Cross-border checkout benchmark 2026

This report summary uses illustrative placeholder data from a prototype study. It does not describe real merchants.

Executive summary

A shopper buying from another country makes more decisions than a domestic one. Will I be charged in my currency? Can I pay the way I usually do? Will there be more to pay on delivery? Each unanswered question is a reason to leave.

For this benchmark, we reviewed 120 checkouts (illustrative sample) run by mid-market and enterprise online sellers in apparel, electronics, beauty, home goods and digital services. We placed test orders from six shopper locations: Germany, Brazil, Japan, India, the United Arab Emirates and the United Kingdom. For each order, we scored the checkout on 24 criteria across five areas.

The headline finding is that most checkouts are localised at the storefront and not beyond it. Prices are translated and converted on product pages, then the payment step falls back to a single currency, a card form and no mention of duties.

Key findings

All figures are illustrative.

1. Currency often changes at the last step. 71% of checkouts showed prices in the shopper's currency on product pages. Only 44% kept that currency through to the payment step. In the rest, the shopper saw a different total at the moment of paying.

2. Local payment methods are missing where they matter most. In Brazil and India, where instant account-to-account payments lead online spending, 38% and 31% of checkouts respectively offered Pix or UPI. Card-only checkouts in those markets had the lowest completion scores in the study.

3. Duties are rarely mentioned. 23% of checkouts showed import duties and taxes before payment. 52% said nothing about them at all, leaving the shopper to discover them on delivery.

4. Translation stops at the form. Error messages for failed payments were shown in English on 64% of non-English checkouts. A shopper who does not understand why a payment failed is unlikely to try again.

5. Address forms assume one format. 58% of checkouts required a postal code field in a format that rejected valid addresses in at least one test market.

6. Speed is not the main problem. Median time from cart to confirmation was 94 seconds, and varied little between markets. Friction came from missing information, not slow pages.

What the best checkouts did

The top decile of checkouts shared a short list of habits:

• One currency from product page to receipt, with the rate locked for the session.
• The leading local method shown first in each market, above the card form.
• A landed-cost line showing duties and taxes, with a short note that nothing more was due on delivery.
• Translated error messages that said what to do next.
• Address forms that adapted to the destination country.

None of these require a redesign. Most are configuration choices in the payment and tax layer.

What's in the full report

The full 28-page report includes:

• Methodology. How we chose the sample, placed test orders and scored each criterion.
• The scoring rubric. All 24 criteria, with pass and fail examples, so you can score your own checkout.
• Market-by-market results. Findings for each of the six shopper locations, including which payment methods were offered and how often.
• Category breakdowns. How apparel, electronics, beauty, home goods and digital services compared.
• A prioritised fix list. Changes ordered by estimated effort and effect on completion.

Related reading

• Localized Checkout and Checkout Components
• Landed Cost for duty-inclusive pricing
• Payment methods catalogue