Refund Method: Original Payment or Alternative?

Every returns portal makes a quiet decision about where the money goes, and most make it badly. The default in consumer law across much of the world is that a refund returns through the same means the customer paid. The default in most returns products is whatever was easiest to build, which is often store credit or a manual bank transfer.

The gap between those two defaults is where compliance risk, support volume and customer trust all leak out at once.

The Legal Default Is the Original Method

Across the EU and many comparable regimes, a trader must reimburse using the same means of payment the consumer used, unless the consumer expressly agrees otherwise and incurs no cost as a result. The consumer-facing summary of these rights is published by the European Commission in its guarantees and returns guidance.

Two words carry the weight. *Expressly* means a real choice, not a pre-ticked box or a path of least resistance. *No cost* means the alternative cannot be worth less, expire sooner, or carry conditions the original refund would not.

When the Original Method Genuinely Fails

Cards expire, get cancelled after fraud, and change when a customer switches bank. Wallets close. Buy-now-pay-later plans complicate the picture further because the money did not come from the customer directly. These are real cases and they need a designed path, not an improvised one.

Most processors will still accept a refund against an expired card if the underlying account exists, because the issuer routes it to the replacement. The failure case is a closed account, which usually surfaces as a rejected refund days later rather than immediately. Your workflow needs to detect that rejection and re-open the case rather than marking the refund complete and moving on. For instalment-funded orders, the sequencing is different again and is covered in BNPL returns and refund timing.

Collecting Bank Details Is Not Free

The obvious fallback is to ask for an IBAN and transfer the money. It works, and it introduces two problems merchants routinely underestimate: you are now storing bank details, which raises your data protection obligations, and you have created a channel where an attacker who compromises a returns request can redirect money to their own account.

If you collect bank details, treat the change of refund destination as a high-assurance action requiring verification, exactly as covered in returns portal authentication. Never allow a bank detail to be supplied and used in the same unverified session.

• Default to the original payment method with no pre-selected alternative.
• Present store credit as an equal-weight choice, never as the easy path.
• Never let credit expire sooner or be worth less than the refund it replaced.
• Verify identity before accepting a new refund destination.
• Detect rejected refunds and re-open the case automatically.

A choice the customer had to look for is not a choice. It is a default with a fig leaf.

Offering Credit Without Steering

Store credit is commercially attractive: it keeps the money in the business, converts at high rates and often exceeds the original order value when spent. None of that is a problem. The problem is the design patterns used to increase take-up: pre-selection, visual prominence, a bonus that is only disclosed after the customer chooses, or a refund path buried one level deeper.

A bonus on credit is legitimate and often effective, provided the underlying refund remains equally available and equally visible. Offering ten percent extra in credit is a genuine choice; making the refund button grey and small is not. The economics of doing this well are set out in store credit bonus economics.

Exchange Sits Beside, Not Instead

The same principle applies to exchange. Making exchange the most prominent option is good product design, because most customers who wanted the item still want the item in a different size. Making refund hard to find is a different thing, and the distinction is whether a customer who knows what they want can get it in one obvious step. The exchange-first playbook argues for prominence, not for obstruction.

Timing Interacts With Method

The statutory refund deadline runs regardless of which method the customer chooses, so a slow manual bank transfer process can put you outside the window even when the customer picked it. Automate the alternative paths or accept that they will be your late refunds.

Store credit has the opposite property: it can be issued instantly, which is why it is genuinely attractive to customers who want their money back today rather than in five working days. That speed advantage is worth stating explicitly in the interface, because it is a real benefit rather than a persuasion technique, and it fits the case made in instant credit and re-spend.