Why your checkout is losing sales you already won

Seven out of ten shoppers who put something in an ecommerce cart never buy it. The Baymard Institute has tracked this figure through nineteen years of checkout research, and it holds steady around 70% no matter how many businesses read the number and assume they're the exception.

The instinct is to blame the product, the price, or the traffic source. Usually it's none of those. A shopper who adds something to a cart has already decided they want it. They didn't leave because the photography wasn't good enough. They left because the steps between "add to cart" and "order confirmed" asked more of them than they were willing to give.

Build an Ecommerce Store That Sells treats that moment as a handoff, not a pitch. By the time someone reaches checkout, persuasion is over. The only job left is getting out of their way.

The account wall, and the fields that don't need to exist

The single largest point of friction in ecommerce, according to the book, is the fight between "guest checkout" and "create an account." Marketing wants the account: it's the return path for newsletters and loyalty programs. The shopper, who doesn't know yet whether they like the brand, experiences a forced registration wall as a stranger asking for a background check before selling them a coffee. Roughly 26% of shoppers abandon a cart rather than clear that wall.

The book's fix, which it calls the "Account Later" approach, reverses the order of operations instead of removing the account option entirely. The shopper enters an email address first, for the receipt. If no account exists, they continue as a guest, no interruption. They fill in shipping and payment and complete the purchase. Only on the confirmation page, once the order exists and the shopper feels good about the decision they just made, does the store ask if they'd like to save their details for next time, and by then, the only field left to fill in is a password, since everything else is already on file. ASOS reported a conversion lift after making exactly this change: removing forced registration and defaulting to guest checkout.

Past the account question comes the form itself, and the book's instruction here is a "Need to Know" rule: if a field isn't essential to shipping the box or processing the payment, it has to go. It names a specific kill list. Address Line 2 confuses shoppers who don't live in apartments. Reveal it behind an "Add Apt/Suite" link instead of showing it by default. Company Name is noise unless the store is strictly B2B. Confirm Email duplicates what browser autofill already handles, and only adds a chance for the shopper to make a second typo instead of catching the first one.

The cost of getting this wrong compounds at the payment screen specifically. Unexpected shipping costs revealed for the first time at the final step cause close to half of cart abandonments. The book calls it "sticker shock," and the fix is structural rather than persuasive: show the shipping calculation in the cart or early in checkout, not as a surprise at the card field. The same screen is also where trust anxiety peaks, which is why security badges and encryption notices belong directly beside the card fields rather than in the footer, and why express wallets (Apple Pay, Google Pay, PayPal) solve two problems in one tap: no typing, and a payment method the shopper already trusts.

Where people go wrong

The first failure is treating checkout as a design problem instead of a facilitation problem, adding pop-ups, chat widgets, and countdown timers to a stage where the shopper has already said yes and just wants to finish. Each one adds a decision the shopper didn't ask to make, at exactly the point where every added decision costs the most.

The second is launching without ever completing an actual transaction with a real card. The book insists on buying your own test product in an incognito window before opening the store, because a broken payment gateway or a confirmation email landing in spam is the kind of failure that only shows up when you check it end to end, and by the time a customer reports it, you've usually lost several days of sales silently.

The third is fixing the wrong stage of the funnel. A spike in "add to cart" activity with few completed purchases points at checkout friction; a high bounce rate on product pages before anything gets added points at a persuasion problem instead: unclear photography, unanswered objections, a hidden price. Confusing the two means redesigning the part of the store that was already working while the actual leak stays open.

This friction audit assumes the shopper already trusts the store enough to start checking out. Getting them that far in the first place (the brand story and category structure that build that trust) is its own system, covered in the e-commerce guide.