Measure
Conversion rate calculator
Your store's conversion rate from visitors and orders, with the formula.
Ecommerce checkout
Checkout optimization removes the reasons shoppers leave between starting checkout and paying. This guide covers why people abandon checkout, 12 ecommerce checkout best practices, how to measure each step, and how to find and test the fixes that matter for your store.
Checkout optimization is the process of making an online store's checkout easier, clearer and more trustworthy so that more shoppers who start checking out finish their order. It covers costs, steps, form fields, payment options, error messages, trust signals and mobile usability, and it is judged on completed orders and revenue, not clicks.
The checkout is the bottom of the funnel: the shoppers here already chose a product and showed they intend to buy. Small changes at this point can win or lose a lot of revenue, which is why, when Convertica starts work on a store, the checkout flow is one of the first things it reviews. You may also see the job called ecommerce checkout optimization, checkout process optimization or, in British spelling, checkout optimisation.
An ecommerce checkout is the sequence of screens between the cart and the order confirmation where the shopper gives contact, delivery and payment details and places the order. Most stores use one of a few checkout layouts, and the right one depends on how many details you need and how your customers like to pay.
| Checkout type | How it works | Strength | Watch out for |
|---|---|---|---|
| One-page checkout | All fields on a single page | Shoppers see everything that is needed | A long page that looks like a lot of work on a phone |
| Multi-step checkout | Contact, delivery and payment as separate steps | Short steps and a clear sense of progress | Extra page loads, and steps that hide costs until late |
| Accordion checkout | Steps open and close on one page | Progress without page loads | Errors in a closed section that the shopper cannot see |
| Express wallet checkout | A wallet button fills contact, address and payment | Very few taps for shoppers who use the wallet | Buttons that crowd out the main checkout path |
| Hosted checkout | The platform or payment provider runs the checkout pages | Security and payment updates handled for you | Limits on what you can change and test |
Shoppers leave checkout without buying mainly because of surprises and effort: extra costs revealed late, forced account creation, long forms, a missing payment method, security doubts, errors they cannot fix, and unclear delivery or returns. Some were never going to buy and were only checking the total. Checkout optimization targets the fixable reasons.
The cart stage has its own leaks, covered in the shopping cart optimization guide. For how to measure the share of shoppers who leave, see the guide to cart abandonment rate.
The ecommerce checkout best practices that matter most are showing the full cost early, offering guest checkout, asking only for what you need, filling in what you can, offering the payment methods your customers use, writing errors that explain the fix, keeping reassurance near payment and designing for the phone first. Each one removes a known reason to leave.
Show shipping, taxes and fees, or a reliable estimate, in the cart and on the product page. Surprise costs at the last step are the classic reason to abandon. If you have a free shipping threshold, tell shoppers how far they are from it.
Let shoppers buy with an email address. Offer to create an account on the confirmation page, where saving their details is a convenience rather than a barrier, and offer sign-in for returning customers without making it the main path. For the sign-in form itself, see login form design.
Remove every field you do not need to fulfill the order. Make company name, second address line and phone number optional unless your carrier needs them, and say why if you ask for a phone number. Fewer fields means less typing and fewer errors.
Use the delivery address as the billing address by default, with a checkbox to enter a different one. WCAG 2.2 success criterion 3.3.7, Redundant Entry, asks for exactly this: information already entered in the same process should be filled in or offered for selection, not typed again.
Use the standard HTML autocomplete values, such as email, given-name, address-line1, postal-code and cc-number, so browsers and password managers can fill fields correctly. Add address lookup if your platform supports it, and use inputmode="numeric" or type="tel" so phones show the right keyboard.
Check which cards, wallets and pay later services your customers expect in each country you sell to. Express wallet buttons can save mobile shoppers a lot of typing, but keep the standard checkout as the clear main path so the buttons do not confuse people who will not use them. Each method also has its own fee: the Stripe fee calculator and the PayPal fee calculator show what a payment costs you at each provider's published US rates.
On a multi-step checkout, show the steps and where the shopper is. Keep a summary of the items, quantities and total in view, with a way to edit the cart without losing what has been entered.
Validate fields as the shopper leaves them, mark the field in error and say how to fix it in text. A declined card needs a message that suggests what to try, such as checking the expiry date or using another card. WCAG 2.2 success criteria 3.3.1 and 3.3.3 cover identifying errors and suggesting corrections.
Shoppers decide whether to trust you at the moment they type card details. Keep the store's look consistent through checkout, show how payment is secured, and put the returns policy, guarantee and contact options near the place-order button.
Show an estimated delivery date or range for each shipping option, not only a carrier name, and say what returns cost and how long shoppers have. Clear terms answer the last questions before payment.
Hide the main navigation, banners and pop-ups in checkout. If you show a discount code field, keep it small or behind a link, because a large empty code box sends some shoppers off to search for codes.
Use one column, large fields and buttons, labels above fields and a place-order button that is easy to reach with a thumb. Test the full checkout on real phones and on a slow connection, because that is where most problems hide.
Good ecommerce checkout UX means the shopper never has to stop and think: every field is expected, every label is clear, every error explains itself and every cost is already known. Most checkout problems are small details, which is why the only reliable way to find them is to place real test orders on real devices.
Measure checkout performance with a funnel from cart to order confirmation: the share of carts that start checkout, the share of checkouts completed, the drop at each step and the payment failure rate. Split every number by device, browser and payment method, because averages hide the step that is broken for one group.
| Metric | Formula | What it tells you |
|---|---|---|
| Checkout start rate | Checkouts started / carts created | Whether the cart page sends people on |
| Checkout completion rate | Orders / checkouts started | How well the checkout itself converts |
| Step drop-off | Shoppers leaving at a step / shoppers reaching it | Which step to investigate first |
| Payment failure rate | Failed or declined payments / payment attempts | Problems with the payment provider, fraud rules or card entry |
| Revenue per visitor | Revenue / visitors | The net effect of a change on orders and order value together |
| Conversion rate | Orders / visitors | The store-wide result that checkout changes should move |
In Google Analytics 4, a funnel exploration built on the standard ecommerce events, from view_cart through begin_checkout, add_shipping_info and add_payment_info to purchase, shows the drop at each step. The conversion rate calculator turns visitors and orders into a rate, and the guide to ecommerce metrics covers the rest.
To find what is wrong with your checkout, start with the funnel report to locate the leaking step, then use recordings, test orders, payment logs and a short exit survey to learn why shoppers leave there. Each source shows a different part of the problem, and a finding that appears in two of them is worth acting on.
A full CRO audit runs the same process across the whole store, from landing page to confirmation.
Test checkout changes with A/B tests judged on completed orders and revenue per visitor, sized with a sample size calculator before launch and read separately for desktop and mobile. Fix outright bugs without a test. Check first what your platform allows, because some hosted checkouts limit what you can change or test.
Many of the best checkout wins happen before the checkout. In a Convertica test on one store, telling shoppers how far their cart was from free shipping improved desktop results, while mobile showed no significant change. In another, showing shipping and refund policies on the cart page helped on desktop and was neutral on mobile. Both point to the same lesson: answer cost and risk questions before the shopper reaches payment, and read each device on its own.
The A/B testing guide covers how to plan and read a test.
An AI shopping agent can complete a checkout only if it can read the costs and policies and fill in the fields. Guest checkout, standard labeled fields with autocomplete values, shipping costs and returns terms in plain text, and no step that relies on a mouse-only widget all help. Agents acting for customers are new, so treat this as a check, not a channel. To check yours, run an AI shopping agent through your checkout: it stops before payment and shows where it got stuck.
Nobody, Convertica included, has years of data on how agents shop yet. Agent readiness is one of the eight checks in the free CRO audit, and the agent readiness checklist goes further.
Use this checkout optimization checklist on your own store. Place a test order on a phone, then answer each line yes or no. Every "no" is a candidate fix: correct the clear problems straight away and test the ones you are unsure about.
The most common checkout mistakes are revealing costs late, forcing account creation, asking for information twice, hiding errors, offering too few payment methods and leaving the store's navigation and promotions in the way. Each one is easy to spot in a test order and usually cheap to fix.
Measure your store and plan checkout tests with Convertica's free calculators.
Measure
Your store's conversion rate from visitors and orders, with the formula.
Plan a test
How many shoppers each checkout version needs before a test starts.
Read a result
Whether a finished checkout test is a real result.
Checkout optimization is the process of making an online store's checkout easier, clearer and more trustworthy, so that more shoppers who start checking out finish their order. It removes the usual reasons to leave: costs revealed late, forced account creation, long forms, missing payment methods, security doubts and confusing errors.
Checkout optimization is the work of making an online store's checkout easier, faster and more trustworthy, so that more shoppers who start checking out complete their order. It covers the number of steps and fields, costs and delivery information, payment options, error handling, trust and mobile usability.
Shoppers usually leave checkout because of extra costs they did not expect, being made to create an account, a long or confusing form, a missing payment method, worries about security, an error they cannot fix, or delivery and returns terms that are unclear. Many were also only comparing prices.
There is no reliable universal figure, because checkout completion depends on your products, prices, traffic and devices. Measure the share of shoppers who start checkout and finish it, split by device and payment method, and compare it with your own store over time. Improving your own rate is what counts.
Not always. A one-page checkout shows everything at once, while a multi-step checkout breaks the same fields into short steps with a progress indicator. Either can work if the total number of fields is low and errors are clear. Test the change on your own store rather than assuming.
Yes. Making shoppers create an account before buying is one of the most common reasons they leave. Offer guest checkout as the default, then invite the shopper to save their details and create an account on the confirmation page, after the order is placed.
A good checkout shows the full cost early, asks only for what is needed, never surprises the shopper, makes errors easy to fix, offers the payment methods your customers use, keeps reassurance close to the payment fields and works well on a phone. Everything else is detail.
In ecommerce, CRO means conversion rate optimization: increasing the share of store visitors who place an order. It covers product pages, the cart and the checkout, where removing friction such as unexpected extra costs, forced account creation and long forms can recover sales from shoppers who already meant to buy.
Set up a funnel report from cart to order confirmation, split it by device, browser and payment method, and look for the step with the biggest drop. Then watch session recordings of shoppers who left at that step, place test orders yourself on a phone, and check your payment logs for declines and errors.
Usually, yes, but hosted checkouts on some platforms limit what you can change or test, so check your platform first. Judge checkout tests on completed orders and revenue per visitor, not on clicks, and read desktop and mobile results separately because they often differ.
It means making sure an AI shopping agent acting for a customer can complete your checkout: guest checkout, standard labeled fields with autocomplete, costs and policies in plain text, and no step that only a person with a mouse can pass. It is a new area, and agent readiness is one of the eight checks in Convertica's free CRO audit.
Ask Convertica
Convertica
Hi, I'm Convertica's AI assistant. Ask me anything about the free CRO audit, the one page and full site rebuilds, or any of the free tools.