Cart and checkout: where 70% of ready buyers disappear

Short answer: the people standing in your checkout have already chosen a product and taken out a card — and roughly 70% of them leave. This is the densest part of the funnel: nobody here needs convincing to buy, you only need to stop getting in the way. Which is why fixing checkout almost always returns more money per hour invested than any other store work.

What the data shows

Baymard Institute aggregates dozens of studies into one figure: the average cart abandonment rate is around 70%. The breakdown of reasons matters more:

One more number sits alongside those: roughly 42% were "just browsing" with no intention to buy. That figure often gets used as an excuse — "people just browse, it's normal". The correct conclusion is different: you were never getting those 42%, so all of your upside sits in the remaining reasons — and every one of them is yours to fix.

Across years of checkout testing, Baymard estimates that the average large store can raise checkout conversion by roughly 35% through checkout design alone — no changes to pricing, range or traffic.

The five fixes that pay the most

1. Show the full cost earlier. The most common reason to leave is not the shipping price, it is the moment people learn about it. A shipping calculator in the cart, an estimate on the product page, a free-shipping threshold with progress ("€12 away from free shipping") — all of it turns a surprise into an informed choice.

2. Allow guest checkout. Forced registration is the second most common reason. It is also the cheapest item on this list to implement: offer the account after payment, as a single "save my details for next time" button, when the customer is already happy.

3. Remove surplus fields. Baymard's measurements put the average checkout at about 11 fields, while most stores need eight. Candidates for removal: separate first/last name fields, "company", a second address line, "how did you hear about us", email confirmation, phone repeat. Every field is one more chance to abandon.

4. Build a real mobile form. Correct keyboard types (inputmode), autofill (autocomplete), city auto-filled from postcode, large tap targets. Mobile checkout is where the gap between "fine" and "painful" is widest.

5. Show that paying is safe. Not an abstract "trust badge" from 2010, but specifics: payment method logos, HTTPS with no browser warnings, a recognisable store name on the card statement, visible contact details and return terms next to the pay button.

Mistakes I see in every second store

• The coupon field taking centre stage. A large "Promo code" box is a reminder that promo codes exist and an invitation to go looking for one. Hide it behind a modest link.
• Registration before the cart. Still out there: "log in first", everything else second.
• Validation errors only after submit. The form reloads, fields are cleared, the error is in small print at the top. Validate inline and never lose what was typed.
• A single payment method. Especially in the EU, where local methods (BLIK, iDEAL, Klarna, SEPA) are more familiar than cards for a large share of buyers.
• No delivery date at the payment step. The customer has agreed to pay but still doesn't know when they'll receive it.
• Hidden fees on the last step. Cash-on-delivery charges, "packaging", a surcharge for paying by card. That is a guaranteed exit and, worse, a permanent loss of trust.
• A mandatory phone number with no explanation. If the courier needs it, say so under the field.

Developer note

• Instrument checkout step by step. The begin_checkout → add_shipping_info → add_payment_info → purchase events from the GA4 ecommerce spec tell you exactly which step loses people. Without them you optimise blind.
• WooCommerce block checkout. The modern block-based checkout gives more control over fields and steps than the classic shortcode, but some plugins still aren't adapted to it — verify compatibility before the migration, not after.
• Autofill must work. autocomplete="given-name", family-name, street-address, postal-code, cc-number. Custom fields without the right attributes break browser and mobile-wallet autofill, which is a direct conversion loss.
• Never let scripts block the pay button. If payment waits on a chat widget or a pixel to load, you are paying for analytics with revenue.
• Log payment gateway failures. Part of your "abandoned carts" are actually failed payments. They only show up in gateway logs, and they are a completely different problem.
• Check order status webhooks. A WooCommerce classic: the order is paid but stays pending because the gateway webhook never arrived. The customer thinks there's no order; you think they abandoned the cart.

Abandoned cart emails: do them, but properly

The most obvious recovery tool and the most common source of irritation. What works, in my experience:

• First email after 1–3 hours, while the intent is alive. One email, not a five-part series.
• No discount in the first email. Otherwise you train your audience to abandon carts for a coupon.
• Discount only in the second email after 24–48 hours, and only for segments where the margin allows it.
• The link returns them to a filled cart, not to the homepage. Sounds obvious; breaks most often.
• Check the legal side. Sending these requires a lawful basis under GDPR — consent or otherwise. Capture the email before the person leaves, and do it honestly.

More on working with the list in retention and LTV.

Practical check

• Placed a test order on my own phone, start to finish
• Full cost including shipping is visible before the payment step
• Guest checkout is available without registration
• Counted the form fields: eight or fewer
• Browser autofill works on every address and card field
• At least two payment methods, including the familiar local ones
• Delivery date is visible at the payment step
• No new amounts appear on the final step
• Checkout step events are tracked and the drop-off point is visible
• The abandoned cart email returns people to a filled cart

What's next

Checkout is the end of the road. But some people never get there because they never found the product: site search and filters.

If you want this diagnosed and implemented: conversion optimisation — checkout is usually the first area I take on, because it pays back fastest.