Cart and Checkout Abandonment Statistics (2026)

Baymard's figure of 70.22% is calculated as the average of 50 different studies that each report an abandonment rate. The individual studies range from roughly 57.6% to 81.1%, because they measure different merchants, categories, devices and definitions of a cart.

So the number is trustworthy as an order of magnitude and unreliable as a benchmark. Two honest stores in different categories can sit 20 points apart without either having a checkout problem.

Baymard's central methodological point is that a large share of abandonment is not addressable by design at all: shoppers comparing prices, saving items for later, or reacting to a delivery cost that is simply what shipping costs. What remains is the friction introduced by the order flow itself.

That split is where optimisation budgets should be allocated. Removing avoidable friction is a compounding gain; trying to win back window shoppers is mostly discounting.

Baymard's checkout usability research, built on thousands of hours of large-scale testing, consistently points at the same order-flow problems rather than at exotic ones: costs that appear only at the final step, mandatory account creation, form fields that are longer or less forgiving than they need to be, and unclear delivery or returns information at the moment of decision.

None of these are novel. They persist because each one is owned by a different team, and none of them individually looks fatal in an analytics dashboard.

Abandonment is a ratio whose denominator you control by definition. A store that lets shoppers add to cart in one tap from a listing page will always look worse than one that requires a product page visit first.

Segment before you conclude anything: device, new versus returning, paid versus organic, and cart value band. Most of the alarming aggregate movement in an abandonment chart turns out to be a shift in traffic mix.

Published abandonment rates vary widely because the denominator varies. Cart abandonment counts sessions that added an item; checkout abandonment counts sessions that reached the first checkout step. The second number is always lower and is the one worth optimising against.

Before benchmarking your own figure, write down the exact event that starts the funnel. Comparing your checkout-step rate against someone else's cart-level rate will make a healthy store look broken.

Across surveys, unexpected shipping, tax or fee totals appear ahead of form length as a stated reason for leaving. That ordering matters for prioritisation: showing a landed total earlier is usually cheaper than rebuilding a checkout flow, and it moves the same shoppers.

Forced account creation and slow, uncertain delivery estimates sit close behind. All three are information problems rather than interface problems, which is why redesigns that only reduce field count often disappoint.

A share of abandoned checkouts return without intervention, so any recovery measurement needs a holdout group. Without one, an abandonment email programme takes credit for buyers who were coming back regardless.

The honest test is a split: identical abandoners, one group contacted, one not, measured on completed orders over the same window. It usually shows a smaller lift than vendor case studies quote, and a real one.