Checkout abandonment rate: benchmarks and how to reduce it

Checkout abandonment rate is the share of people who begin your checkout flow and leave without paying. It's narrower than cart abandonment, which counts everyone who added an item and wandered off — a group full of browsers, price-checkers and tab hoarders who never intended to buy today. Someone who starts checkout has signaled intent. Losing them is the expensive kind of loss, because everything upstream — the ad spend, the landing page, the product page — already worked.

The two rates get conflated constantly, and it matters: cart abandonment averages around 70% and is partly noise, while checkout abandonment sits lower and is mostly signal. A high cart rate says people like to browse. A high checkout rate says the checkout is broken.

Industry studies put average cart abandonment around 70% — Baymard Institute's long-running average is just over that — and typical checkout abandonment in the 50–70% band, though anything under 50% is genuinely strong. Mobile runs worse than desktop, often by 10–15 points, which is uncomfortable given mobile is most sites' majority traffic. Travel and luxury goods run higher than average; groceries and repeat-purchase categories run lower.

Benchmarks deserve a caveat, though: they swing on where 'checkout started' is measured. Counting from the cart page produces a scarier number than counting from the first checkout form field. Pick one definition, keep it, and treat your own trend as the real benchmark — a rate falling from 68% to 55% is worth more than any industry table.

Survey after survey finds the same culprits, in roughly this order — note how many are fixable in an afternoon:

• Surprise costs — shipping, taxes and fees appearing at the last step
• Forced account creation before paying
• Long or confusing forms — every field is a toll booth
• Missing payment methods — no PayPal, no Apple Pay, card-only
• Trust wobbles — sketchy design, no security cues, unfamiliar brand
• Comparison shopping and plain distraction — some loss is just humans

The overall rate says you have a problem; it never says where. That takes a funnel over the checkout steps — cart → details → payment → done — showing the drop at each stage. A 40% cliff at the payment step points somewhere very different (payment methods, surprise total) than a slow bleed through the address form (form length, mobile keyboards).

Numbers locate the leak; they rarely explain it. This is where session recordings earn their keep: watch five replays of people quitting at the same step and the cause is usually embarrassing and obvious — a coupon field that sends everyone off hunting for codes, a validation error that clears the form, a shipping dropdown that doesn't work on iOS. In Bigdelta the funnel and the recordings sit on the same data, so 'who dropped at payment' is a segment you can watch, not a hypothesis.

Fix the step the funnel points at, not the checklist the internet points at. That said, the fixes with the best hit rate: show the full cost as early as possible — surprise shipping is the single most-cited abandonment reason; offer guest checkout and ask for the account after the receipt; cut every form field that isn't strictly needed to fulfill the order; add the wallet buttons (Apple Pay, Google Pay, PayPal), which collapse the whole form into one tap; and make the error states humane, because a form that wipes itself on a card typo deserves its abandonment rate.

Then follow up on the losses that still happen: an abandoned-checkout email within a few hours recovers a real single-digit percentage of completions, which at checkout-level intent is meaningful revenue. Recovery is a patch, though, not a cure — a checkout that needs heavy recovery emails is a checkout with a fixable leak.

Instrument the checkout steps before touching anything — one week of funnel data turns the redesign debate into a one-line diagnosis. Fix the biggest cliff, watch the rate for two weeks, repeat. Chasing the last 30% is futile — a chunk of abandonment is humans being humans — but the gap between a 70% rate and a 50% one is usually two or three concrete fixes, and at checkout intent levels, each point recovered is revenue that was already standing at the till.