How to Reduce Cart Abandonment: Checkout UX Fixes Ranked by Why Shoppers Leave

Baymard Institute, which publishes large-scale e-commerce usability research, keeps a running average of published cart abandonment studies. Its October 2026 update puts the average at 70.22%, calculated as the unweighted average of 50 studies published between 2006 and 2025, with individual results ranging from 55% to 84%. In plain terms, roughly seven of every ten carts never become orders.

That number is a benchmark, not a target. A large share of abandonment is not a checkout failure at all: Baymard reports that 42% of US online shoppers have abandoned a cart because they were just browsing or not ready to buy. People use carts as wish lists and price calculators. The useful question is not how to reach zero, but how much of your abandonment is caused by your own checkout.

Excluding the just-browsing group, Baymard's survey of reasons for abandoning during checkout produces a ranking that doubles as a priority list. Baymard does not publish the survey's sample size on the public page, so treat the percentages as directional rather than exact.

Cost surprises lead the list twice: 40% cite extra costs and another 12% could not see the total up front. The fix is to move cost information earlier. Show a shipping estimate or free-shipping threshold on product and cart pages, let shoppers estimate shipping from a postal code before they enter checkout, and never introduce a mandatory fee on the last step.

For Canadian stores this is also a legal issue. The Competition Bureau describes drip pricing as offering a low price and then adding mandatory fees so the advertised price is not attainable. Its guidance to businesses is never to promote an unattainable price, unless the extra charge is imposed by an act of Parliament or a legislature, such as sales tax. A service fee, booking fee or handling fee that every buyer must pay belongs in the price you show first.

Forced account creation (18%) and long or complicated checkouts (17%) are design decisions, so they are the cheapest to change. Baymard's benchmark found that the average US checkout shows 23.48 form elements by default, or 14.88 counting only form fields, while an ideal flow needs about 12 to 14 elements, or 7 to 8 fields. Google's web.dev guidance on payment and address forms gives the implementation details.

The next cluster is about confidence: distrust of the site (19%), errors or crashes (17%), declined cards (10%) and too few payment methods (9%). These need engineering work rather than copy changes. Monitor checkout errors as a release metric, test the full flow on real mobile devices after every theme or app change, and offer the wallets your customers already use so they do not have to type a card number at all.

Delivery speed (20%) and returns policy (13%) are operations decisions, but checkout can still state them clearly. Show an estimated delivery date next to each shipping option and link a short returns summary near the payment button instead of hiding it in the footer.

If you sell to consumers in Japan, checkout design is regulated directly. Article 12-6 of the Act on Specified Commercial Transactions requires the screen where clicking the order button completes the purchase to display six items, and prohibits displays that mislead the buyer. The Consumer Affairs Agency's guideline explains each one, and a buyer who places an order after being misled by a display that breaks this rule can cancel it.

Once the checkout itself is fixed, recovery messages catch shoppers who were interrupted. They depend on capturing an email address early. Shopify, for example, treats a checkout as abandoned when it stays incomplete for more than ten minutes after the customer has entered their email, and it cannot send a recovery email if the customer gave only a phone number. Its Messaging app lets you choose the delay before an automated reminder is sent.

Ask for the email address as the first checkout field, keep the reminder short with a direct link back to the cart, and confirm your consent basis before sending. In Canada, reminder emails can be commercial electronic messages under Canada's Anti-Spam Legislation, so confirm with counsel how your consent wording covers them.

Measure abandonment by step rather than as one number: cart to checkout start, checkout start to shipping, shipping to payment, payment to order. A drop at the shipping step usually points to cost or delivery surprises, and a drop at payment points to errors, declines or missing methods. Change one thing at a time, compare the same weeks of demand where you can, and keep the abandoned checkout report as a list of real sessions to watch and test.

Baymard estimates that large e-commerce sites can gain a 35.26% increase in conversion rate through better checkout design alone. Your own result will differ, which is why step-level measurement matters more than any industry average.

Before changing anything, run a short audit of the live checkout. It usually shows which of the fixes above applies to your store, and gives you a baseline to compare against.