A checkout can get harder to complete without anyone changing the checkout
Checkout friction is usually treated as a design problem: too many fields, a hidden guest option, a slow page. A benchmark PYMNTS published in October 2021 with the payments company Checkout.com points at another cause. The first edition of the report, Removing Friction at the Checkout, was reported on 12 October 2021, and on 25 October PYMNTS wrote that the friction it measured had gone up between the first and second quarter of that year. The authors name a reduction in free shipping as part of the reason. No screen had to change for that to happen. What changed was what the shopper was asked to accept.
The Checkout Conversion Index, which PYMNTS has run since 2016, looks at non-Amazon online merchants and scores the features that decide how much friction a shopper meets on the way to an order. For this report its analysts shopped anonymously at 510 leading merchants in the second quarter of 2021 and surveyed nearly 2,140 US consumers. The overall score went from 59.5 in the first quarter to 58.9 in the second. The authors call that a minor increase in friction, and the published articles give no margin of error, so the headline score proves little on its own. What the index measures is more useful than the score: it counts what merchants offer, so when an offer is withdrawn, it shows up as friction.
The shopper meets a pricing decision as a checkout problem
A free shipping threshold is often set by whoever manages margin, a returns policy by customer service and legal, the list of payment methods by finance or IT. None of these is normally logged as a checkout change. The shopper meets every one of them in the checkout all the same. The delivery charge appears when the total is calculated, the returns terms are read before the order or not at all, and the payment methods appear on the last screen. When one of those decisions becomes less generous, the funnel shows a drop at a step whose design did not change, and the team looks for the cause in the layout.
The consumer survey behind the index shows how directly those decisions reach the shopper. Among respondents who had abandoned a cart, the largest group, 37%, said they were simply not ready to buy. A further 17% named shipping costs as the main reason and another 17% high prices, two answers about terms that are set outside the checkout and only become visible inside it.
The pattern has not faded since 2021. Baymard Institute’s survey of US online shoppers, on a list last updated in September 2025, ranks extra costs for shipping, tax or fees first among reasons for abandoning a checkout, at 40% once shoppers who were only browsing are set aside. Another 12% say they could not see or calculate the total cost up front. Baymard’s checkout benchmark of more than 180 leading US and European sites, updated in November 2025, rates 64% of desktop checkouts mediocre or worse. Four years on, the commercial terms and the design are both still costing orders.
A new promise does not cancel a new charge
The 2021 data contains a telling pair of movements. The share of merchants offering free shipping fell from 74% in the first quarter to 63% in the second, while the share offering guarantees and refunds rose to 86%, nearly ten points above the first quarter and 18 points above the end of 2020. In aggregate, merchants were offering less on delivery cost and more on reassurance; the report does not say whether the same merchants made both moves.
On paper, trading free shipping for guarantees and refunds would look like a fair exchange. For the shopper the two arrive at different moments. A delivery charge lands in the total on the payment step, exactly when the decision is made. A refund guarantee only helps if the shopper has read it, and if it sits on a separate policy page, the shopper has to go looking for it. This is a conclusion, not a finding of the study: a merchant who replaces free shipping with a stronger returns promise has to put that promise where the charge appears. One sentence next to the delivery line in the basket and again on the payment step does more than a policy page the shopper has to find during checkout.
The same applies to every change in that direction. If a free shipping threshold goes up, show the remaining amount in the basket before the shopper reaches the delivery step. If a return window gets shorter, change the line on the product page on the same day as the policy. The shopper accepts a less generous term more readily when it is stated early than when it turns up as a surprise in the total.
More payment methods did not make a top performer
The index holds a second lesson. Middle-ranked merchants increased the average number of payment methods they support by 13% and bottom-ranked merchants by 18%, and both groups beat the top performers on payment choice. They still ranked lower, because the ranking rests on the overall score across all the features the index looks at, not on payment choice alone. A longer list of methods on the payment step does not make up for a charge that appears too late. What counts there is that the shopper finds their own method at once, not how many others stand beside it.
Log commercial changes next to design changes
Most teams keep release notes for the checkout: a new form, a redesigned payment step, a changed button. Few record the date a shipping threshold went up, a return window was shortened or a payment method was dropped in one market. Start that log now, in the same place as your release notes, and mark every entry on your checkout funnel. When completion at the delivery or payment step drops, check the log before you open the design. And before the next change to shipping, returns or payment methods goes live, decide where the shopper will first see it and put the explanation there.
Sources: pymnts.com, pymnts.com, baymard.com and baymard.com
