Payments and trust at the AU/NZ checkout

An Australian shopper reaches your payment step carrying a quiet checklist: Is my preferred way to pay here? Will I get slugged a surcharge for it? Does this site look like it will still exist if I need a refund? Australian e-commerce has its own payment weather — BNPL adoption is among the highest in the world, card surcharging is legal and common, and shoppers have been burned by enough lookalike scam stores to be genuinely watchful. A checkout transplanted from a US playbook leaks money here.

This is how we approach payments on AU/NZ builds, after shipping them for a skincare brand with nothing to hide, a providore shipping cold goods nationally, and a florist racing the same-day clock. None of it is exotic. All of it is deliberate.

The method mix is a research question, not a default

Before we touch layouts, we pull the store's payment-method report and cross-reference it against traffic by device and audience segment. The numbers usually hold three surprises:

Wallets punch above their weight on mobile. Apple Pay and Google Pay routinely take a third or more of completed mobile checkouts once they're actually visible — which they often aren't. If your express buttons sit below the card form, you're measuring demand for a door hidden behind a cabinet. We covered placement in the checkout friction audit; it applies here doubled.
BNPL is a category question, not a demographics question. Afterpay and Zip index highest in apparel, beauty and gifting — impulse-priced purchases of $60–$300. On a $900 providore hamper, BNPL share collapses; on a $79 candle trio, it can carry a quarter of orders. Merchants who assume "our customers don't use Afterpay" are often wrong in both directions.
PayPal still matters more than teams think. Especially for first-time buyers of a brand they found through a marketplace or social ad. PayPal's pitch isn't speed — it's "the merchant never sees my card." That is a trust product, and for new brands it converts like one.

The mix shifts with AOV, category and acquisition channel, so the only universal rule is: measure per-method conversion before you arrange the buttons.

Ordering is a design decision with revenue in it

Payment-method order is ranking UI, and all the rules of ranking apply: position one gets a premium, and whatever you put first tells customers what you want them to do. Our default order, and the reasoning:

Express wallets (Apple Pay / Google Pay / PayPal) at the top of the payment step, and again in the cart. For a large share of phone buyers the wallet is the checkout — address, card and confirmation in one authenticated sheet. Everything beneath is a fallback path.
Card next, because it's the universal fallback and carries the least explanation cost. Card inputs should detect the brand from the BIN, format the number to the card's spacing, and never, ever clear themselves when another field errors.
BNPL (Afterpay first in AU, Zip second) after card, with the instalment maths shown. "4 payments of $24.75" is the persuasive copy — not the logo. Afterpay's own research keeps finding that showing the split amount at PDP lifts conversion on considered purchases; showing it only at checkout is money left at the door.
Bank transfer (PayID/Osko) only where it genuinely fits. For high-AOV B2B-ish orders — commercial catering kit, trade supplies — a PayID option with a clear "orders ship after payment clears" note can convert buyers who don't want $4,000 on a card. For a $60 t-shirt, it's friction dressed as choice.

One caveat we give every client: more than four visible methods is not inclusivity, it's a decision problem. Every extra option taxes the fast paths. Choose for your audience, not for the completeness of the payment industry's catalogue.

Surcharges: the honesty tax

Australia is one of the few markets where card surcharging is legal — capped by the RBA at the merchant's actual cost of acceptance. It's also one of the fastest ways to poison a checkout. The pattern we push:

Amex and international cards cost more. If you're going to surcharge, say so before the card field is filled — "a 1.5% fee applies to Amex" as inline hint text — never as a surprise line on the review step. Surprise fees at the final step are among the most-cited abandonment triggers in every study we've seen, and our own session recordings back it: the cursor stalls, the tab closes.
The better move is usually to absorb it and adjust pricing. A store that quietly builds 40 cents into its pricing converts better than one that itemises its internal costs at the customer. Nobody has ever felt warm toward a business because the payment processing was transparently billed.
NZ note: cross-border shoppers are common in both directions. Show the currency early, in the header and on the PDP price, and don't let someone discover at payment that $129 was NZD. Currency surprise reads as a trick even when it's a mistake.

Trust signals that don't look like clip-art

The classic move — a footer strip of Norton, McAfee and "100% Secure" badges from 2011 — not only doesn't build trust, it actively marks the site as dated. Shoppers under forty have never consciously read one. What actually works:

Payment-method marks, rendered small and grey, at the payment step. Not "trust badges" — just the quiet confirmation that Visa, Mastercard, Afterpay and PayPal are real options here. We render them as inline SVGs, full monochrome, sized to sit inside a hairline rule. They function as a UI label for the methods, not a security claim.
The returns and shipping story, one line, near the pay button. "Free 30-day returns. Ships from Marrickville." Locality is a trust signal: a stated Australian warehouse answers the unasked question "how long is this taking and can I send it back." For GLADE's skincare storefront, surfacing the dispatch suburb and the returns policy beside the payment button measurably cut support tickets asking exactly those two questions.
Real policy pages, linked in footer, written by humans. A refund policy that reads like a person wrote it — conditions, timeframes, Australian Consumer Law acknowledgement — does more than any badge. On that note: in Australia you cannot contract out of ACL guarantees. A policy that says "no refunds on sale items" is not just untrustworthy, it's unlawful. Get the legal page right and link to it.
Order confirmation as a trust asset. The email and the confirmation screen are where first-time buyers decide whether they made a mistake. Instant, specific confirmation — what was ordered, when it ships, how to reach a human — pays back in repeat purchase rate. We treat it as part of checkout design, not an afterthought.

Per-method analytics, or you're guessing

Payment reporting should answer four questions per method: selection rate (who picked it), completion rate (of those, who finished), decline rate by reason, and share of revenue. The failure modes we've found with this lens:

A store whose Afterpay completion rate ran 15 points below card — traced to the Afterpay redirect killing the session on older Android WebViews. Nobody had ever completed a purchase on the actual devices after the actual redirect.
Declines on card spiking for one issuer — a 3DS challenge that wasn't rendering inside the embedded payment frame on iOS. The fix was a one-line CSP change; the diagnosis took a session recording.
BNPL selection high but completion low on a high-AOV store — the instalment limit was below the median basket. Merchandising fix, not a checkout fix.

Tag every checkout event with the method selected, and watch declines separately from abandonments — a decline is not a lost customer, it's a customer your payments stack turned away. Then run the full flow yourself, monthly, on a real phone, on 4G, with each method. Merciless dogfooding catches what dashboards aggregate away.

Key takeaways

Payment-method order is revenue-bearing UI: wallets on top, card as fallback, BNPL with the split maths shown, everything else only if it earns its slot.
Measure conversion per method per device; the mix you assume is rarely the mix you have.
Surcharges are legal in AU and still usually a mistake — if you must apply them, disclose before the card field, never at review.
Trust comes from specific, local, human information — dispatch suburb, real returns terms, instant confirmation — not from clip-art security badges.
Declines are stack failures, not customer failures. Instrument them separately.

FAQ

Should we offer Afterpay if our margins are thin? Run the arithmetic on incrementality, not ideology. BNPL fees run roughly 3–6% versus ~1.5% for cards, so the question is whether offering it wins orders you'd otherwise lose. Test it for a quarter, measure per-method completion and new-customer share, then decide. For gift-heavy or apparel stores it usually pays; for commodity restock purchases, often not.

How many payment methods is too many? Past four visible options, each addition slows the fast paths. If a method takes under 2% of orders after you've given it fair placement for a quarter, retire it.

Do we need 3-D Secure on everything? Let the payment provider's risk engine decide per transaction rather than hard-forcing it — intelligent 3DS keeps fraud protection while sparing low-risk orders the challenge step. Hard-forcing it costs you real conversions on mobile.

Is PayID worth supporting for a normal consumer store? Usually not as a checkout default — the cleared-payment delay breaks instant fulfilment. It's worth it for high-AOV, made-to-order or B2B-flavoured stores where a bank transfer genuinely suits the buyer.

Where should the split-payment maths live? On the PDP under the price and in the cart line — that's where it changes the buy/no-buy decision. Repeating it at payment is fine; only showing it at payment is too late.

If you're weighing a checkout rebuild, our e-commerce practice starts with exactly this audit — or see how it played out in the Tallow & Co. case study.