Last CallLast Call Pool
Research & Analysis

3 Key Findings from Testing the World's Leading Checkout Flows

September 19, 2024 3:55 PM
Long Lecture
🇬🇧 English
Bratysława 2

About

Baymard's large-scale checkout usability research found that the average checkout contains 11.2 form fields. Yet the same checkout usability testing also reveals that most sites can achieve a 20–60% reduction in the number of form fields displayed by default. In this presentation, Baymard will share three key findings from testing the world's leading checkout flows and show you how it can be optimized to decrease checkout abandonment.

Talk by

Photo of Rinaldo Ugrina, Product Manager at Baymard Institute
Rinaldo Ugrina
Product Manager at Baymard Institute

Watch the full talk

Watch this WaysConf session, then continue with related talks or explore the current programme.

Explore the 2026 agendaMore WaysConf recordings
From the recording

Talk in brief

Rinaldo Ugrina of Baymard Institute explains why checkout optimization should focus on the number and difficulty of form fields rather than simply the number of pages. Drawing on Baymard’s research as presented in 2024, he walks through a checkout that begins with 17 visible fields and reduces it to seven. The changes are intentionally practical: use one full-name field, collapse optional minority fields such as address line two, company, and coupon code, provide resilient address lookup, default billing to shipping when order data supports that choice, and delay optional account creation until after the order. The broader principle is to reduce both real work and perceived work. Open text fields attract attention and appear to be required tasks even when they are optional. A shorter-looking form is not enough; each remaining input must match how users think and provide a reliable fallback when automation fails.

Key takeaways

  1. 01

    Count tasks and fields, not only checkout steps

    The talk finds no consistent relationship between fewer pages and better checkout UX. Visible fields more closely represent the work users believe they must complete.

    Watch from 7:46
  2. 02

    Match the form to how people think about their name

    A single full-name field avoids the correction behavior observed when users enter their whole name into a first-name field.

    Watch from 12:03
  3. 03

    Collapse optional minority fields

    Address line two, company, and coupon inputs remain available but stop presenting themselves as mandatory work to every customer.

    Watch from 14:38
  4. 04

    Use address automation with error tolerance and a fallback

    Lookup should understand misspellings, rank useful nearby results, and still expose the completed fields for review or manual entry.

    Watch from 17:55
  5. 05

    Delay account creation instead of blocking checkout

    Guest checkout followed by an optional password on the confirmation page makes account creation feel like the one or two extra fields it actually requires.

    Watch from 22:40
Read edited transcript highlights

These concise notes were edited from automatic captions and checked against the talk structure. They are not a verbatim transcript.

From observed behavior to design guidance

Ugrina begins by explaining Baymard’s method as it stood at the time of the talk: moderated usability sessions on live commerce sites are analyzed for recurring behavior, converted into design guidelines, and then evaluated against major sites. That method distinguishes a pattern that repeatedly causes difficulty from a preference or isolated anecdote. The session focuses on a subset of checkout findings that teams can act on through layout and form behavior rather than a complete redesign of commerce infrastructure.

Watch from 0:01

The wrong optimization question

Teams often ask how many pages a checkout should contain. The talk argues that page count has no reliable, direct relationship with usability because a single page can still contain many demanding tasks. Form fields are a better representation of both actual and perceived effort. Reducing visible inputs therefore matters more than compressing every step into one screen. The target is not a universal page count; it is the smallest set of understandable tasks needed to complete the order.

Watch from 7:46

One field for a person’s name

Eye-tracking sessions showed people entering their complete name into the first visible name field, then moving parts of it after realizing the form expected separate first, middle, or last names. A full-name field better matches the user’s mental model and accommodates titles, middle names, prefixes, and suffixes. On a typical checkout, merging first and last name removes one visible input. It is a small change, but the session emphasizes that checkout improvement is usually the cumulative effect of several such reductions.

Watch from 12:03

Optional fields should not look like required tasks

Address line two, company, and coupon code are needed by a minority of customers, so removing them entirely would create other problems. Leaving them as open text boxes, however, makes every user stop and decide whether to act. Collapsing each field behind a clear text control keeps it discoverable without demanding the same attention. Coupon fields are especially disruptive because an empty visible box can send customers away from the checkout to search for a discount.

Watch from 14:38

Automation must remain resilient

Automatic address lookup can reduce typing and mistakes, but the implementation needs to tolerate misspellings, prioritize relevant nearby results, and fill the individual address fields for review. Manual entry remains necessary when a lookup fails. Postal-code detection is a smaller alternative that can populate city and region. In both cases the interface reduces what users must actively consider while preserving transparency and control over the final shipping address.

Watch from 17:55

Remove perceived work as well as real work

When order data shows that billing and shipping are usually identical, a checked default can hide the redundant billing fields until they are needed. Account creation follows the same principle. Starting checkout with forced registration makes all later shipping and payment work feel like part of creating an account. A prominent guest path followed by an optional password on the confirmation page presents registration as the small additional task it is. Together, the patterns in the talk reduce a representative flow from 17 visible fields to seven.

Watch from 20:29
Explore WaysConf 2026

View the agenda

Get tickets