Ecommerce checkout: where the decision gets interrupted

The decision to buy rarely ends at the buy button. It breaks earlier, in a price that never appears, a cart that gives no signal, a size chart that answers nothing. The ecommerce topic index holds the articles behind this post.

Where the decision gets interrupted

The product page study looked at 49 ecommerce sites and put the decision on the product page itself. A weak page costs more than a lost sale. It can drive the wrong purchase, a return, and a dent in trust.

The basics are short. The page needs a descriptive name, a recognizable image with zoom and a price that includes fees. It also needs a clear color and size choice, availability, an obvious add button with feedback, and a short description. Scanning readers give the start of the description most of their attention, so the first lines carry the decision. One image rarely answers enough. Comparison demands consistent information across variants, categories, sites, and competitors. In reviews, people jump to the negative ones first.

A clear choice before checkout follows the same rule.

Price and total cost

The B2B price study names price the most requested piece of information, and plenty of sites hide it. Buyers who couldn’t find a price left for a competitor, and the sites that showed price first earned a place on the short list. Price sets the category, and it lets a buyer compare and plan. An estimate already helps.

Hiding what the person came for stains the rest of the brand. When the exact number can’t be published, show a typical scenario, a range, or a suggested price. A complex price calculator usually fails.

Scope note: this framing of hidden price as a central failure comes from the B2B article. In the other articles here, price is a display detail.

Cart and the feedback after adding

The shopping cart study treats the cart as a waiting room: that is where people decide. It lists four supports: a full cart page, a clear image, a cart item linked to its product page, and easy removal. A small minicart forces scrolling and blocks comparison. A wrong or tiny image raises doubt about the selection and about the reliability of the site. Prefer removing an item over deleting it, and let the person zero the quantity.

The cart feedback study asks for three signals when an item is added: a counter on the icon, a visible confirmation, and a new button label. None of them should block adding more units. Without them the person can’t tell whether the action worked or whether the right item went in. A subtotal near the icon helps catch an error. Avoid a message that disappears on its own, because nobody can check it before it vanishes. Don’t disable the button after adding, and offer another add.

Size and measurement

The sizing study found that the main worry is whether a piece will fit, especially when returning it from another country is hard. A localized site should show the local size as the default and the original next to it. A global site needs a table that covers several regions and explains how the measurement was taken. One table for every brand reads as imprecise. What works: a table per brand, body dimensions per size, and the measurement of the garment itself, with the method explained. On mobile, hide the irrelevant column and keep the local size in the next column.

Mobile checkout

The mobile checkout study pushes for fewer steps, less typing, and the features the device already has. The cart sits one tap away. Removal gets a clear button without forcing the quantity to zero, and a change applies without a separate update button. Guest checkout stands out, above signing in or creating an account. The order summary sits at the top of the checkout. Location, with permission and an option to change it, feeds tax, shipping, and pickup. Each field gets the right keyboard, a calculated field fills itself, browser autofill works, and state and card expiry open as a field instead of a list.

A headless storefront helps control those field rules.

Save for later against the cart

The wishlist study, also based on 49 sites, describes the cart as external memory and a waiting area. Saving for later has to be easy to find and low effort, with no signup required. The label “wishlist” carries a bad connotation, so “favorites” works better. Hiding the function behind a gesture or a menu hides the feature itself. An item parked somewhere hidden makes the person think it is lost, and that sends them to a competitor. One warning: cart abandonment misleads as a metric, because an item left in the cart can become a purchase later.

Read that as one step in the funnel, not as a final sentence.

After the buy button

The 2004 article frames trust as the biggest challenge of ecommerce, since the company is virtual and nobody touches the product. The first order is the hardest, and treating the buyer well after the order raises the odds of a second one. The email confirmation, the packaging, and the receipt all belong to the total experience, and they calm the anxiety that grows when delivery runs late.

Where the sources disagree

Removing an item: the 2014 article allows removal by zeroing the quantity, and the 2018 article asks for a removal button and forbids forcing the zero. The intent is the same and the mechanism is opposite, so decide by context: a clear removal button for mobile, the quantity path for desktop.
Add confirmation: one article recommends a message in most cases and reserves an intermediate page for sites with few items, while the other treats both options as equal, with no ranking.
Minicart: one article criticizes the minicart for blocking comparison and asks for a full cart page, while the other wants the icon tap to land on the cart with no intermediate step.
Save for later: one article treats it as an accessory of the product page, and the other shows the label gets in the way and that requiring an account kills the feature.
The 2022 frame against 2018 and 2019: the state of ecommerce UX study ran 39 participants in three countries, and nearly all of them finished their tasks, so ease alone no longer decides. The field has moved a long way from the studies of 2000 and 2001. The other articles describe basic add-to-cart feedback that still fails. The facts don’t clash: the innovation layer advanced while the basics kept failing. That tension is the useful part.
Price: the hidden-price failure sits only in the B2B article. Elsewhere price is a display detail.