Merging checkout screens rarely shortens a checkout, because the shopper still has to face every field
When a checkout feels slow, the usual reflex is to cut screens: put delivery and payment on one page, switch to a one-page layout, show a progress bar with fewer dots. The result looks shorter on a roadmap. For the shopper it often is not. Every field that was spread over four screens is still there, now stacked in a single column, and on a phone the shopper scrolls through all of it with the keyboard open. The number of clicks on a continue button goes down. The amount of reading, deciding and typing stays the same.
Baymard Institute’s research supports measuring the checkout in fields rather than screens. In an article published in June 2024, research lead Edward Scott reports that in Baymard’s usability testing the number of screens mattered much less to shoppers than the work each screen asked of them. The figures point the same way. A new customer meets 5.1 steps on average, counted from the cart up to and including the order review, a number that has hardly moved since Baymard began tracking it in 2012, and Baymard’s view is that usability does start to suffer once a flow reaches eight steps or more. Fields are a different matter: 11.3 on average in 2024, down from 12.7 in 2019, against an estimate that most sites need only 8. The screens of a typical checkout are mostly fine. The excess is in the fields.
The length of your checkout is a number you can take in ten minutes
Open your checkout as a new guest, on a phone, and list every input and every choice that is visible on the default route, step by step, optional ones included. That total is the length of your checkout. It is a more honest figure than the step count, because it is what the shopper actually has to deal with.
Count the fields that are visible but usually left empty as well. A field on screen has to be read and judged as relevant or not, even when the answer is to skip it. The article shows a gaze plot from an earlier round of testing on an outdoor retailer’s payment step: across 32 participants, a large share of attention went to membership sections that had nothing to do with the purchase. A newsletter block or a gift message box on the default route can take the same toll. On a phone every extra field weighs more, because typing and correcting are slower there, a point Baymard makes as well.
Shoppers notice the total even if they never count it. In Baymard’s survey of US online shoppers, 17% say they have abandoned an order because the checkout was too long or complicated. What they describe is an impression, and that impression is formed by everything the form puts in front of them, not by the number of screens.
Every field in the checkout was requested by someone
Checkout fields are rarely added in one go. They arrive one at a time, each with a reason someone in the company could defend: a phone number for the carrier, a company name for invoices, a separate first name for personalised emails, a date of birth for a loyalty programme. Nobody is asked to remove any of them, so the total only grows. That is why a field count needs to be reviewed field by field, with the question of who uses the data and whether every shopper has to see the field for that purpose.
The name is a good test case. Baymard found that 89% of the sites in its benchmark split the name into more than one field, while in testing 42% of participants typed their full name into the first-name field at least once. With a single name field, only 4% hesitated briefly. A second address line shown by default is another: it stopped 30% of participants in testing, and 75% of sites still show it. Company name and extra phone fields belong in the same review.
Most of these fields do not have to disappear. They can move out of view. A second address line can open from a link, the company name can sit behind a link of its own, and an account can be offered on the confirmation page instead of before the order. Billing details can appear only when the shopper says they differ from the delivery address. For the billing address Baymard makes an exception for shops where the order data shows that many customers use different addresses, such as gifting and business-to-business sites. In each case the department that wanted the data still gets it from the shoppers who have it to give, and everyone else no longer has to read the field.
For the address itself, Baymard recommends going further, with a fully automatic lookup that fills in the complete address. If you sell in the Netherlands, a lookup on postcode and house number matches the way shoppers already give their address.
Give the field count an owner
Treat the number of visible fields on the default guest route as a figure that is measured and reported, separately for mobile and desktop, in the same way page speed is. Baymard’s 8 fields is a practical reference point for most shops. Any request for a new field should then state what it adds and what it costs against that figure, and someone has to be able to say no.
Start this week. Count the visible fields in your own guest checkout on a phone and write the number down. Mark every field that the order does not strictly need and name the team that asked for it. Merge the name fields, put the second address line behind a link and add an address lookup, then count again. That second number is the one to defend from now on.
Sources: baymard.com and baymard.com
