A mistyped card number costs more orders than the typing itself, and a holiday peak raises that cost
In November 2023, a few weeks before the holiday season, Retail Customer Experience published a piece by Matt Miller, vice president of product at Paze, a digital payment method offered by banks and credit unions. His case is a familiar one. Holiday shoppers are short on time, often on a phone, and a checkout that makes them retype what they already have stored somewhere else loses them. The remedy he points to is the category his own company competes in: standardised digital wallets that fill in card, billing and shipping details. The argument holds in outline. The evidence offered for it is thinner than it looks, and the part that does hold up leads to a piece of the payment step most shops have not opened in years: the card form itself.
What merchants believe is not where shoppers stop
The headline figures in the article describe opinions held by merchants. It states that 55% of merchants believe shoppers abandon a cart when their preferred payment method is not offered, and that 74% believe shoppers abandon when they have to type a 16-digit card number by hand. The survey behind those figures is not named, and neither are its sample or its year. A third figure, on how many merchants see trust and security as critical, carries a footnote marker with no footnote on the page. A belief shared by three quarters of merchants may well be correct. It still tells you what your peers expect, not what happens in your own payment step.
Research that watches shoppers is more precise about where manual entry goes wrong. In Baymard Institute’s moderated usability tests, participants often double-checked the number they had typed before submitting, and tended to type it in the same groups of digits printed on the card. One wrong digit triggers a validation error. On more than a third of the sites Baymard tested, an error in the form wiped the card details the shopper had already entered. That is where manual entry really costs orders: not in the typing, but in being asked to start again after a mistake the form could have caught.
A card form that forgives a typo
Most of the repairs are small, and none of them requires a new payment method. Show the number in groups of four while the shopper types, the way it is printed on the card, and accept spaces if the shopper types them. Baymard recommends both. It also advises against a fixed length limit on the field, because card numbers differ in length; Maestro numbers, its example, run from 12 to 19 digits. When something is wrong, keep everything else the shopper entered, point to the one field that needs attention and say what is expected there.
Then let the phone do the typing. Give the name on the card, the card number, the expiry date and the security code each the autocomplete value meant for that field, so the browser or the phone’s password manager can fill in a stored card in a couple of taps. A form that leaves them out makes the browser guess which field is which, and when it guesses wrong the shopper is back at the keyboard. If your card fields are loaded inside a frame from a third party, check autofill on a real phone, because that is where the holiday shopper meets it.
Place the wallet buttons above the card form, not under it. A shopper with a card already in Apple Pay or Google Pay should see the route that skips the empty fields before they reach the fields.
Mastercard is working to remove the form, and it is not finished
The article is almost three years old, and one of the card schemes has since moved in the direction it argued for. Mastercard has set itself the goal of phasing out manual card entry online across Europe by 2030, built on tokenisation, Click to Pay and payment passkeys. In a Click to Pay checkout the shopper no longer types a card number and the merchant no longer receives one. In February 2026 Mastercard said it was halfway towards that goal.
A goal set for 2030 also means the form is still in use this season. A first-time buyer whose card is not enrolled in any of these services has no other way to pay by card on your site. The form is the fallback behind every card route, and the fallback is what the shopper sees when the faster route is not available on their device or does not recognise them.
In the Netherlands the card form serves a minority, which makes it easy to overlook
The Dutch market changes the weighting without removing the problem. The Dutch Payments Association reports that iDEAL accounts for 71% of all online purchases, credit cards for 8% and direct debits for 5%. iDEAL is a payment by bank transfer, so a shopper who chooses it never types card details on your site. A form that serves a minority of orders tends to be built once at launch and not tested again on a current phone.
The share is small, but a peak raises the absolute numbers, and the shoppers on that 8% route are by definition the ones who did not choose iDEAL. They came to your payment step with a card in hand, and the form is the only thing between them and the order.
Before the next peak, buy something from your own shop on a phone, as a guest, by card. Do it twice. The first time, let the phone fill in the card. The second time, type the number by hand and make one deliberate mistake. Note whether the digits group as you type, whether autofill completed every card field, whether a wallet button was visible before the form, and what was still on screen after the error message. Every field you had to type twice is one to fix before the shoppers who will not try a second time arrive.
Sources: retailcustomerexperience.com, baymard.com, card number field, mastercard.com and betaalvereniging.nl
