles the user experience friction. The consumer experience is simple: tap the recipient, enter the amount, confirm. The fact that it is A2A underneath is invisible.
P2P apps demonstrated that A2A can achieve massive adoption if the user experience is optimised correctly. Venmo moved $200+ billion in volume in 2024. Zelle, the bank-backed P2P network, processes over $1 trillion annually. The P2P layer proved that consumers will use A2A at scale if the experience does not require them to interact directly with bank authentication and account selection.
High-Value B2B Payments
Business-to-business payments have migrated to A2A over the past decade. Large enterprises with significant invoice volumes process through bank wires, ACH, or real-time payment networks. The reason is pure economics: at $10,000 or $100,000 per transaction, the card interchange fee (1 to 3 percent) becomes unaffordable. A flat fee of $10 to $50 for an A2A payment is trivial compared to the savings.
B2B A2A does not require elegant user experience because both parties are operating in controlled environments. The business sending payment has payment authorisation workflows. The business receiving payment has accounting integration. Both are willing to tolerate friction if the economics make sense. And the economics overwhelmingly favour A2A.
Government Disbursements and Social Programs
Governments worldwide are using A2A to disburse benefits, tax refunds, and social payments. The US Treasury uses A2A for direct deposit of tax refunds. UK DWP (Department for Work and Pensions) uses A2A for universal credit and jobseeker payments. The volume is enormous. The cost per transaction is critical because governments are processing millions of disbursements monthly.
Government use of A2A also solves the financial inclusion problem: A2A requires a bank account, which creates incentives for citizens to maintain banking relationships. It also eliminates the middle layer of payment card issuers or check processors, reducing the total cost to the government.
Where A2A Struggles at Checkout
E-commerce checkout is where A2A loses decisively against cards. The reasons are well understood but not easily solved.
Redirect Friction
When a customer selects A2A at checkout, they are redirected away from the merchant site to their bank. The browser leaves the merchant domain. The authentication happens at the bank. The customer returns to the merchant. This redirect creates three problems. First, it is slower. A redirect requires loading a new domain, running authentication, and redirecting back. That takes 5 to 15 seconds for a smooth flow. Cards are instant.
Second, it creates trust anxiety. The customer sees a URL change. They are on their bank's domain, not the merchant's domain. Some customers perceive this as a security risk (it is actually safer, but perception matters). A small percentage abandon the checkout the moment the redirect happens.
Third, redirect creates technical complexity. Some payment processors do not handle redirects elegantly. Network interruptions during the redirect can leave the transaction in an ambiguous state (did the payment go through? Did it fail?). Mobile applications handle redirects poorly compared to desktop browsers.
No Saved Credentials
With card payments, the merchant can save the card details and offer "one-click checkout" on repeat purchases. The customer clicks once, the payment processes, the transaction completes. Saved credentials are a core feature of card checkout. They are why repeat customers do not re-enter payment details.
A2A has no standardised saved credentials mechanism. The customer selects their bank, authenticates, completes the payment. On the next purchase, they must repeat the entire flow. They select the bank again. They authenticate again. The redirect happens again. This is why A2A checkout has much higher abandonment on repeat customers.
Some payment processors have attempted to solve this using Variable Recurring Payments (VRPs) and persistent consent tokens. The first payment grants the merchant access to future payments from that customer, which allows the merchant to initiate payments without the customer's explicit involvement. But VRPs are not widely available, not all banks support them, and customer trust in persistent consent is still building.
Bank App Switching
Mobile checkout is increasingly important for e-commerce. Mobile conversion rates are lower than desktop, but mobile volume is higher. Mobile A2A checkout has a specific problem: bank app switching. When a mobile customer selects A2A at checkout, the checkout redirects to the bank's mobile authentication. If the customer has the bank's app installed, the browser redirects to the app. The customer authenticates in the app. The app must then redirect back to the checkout. This requires the merchant's checkout to have proper deep link handling, the bank's app to correctly implement the redirect, and the operating system to handle the app-to-web transition correctly. Any of these can fail silently.
Many customers do not have their bank's app installed. The redirect goes to the mobile web login, which requires them to remember their banking password. They attempt authentication. The session times out. The redirect fails. The customer abandons the checkout. From a conversion perspective, mobile A2A is still significantly worse than mobile card checkout.
No Saved Bank Selection
With cards, customers trust that their card is saved. They do not worry that the merchant will forget it. With A2A, even if the payment succeeds, the customer cannot know whether their bank is saved for future use. Some providers save bank selection, some do not. There is no standard. A customer might complete a purchase with their bank, return three months later, expect their bank to be pre-selected, and find that they must choose it again.
This lack of predictability creates friction on repeat purchase. The customer experience is inconsistent. They lose confidence in whether A2A will work smoothly on the next purchase. This is a small problem compared to the redirect friction, but it compounds the perception that A2A at checkout is not ready for consumer use.
Conversion Rate Reality: A2A vs. Cards
The data on conversion rates is unambiguous. Payment processors who offer both cards and A2A report that card conversion rates are substantially higher. Industry benchmarks vary, but the pattern is consistent: cards convert 2 to 5 percent better than A2A. For a merchant with a two percent checkout conversion rate, offering A2A instead of cards would drop conversion to 1 percent or lower.
The gap is larger for first-time customers and smaller for returning customers. Returning customers who have used A2A successfully before have lower abandonment on A2A (because they know what to expect). First-time customers abandon at much higher rates because the experience is unfamiliar.
However, the total revenue impact is more nuanced than conversion rates alone suggest. A2A has lower total failure rates (cards have chargebacks, A2A often does not). A2A has lower payment processing costs. These economics mean that even with lower conversion rates, A2A can be more profitable per transaction for specific merchants. A subscription business processing recurring charges would prefer A2A even with lower initial conversion because the recurring success rate and lower failure rate more than compensate.
For e-commerce checkout, cards convert 2 to 5 percent better than A2A. For recurring payments and subscription billing, A2A converts as well or better than cards because the customer only checks out once, and recurring success rates are higher.
Where A2A Already Won: iDEAL and Bancontact
The Netherlands and Belgium have already solved the e-commerce A2A checkout problem. In both countries, A2A payment methods have achieved dominant market share. Understanding how they did it provides a playbook for the rest of the world.
iDEAL in the Netherlands
iDEAL is an A2A payment method optimised for Dutch e-commerce. It launched in 2005 and has achieved extraordinary adoption. Today, iDEAL accounts for 70 percent of Dutch e-commerce transactions. Credit cards account for less than 10 percent. iDEAL is the default payment method for Dutch online shopping.
Why did iDEAL succeed? First, banking infrastructure in the Netherlands is unified. Most Dutch consumers have accounts with one of a few major banks. The bank selection at checkout is simple: the customer sees three to five banks and immediately recognises their bank. The friction is minimal.
Second, iDEAL was built with e-commerce in mind from the start. The user experience was optimised for checkout. Banks invested in ensuring that authentication was fast and reliable. The payment success rate is over 99 percent. Failures are rare.
Third, iDEAL was backed by the banking industry. All major Dutch banks were required to implement it and support it. There was no fragmentation. There was no choosing between different A2A providers. It was a single standard that all banks followed. This unified implementation reduced complexity for merchants and eliminated the problem of bank interoperability.
Fourth, government and regulatory support. Regulators in the Netherlands encouraged the adoption of A2A payment methods over card networks. This created a push from the regulatory side that complemented the pull from merchants and customers.
Bancontact in Belgium
Bancontact followed a similar pattern in Belgium. It became the dominant payment method for Belgian e-commerce, with 60+ percent market share. Both iDEAL and Bancontact share several characteristics that enabled their success: unified banking infrastructure, early regulatory support, merchant adoption, and a checkout experience that was optimised from day one.
The Key Differences from Global Markets
The Netherlands and Belgium are small, homogeneous markets with unified banking infrastructure. iDEAL and Bancontact could succeed because all major banks in those countries were required to participate. In larger, more fragmented markets like the US or UK, unified A2A adoption is more difficult. In the US, there are thousands of regional banks. A customer in Ohio might not recognise a bank that serves California. The bank selection problem is orders of magnitude more complex.
Second, iDEAL and Bancontact benefited from very limited card penetration. In both countries, debit cards were not as dominant as in the US. This meant the cultural resistance to moving away from cards was lower. In the US, the credit card is deeply embedded in consumer behaviour. Shifting that requires overcoming two decades of habit.
What Needs to Change: The Technical Playbook
For A2A to become viable at checkout globally, several technical and experience problems need to be solved.
Embedded Bank Selection
The redirect to the bank's authentication page is the single largest source of friction. What if the bank selection and authentication happened without leaving the merchant? This is possible through embedded flows where the bank's login interface is embedded within the checkout iframe. Some payment processors have implemented this, and conversion rates improve significantly. The customer selects their bank, authenticates without redirecting, and completes the payment in context.
Embedded flows require API-level cooperation from banks and payment processors. Not all banks support embedded authentication today. But the technical path is clear. As more payment processors implement embedded flows, bank adoption will follow.
Persistent Consent Tokens and VRPs
The lack of saved credentials is the second-largest conversion killer for repeat customers. VRPs solve this by allowing the merchant to initiate payments without customer interaction after the first payment grants consent. Once a customer has authorised the merchant to collect payment, the next transaction requires no customer action.
VRP availability is the constraint. In Europe, VRPs are now mandated by regulation (PSD2 and subsequent updates). US regulators are discussing similar requirements. As regulations mandate VRP support, bank adoption will accelerate. Once VRPs are standardised and widely available, the saved credentials problem goes away.
Smart Bank Pre-Selection
Merchants can reduce friction by pre-populating bank selection based on customer data. If a returning customer has paid through A2A before, the checkout can pre-select their bank. If the customer's location or IP suggests they are in the Netherlands, iDEAL can be pre-selected. These micro-optimisations reduce the friction on repeat purchases and increase conversion.
Asynchronous Authentication
A longer-term opportunity is moving away from synchronous authentication (authenticate now, complete the payment now) to asynchronous authentication. The customer initiates the payment. The payment request is sent to the bank. The bank authenticates the customer through a push notification to the banking app. Once authenticated, the payment completes. From the checkout perspective, the payment is asynchronous. The customer does not need to wait for the authentication to complete before leaving the checkout.
Asynchronous authentication reduces perceived friction because the customer does not feel like they are waiting for the bank to authenticate them. The authentication happens in the background while they are still on the checkout page.
The Merchant Perspective: Lower Costs, Higher Abandonment
From a merchant economics perspective, A2A is deeply attractive. Interchange fees on cards are 1.5 to 3.5 percent. A2A payment costs are typically a flat fee of $0.25 to $0.50 per transaction, or a lower percentage (0.5 to 1.5 percent). For a $100 transaction, A2A costs $1 to $1.50, while cards cost $1.50 to $3.50. The savings are substantial.
But the savings are only valuable if customers actually complete purchases. If A2A checkout abandonment is 10 percent higher than cards, the merchant loses more revenue from abandonment than they save from lower fees. The merchant calculates: $1,000 in sales, card fee cost is $25, card abandonment cost is $0. Switch to A2A, fee cost is $10, but abandonment cost is $100 (lost sales). The merchant stays with cards.
This is the core dilemma A2A faces at checkout. The economics favour A2A for merchants at scale, but the conversion rate disadvantage prevents merchants from switching unless abandonment rates improve. Merchants will adopt A2A as an option (to offer variety to customers who prefer it), but they will not make it the default payment method until conversion rates match cards.
Click to Pay and Alternative Framing
Visa and Mastercard are attempting to solve the A2A checkout friction through Click to Pay, a wallet-based authentication method that reduces redirects and pre-populates payment details. Click to Pay is not technically A2A (it is still card-based), but it reduces some of the friction that makes cards slower than optimised A2A flows.
Wallet-initiated payments (Apple Pay, Google Pay) solve friction differently. The customer stores their preferred payment method in the wallet. At checkout, they tap the wallet, authenticate once (through biometrics or device PIN), and the payment completes. The friction is minimal. The payment method underneath (card, bank account, or something else) is invisible to the customer.
This is why wallet-initiated payments are capturing increasing share of mobile checkout. They are not technically A2A, but they provide the frictionless experience that A2A promised. Merchants increasingly view wallet payments as the optimal checkout flow, potentially more important than optimising A2A specifically.
The Multi-Rail Checkout of the Future
In three to five years, the optimal checkout will offer multiple payment rails simultaneously. The customer will see cards, A2A (with embedded authentication), buy-now-pay-later, and wallet payments all as options. The merchant's payment orchestration system will route the customer to the rail most likely to convert based on their location, device, purchase history, and other signals.
A2A will win in specific scenarios: recurring payments, high-value transactions where cost matters, and geographies where A2A infrastructure is mature (like the EU). Cards will win where consumer protection and universal acceptance matter. Wallets will win where frictionless experience is the priority. BNPL will win where purchase size and financing matter.
The future of A2A at checkout is not "A2A replaces cards." It is "A2A becomes a critical option for specific use cases and geographies, with conversion rates approaching parity with cards as user experience optimisations are implemented." The playbook is now clear. The constraint is implementation speed and adoption.
In your e-commerce checkout, if A2A conversion rates matched cards, would the lower payment costs make it your preferred payment method, or are there other factors beyond cost and conversion that would prevent adoption?
Key Takeaways
- A2A checkout loses because of user experience friction: Redirects, lack of saved credentials, and bank app switching create higher abandonment than cards, despite lower payment costs.
- A2A dominates in non-checkout use cases: Recurring payments, P2P transfers, B2B, and government disbursements have achieved massive A2A adoption because user experience friction is eliminated through app layers or background processing.
- Conversion rate gap is real: Cards convert 2 to 5 percent better than A2A at checkout. This gap exists due to consumer familiarity and streamlined user flows developed over two decades.
- iDEAL and Bancontact prove the playbook works: In the Netherlands and Belgium, A2A payments (iDEAL and Bancontact) achieved 60 to 70 percent e-commerce share through unified banking infrastructure, regulatory support, and optimised checkout experience.
- Embedded authentication reduces friction: Moving bank authentication inside the checkout (embedded flows) instead of redirecting away can significantly improve conversion rates for A2A payments.
- VRPs solve the repeat purchase problem: Variable Recurring Payments and persistent consent tokens allow one-click A2A checkout for repeat customers, but adoption requires regulatory mandate and bank implementation.
- Merchant economics favour A2A at scale: Lower payment processing fees make A2A attractive, but only if conversion rates do not suffer. The merchant calculus is: lower costs per transaction vs. higher abandonment cost.