The most useful figure in the Shopify and BCG study is a time, not an uplift
In August 2023 Retail Dive ran a sponsored article from Shopify on what makes a checkout convert. It rests on a study Shopify carried out with Boston Consulting Group, published by BCG in June 2023 as Leading Online Shoppers to the Finish Line, which covered more than 220,000 ecommerce sites, more than a billion data points and more than 100 variables. The article condenses that into five drivers, each with an upper bound: up to 50.4% more conversion from accelerated payment methods, up to 18.9% from simplified checkout flows, up to 15.5% from customer accounts and subscriptions, at the expense of total traffic, up to 7.7% depending on device type, and up to 3.5% from apps and customisations. Figures like these travel well in presentations. The finding that deserves the most attention does not appear in the article at all. It sits in BCG’s own report: how long the checkout takes.
Read every “up to” as the ceiling of a correlation
BCG’s own announcement of the study contains the caveat the sponsored version leaves out. Of the more than 100 variables examined, not one explains more than 1.4% of the variation in lower funnel conversion, the stage that starts when a shopper begins the checkout. The large percentages describe the best cases in a pattern, not the effect a particular shop can expect from switching one feature on.
The basket finding shows why that distinction matters. According to BCG, conversion can rise by as much as 63% when a shopper who started with one or two items ends up with more than ten. A shopper who has filled a basket with ten items has largely decided to buy already. The size of the basket did not cause the purchase; it signalled it. Several of the drivers in the article carry the same risk of reading a sign of intent as a cause.
The wallet figure needs the same care. BCG states that its finding on accelerated payment methods applies to North America, and the methods it names are Meta Pay, Amazon Pay, PayPal, Shop Pay, Apple Pay and Google Pay. That is a list of the ways a North American shopper avoids a form, not a payment mix to copy into a Dutch checkout. The device finding is limited as advice for a different reason. Shoppers are much more likely to complete a purchase on a desktop, the report says, but you cannot move your shoppers to a desktop. You can make the phone checkout shorter.
Ninety seconds is a budget for the whole checkout
This is where the timing finding comes in. BCG reports that checkouts completed in under 90 seconds convert significantly better, and that lower funnel conversion may be around 47% lower on average when the checkout takes longer. It is a correlation too, and some checkouts are slow because the shopper hesitates for reasons that have nothing to do with the form. Unlike basket size or device, though, time is something your design spends directly.
Treat the 90 seconds as a budget and decide where it goes. A postcode that completes the address, a single name field, delivery options shown with their price and date, and a payment method the shopper recognises at a glance each take only seconds. A forced account, a label that has to be read twice, an empty promotion code box that sends the shopper off to search for a code, and a redirect that loses the basket each spend the budget without bringing the order closer. BCG makes two of these points itself: it advises against blank coupon code fields because they introduce doubt, and says fraud detection should not hold up the checkout flow.
Autofill shows what the budget rewards. According to the study, 99% of stores use autofill where it is available, so switching it on no longer sets one shop apart from another. Whether the fields are built so the browser can actually fill them, with the right field types and in an order it can follow, is where the seconds are won or lost.
In the Netherlands the fast route is a bank transfer, so time it separately
The accelerated methods in the study save time by filling in details the shopper would otherwise type. Most Dutch shoppers already skip that typing. The Dutch Payments Association describes iDEAL as an established Dutch payment method via bank transfer and reports that it accounts for 71% of all online purchases. Because it is a bank transfer, the shopper completes it with their own bank rather than in a form on your site. The same association notes that the transition from iDEAL to Wero was set to start in early 2026, so the name shoppers look for in your payment step is changing.
That step counts against your 90 seconds, but most of it happens outside your site. What you control is either end of it: how quickly the shopper reaches the payment step with a final total, whether the method appears first, under the name and logo it carries today, and whether the return from the bank lands on a confirmation page with the order intact. A shopper who comes back to an empty basket or an unclear payment status spends the rest of the budget, and usually more, finding out whether the order went through.
Put a clock on your own checkout
Most shops report conversion per checkout step and never the time the checkout takes. Add that measure: the time from the first checkout page to the order confirmation, as a median, split by device and by payment method. Then go through the checkout yourself on a phone, as a guest, with a stopwatch running, and write down where the seconds go. If the total is well over ninety seconds, the slowest step on that list is your first change, whatever the feature list of your platform promises.
Sources: retaildive.com, bcg.com, bcg.com and betaalvereniging.nl
