Improve the checkout.

Read the list as a product manager would. Most of these reasons are design decisions, not payment problems. Only one, the declined card, is about the payment itself, and lessons 3 and 12 cover how to recover from those. The rest are about price, trust, effort, and speed, which a team can change.

Walk the checkout, part by part

Tap each part of the checkout below. Switch between a returning shopper, whose details are saved, and a new shopper, who sees the form: most checkout trouble happens to the second. The panel says what goes wrong in each part, what to do about it, and how strong the evidence is. Research means a usability study or survey backs it. Practice means it is common practice that you should still test on your own checkout.

What not to do

Some tactics lift a number this week and cost trust for good. Fake urgency (a countdown that resets, “only 2 left” when it is not true) teaches shoppers not to believe you. Pre-checked add-ons, such as protection plans or donations ticked by default, put a charge in the cart the shopper did not choose. Hiding the real price until the last step is the top reason people leave. A good checkout is honest and fast, not clever.

Design for the phone first

Adobe reports that 56.4% of US online holiday transactions in 2025 happened on a smartphone, up from 54.5% the year before. Test every change on a phone first. Make tap targets big enough to hit with a thumb: WCAG 2.2 sets a 24 by 24 pixel minimum at level AA, and Apple’s guidelines recommend 44 by 44 points. Use the right keyboard for each field (numbers for ZIP and card, email for email) and autocomplete attributes, so the phone can fill the form for the shopper.

Accessibility is part of conversion. The US Department of Justice says the ADA applies to the websites of businesses open to the public, and points to WCAG as the guide. For checkout, the criteria that matter most are labeled fields with their purpose declared (1.3.5), errors that are identified and come with a suggestion (3.3.1 and 3.3.3), not making people re-enter what they already gave (3.3.7), and targets large enough to tap (2.5.8). Test the whole flow with a keyboard and a screen reader.

Measure before you fix

You cannot improve what you have not measured. Count how many shoppers reach each step and how many move on, so you can see exactly where they leave. Split it by device and by payment method: a checkout can be fine on desktop and broken on a phone, or fine for cards and broken for one wallet. Then change one thing at a time, and test it against the old version (an A/B test) before you roll it out to everyone.

Try it below with made-up numbers. Move a step’s completion rate and watch how many orders come out the end. Small gains early in the funnel multiply through every later step.

Keep the size of the prize honest. Baymard estimates that the average large ecommerce site could gain a 35.26% increase in conversion rate through better checkout design alone. That is an estimate of what is possible, from usability testing, not a result any one site is promised. Your own funnel and your own tests tell you what is true for your checkout.

Sources for this lesson: Baymard Institute (cart abandonment statistics, last updated September 22, 2025; checkout form fields, June 2024; checkout UX benchmark, 2025; inline validation and error message research), Adobe (2025 US holiday shopping season), US Department of Justice ADA web guidance (2022), W3C WCAG 2.2, Apple Human Interface Guidelines, and Stripe’s checkout optimization guides.