The average online checkout asks for 11.3 form fields. Baymard Institute, which has been measuring this since 2012, says most sites need eight. The average flow runs 5.1 steps.

Those three numbers sound small. They describe the most expensive page you own, because unlike every other page on your site, the people reaching it have already decided to give you money. They arrive wanting to buy. A meaningful share then leave, and most merchants could not tell you at which field.

The widely quoted figure is that around seven in ten carts are abandoned. That number is real and it is also routinely misused, which is where this gets interesting.

---

Where shoppers actually go

• The documented average abandonment rate is 70.22%, drawn from 50 separate studies.

  From Baymard's aggregated research. Treat it as the ceiling of the problem rather than the size of the opportunity, for reasons below.



• Extra costs are the single largest stated reason, at 40%.

  Shipping, tax and fees appearing later than expected. Nothing else comes close.



• Forced account creation drives 18% away. A long or complicated checkout accounts for 17%, and site errors or crashes another 17%.

  Those last two are frequently the same page behaving differently on a weak connection.



• Checkout forms complete 54.4% of the time, and only 51.4% on mobile.

  From Zuko's form benchmarks. Your worst-performing device is carrying most of your traffic.



• Digital wallets took 56% of global ecommerce value in 2025.

  Worldpay's Global Payments Report 2026, published in March, puts combined wallet spending above 13.8 trillion dollars. In Brazil, Pix alone accounted for 42% of ecommerce.

Here is the part the industry tends to leave out. Baymard's own research finds that 42% of US shoppers have abandoned a cart simply because they were browsing and not ready to buy. No checkout design fixes that, and no honest agency will promise to. The reasons listed above are the distribution among the rest. Your realistically recoverable share is a fraction of that 70%, and it is still almost certainly the largest single pool of lost revenue on your site.

---

1. Show the real total before anyone reaches the checkout

Forty percent is the largest number in this dataset and it describes a trust failure rather than a pricing one. People do not abandon because shipping costs money. They abandon because the price changed after they had decided.

Put shipping costs, or a shipping calculator, on the product page and in the cart. State tax handling plainly. If you have a free shipping threshold, show how far away the basket is from it. The surprise is the problem, so remove the surprise rather than the cost.

2. Let people buy without creating an account

Guest checkout is the highest-certainty fix on this list, and plenty of stores still make an account mandatory because the marketing team wants the email address.

They get the email address anyway at the order confirmation step. Offer guest checkout as the default path, then invite account creation after the purchase is complete, when the customer has a reason to want one: order tracking, faster repeat purchase, returns. Asking for a commitment before you have delivered anything is the wrong order of operations.

3. Get to eight fields and stop asking the same thing twice

Audit your checkout field by field against the standard Baymard sets. Most flows carry a second address block that could be a checkbox, a confirm email field that measurably hurts completion, a company field nobody needs, a title dropdown that serves no purpose, and a phone number with no stated reason attached.

Use address lookup so a postcode fills the rest. Detect card type rather than asking. Default the billing address to the delivery address. Every removed field is one fewer chance to lose someone who had already agreed to pay.

4. Offer the payment methods your customers already use

With wallets at 56% of global ecommerce value, a checkout offering only card entry is asking a majority of shoppers to do something harder than they are used to. Baymard's data shows 9% of abandoners citing insufficient payment options.

Which methods matter depends entirely on where you sell. Pix in Brazil, iDEAL in the Netherlands, wallets almost everywhere, and buy now pay later in markets where it has become normal, such as Germany, where it accounts for roughly 18% of ecommerce. Check your own analytics for where orders come from before adding anything, because every extra method carries integration and reconciliation cost.

5. Make the delivery promise specific

Slow delivery is cited by 20% of abandoners, and vague delivery is frequently the real culprit. "Ships in 3 to 5 business days" leaves the customer doing arithmetic that includes a weekend.

Show an actual date range. Show it before the checkout, not inside it. If you have a faster paid option, present it at the point where someone is deciding whether the timeline works, which is the product page rather than step four.

6. Earn the trust that the payment step requires

Nineteen percent of abandoners say they did not trust the site with their card details. This is partly design and partly infrastructure.

Visible contact information, a real returns policy, consistent visual quality, no broken layouts, no stock imagery that contradicts the brand. Keep the payment step on your own domain where possible rather than bouncing to an unfamiliar one. The cues people use to judge safety are mostly unconscious, which is why a checkout that merely looks unfinished converts worse regardless of how secure it actually is.

7. Treat errors and declines as design problems

Seventeen percent abandon because of errors or crashes, and 10% because a card was declined. Both are usually survivable if the page handles them well.

Validate fields as people type rather than after submission. Never clear a filled form on an error. Keep the basket intact through a failed payment. When a card declines, say what to try next and offer an alternative method on the same screen rather than returning a generic failure. Test the whole flow on a throttled connection, because the error rate on a bad mobile network is nothing like the one on your office wifi.

8. Put the returns policy where the hesitation happens

Thirteen percent cite an unsatisfactory returns policy. For a certain class of product, especially clothing and anything with fit or taste involved, the returns terms are the purchase decision.

Write the policy in plain language, state who pays for return postage, and link it from inside the checkout rather than burying it in the footer. If your terms are genuinely restrictive, that is a commercial decision worth revisiting rather than a copywriting problem.

---

The number everyone quotes and nobody qualifies

Seventy percent abandonment sounds like a 70% opportunity, and it is not. Roughly four in ten people who abandon were never buying today. They were comparing, saving for later, killing ten minutes, checking a price against a competitor. Treating that group as recoverable is how conversion projects end up with targets nobody can hit and a nagging sense that the work failed.

The honest framing is narrower and more useful. There is a specific, identifiable group of people who intended to buy and were stopped by something you control: a price that changed, a form that demanded an account, a field that errored, a delivery date that was never stated. That group is smaller than the headline and it is reachable, and the fixes above are ordered by the evidence of how many of them each one affects.

One technical caveat worth knowing, since wallet growth is driving a lot of replatforming conversations. Worldpay counts card-funded wallet transactions as wallet volume rather than card volume, so some of that 56% is card spending that moved into an app rather than new behaviour. The direction is still clear and the implication for your checkout is unchanged, but it is not quite the clean displacement of cards that the headline suggests.

The last thing worth saying is about where to spend. Most stores measure their checkout on a desktop, fix it on a desktop, and sell to a majority on mobile, where completion sits several points lower. If you only do one thing from this list, do it on a phone, on a bad connection, with one thumb.

---

Three tests to run this week

• Buy something from your own store on your phone, on mobile data, with a card you have not saved.

  Count the taps. Note where the total first appeared and whether it matched what you expected on the product page.



• Count your checkout fields and compare against eight.

  List every field above that number and write down what breaks if you remove it. Most of the answers will be nothing.



• Deliberately fail a payment.

  Use a test card that declines, or an expired one. See whether your basket survives, whether the form keeps what you typed, and whether the message tells you what to do next.

An hour of work and no budget, and at the end you will know which of the eight items above is actually costing you the most.

Want your checkout measured, rebuilt and tested where your customers actually shop? Talk to Spark about the flow between a full basket and a paid order.