Gift cards are products. Treat them like it.

Every November, the gift card becomes the highest-margin item in the store and the least designed. The purchase flow lives behind a footer link. The "design" is a dropdown of round numbers and a stock email with a code in it. The balance checker — if one exists — demands the sixteen-digit code, the four-digit PIN, and an act of faith. And yet gift cards routinely account for a startling slice of Q4 revenue, carry near-perfect margins until redeemed, and arrive with a built-in acquisition channel: every card is an introduction to a customer who didn't choose you and now holds your money.

A gift card is not a payment instrument. It's a product with three users — the buyer, the recipient, and your support team at 11pm on 24 December — and most stores ship it with a fraction of the care they'd give a candle. Here's how we design them as first-class products, learned from a few holiday seasons watching what actually breaks.

The buyer is shopping for a feeling, not a number

The person buying a gift card is usually buying their way out of uncertainty: they don't know the size, the roast, the shade, and a card lets them give generosity without guessing. The purchase flow should honour that emotional job.

Sell the occasion, then the amount. "For the one who drinks too much coffee" with cards styled to occasions outperforms a bare grid of $25/$50/$100. The amounts matter less than the story the buyer tells themselves. This is the same gifting psychology we designed around in designing for the gift buyer — the buyer is purchasing confidence that the gift will land.
Suggest amounts against real baskets. "Covers a month of beans" or "covers dinner for two with a little left for dessert" anchors the number in a use, not a wallet. Round-number ladders with no context make people pick the middle and feel vaguely cheap about it.
The message field is the product. A generous, free text message — with a live preview of what the recipient actually receives — is the difference between a gift and a bank transfer. Show the recipient view on the purchase page. Nobody has ever bought a card faster because of a dropdown; they buy faster when they can see the moment of delivery.
Delivery scheduling, treated as sacred. Choose the date, choose the hour where the platform allows it. A birthday card that arrives at 11pm the night before is a small tragedy with a ticket attached. Confirmation should state the scheduled delivery in plain words: "Arrives by email on the morning of 14 March." If you can't guarantee timing, say what you can guarantee.

One honest constraint: physical cards. If you sell them, the shipping promise needs the same candour as the rest of your store — cut-off dates stated early, on the purchase page, not discovered at checkout on 21 December.

The recipient is a first-time customer holding your money

The recipient experience decides whether the card becomes a customer or a support ticket. The design brief is one line: someone who didn't choose you must be able to spend this without friction or humiliation.

Balance check without an account wall. One input for the code, one button, one number. No login, no PIN unless regulation truly requires it, no "create an account to view your balance" — which is a hostage note wearing a product hat. We've audited stores where checking a balance required three screens; each screen is a promise the brand doesn't want you to redeem.
Partial redemption as the default assumption. Stored value rarely matches basket value. The redemption UI must handle "card covers $50, basket is $73" without drama: apply the card, show the remainder due, keep both numbers visible. And the reverse — basket is $31 — should end with the remaining balance stated on the confirmation screen and emailed, not implied. A balance that vanishes into a database field is value the customer will assume you stole.
The code must survive the real world. Codes get forwarded, screenshotted, printed, and pasted with spaces. Accept pasted codes with whitespace and dashes stripped. Autofocus the input. If your code format includes easily confused characters (O/0, I/1), your code generator should exclude them — a decision made once, at issuance, that saves ten thousand support tickets.
The recipient email itself is a landing page. It will be opened on a phone, possibly months late, by someone mid-thought. Lead with the amount and the sender's message, one button ("Choose something lovely"), and a secondary link to check balance later. Any upsell above the amount is noise in the one email guaranteed to be opened with goodwill.

Expiry: honesty in every jurisdiction

Expiry rules differ — in Australia, gift cards sold since late 2019 must carry a minimum three-year expiry with the date disclosed; other markets have their own rules and their own ideas about fees. Two design principles travel everywhere:

Disclose the expiry at purchase, on the card, and in the email. Not in terms linked from a footer. "Valid until 14 March 2029" in the receipt email costs you nothing and buys staggering goodwill. Brands hide expiry dates because they quietly benefit from breakage — value purchased and never redeemed. Breakage income earned through concealment is a reputational debt with compounding interest.
Design the expiry horizon into the product. If cards expire, the recipient email and balance page should carry the date, and a gentle reminder flow at the nine-months-left mark is a service, not a spam. A brand that reminds you to spend its money is a brand you trust with money.

Where the law is stricter than your defaults — no expiry, no dormancy fees — that's not a constraint on the design, it is the design. Build the rule engine per market and stop shipping one global card into nine jurisdictions.

Fraud friction without punishing the honest

Stored value attracts abuse: bulk purchases on stolen cards, code-guessing, rapid resale. The honest store's dilemma is that every defence taxes the 99%.

The measures we've seen work without wrecking the experience: velocity limits on code-guessing at the balance endpoint (rate-limit silently, never reveal whether a guess was close); modest purchase caps per buyer per day with a human route for exceptions; and delaying digital delivery a few minutes for first-time buyers at large amounts, with the delay stated plainly ("security check — arrives within 15 minutes"). What doesn't work: making legitimate buyers verify their identity to buy a $50 card. The fraud team and the conversion team need to sit in the same room, with the conversion team holding veto over anything that touches the gift-buying flow in December.

The service-recovery superpower

Here's the underused one. A gift card issued by a human, unprompted, is the most graceful apology in commerce. Shipment lost in transit, delivery a week late for a birthday, a product that missed the mark — the resolution menu is usually refund or replacement, and both are mechanical. Replacing a ruined birthday present with its exact replica is logistics; replacing it with a sincere note and a card "for the next attempt, on us" is a relationship.

This only works if issuing a card is a two-minute support action with a personal message field — which is exactly what building gift cards as a product gets you. The same infrastructure, pointed at recovery. We wrote about the wider pattern in returns as a retention channel: the moment something goes wrong is the moment loyalty is actually priced.

Build checklist

For anyone scoping this properly — which, when we take on e-commerce engagements, means a dedicated slice of the roadmap rather than a plugin checkbox:

Purchase flow: occasion framing, use-anchored amounts, live recipient preview, delivery scheduling with plain-language confirmation.
Recipient flow: one-input balance check, no account walls, partial redemption both directions, remaining balance stated and emailed.
Compliance: per-market expiry rules, disclosed dates everywhere the value is shown, reminder flow on the horizon.
Operations: rate-limited balance endpoint, sane purchase caps, a manual issue-and-message tool for support.
Measurement: redemption rate by cohort, time-to-first-redemption, recipient-to-customer conversion, breakage — reported honestly, including what you earned from value never spent.

Key takeaways

A gift card has three users: buyer, recipient, and your support team. Design for all three or support absorbs the cost in December.
Buyers are purchasing confidence. Sell the occasion, anchor the amount to a real basket, and make the message field the hero.
Recipients didn't choose you. Balance check without login, partial redemption both ways, remaining balance always stated.
Expiry disclosure is a trust decision hiding inside a legal requirement. State the date everywhere; remind before it bites.
A manually issued card with a personal note is the best service-recovery instrument in commerce — if issuing one takes two minutes.

FAQ

Digital, physical, or both? Digital first, always — the margins, delivery guarantees, and data are better. Physical cards earn their place when your customers shop in person (grocers, florists, hospitality venues) or when the unboxing is the gift. If you do physical, the shipping promise discipline matters more than the card stock.

Should recipients be able to split a card across orders? Yes — stored value should behave like a wallet, not a coupon. One-shot redemption codes push people to pad baskets they can't afford or abandon value they can't retrieve. If your platform makes partial redemption hard, that's a platform problem worth escalating, not a design constraint to accept.

Isn't breakage just... good revenue? It's reported as such in plenty of board decks, and it's real. But breakage earned through concealment — hidden expiries, humiliating balance checks, redemption friction — is borrowed against brand trust. The healthy version is breakage that happens despite a genuinely easy redemption experience. Design for redemption; let breakage be an outcome, not a strategy.

How do gift cards interact with loyalty programs? Purchase earns loyalty on the buyer; redemption is just payment, not a second earn event — double-dipping both confuses the economics and invites gaming. The subtler play is recognising the redeemer as a new customer and treating their first post-card purchase as the start of a relationship, which is a lifecycle problem — the kind we map in lifecycle email architecture.