A cart left before checkout is often an order postponed, so build the checkout for the shopper who comes back
In November 2021 PYMNTS published Consumer Choice at the Checkout, a report produced with Checkout.com on what keeps online shoppers with a merchant. It draws on a survey of 2,139 US consumers and on the quarterly test in which PYMNTS goes through the checkouts of more than 500 leading merchants. The findings are five years old, American and collected during the pandemic, so none of the percentages describes a European shopper today. One finding has aged better than the rest, because it is about behaviour rather than mood: 67% of the shoppers surveyed said they abandon carts before checkout and later return to complete at least one of those abandoned purchases.
That figure changes what an abandoned cart is. In most reports it is a loss, counted on the day it happens. For two in three of the shoppers surveyed, leaving a cart has at least once been a pause rather than a final decision. PYMNTS’s published summary does not say how long they stayed away, how often the return ended at the same shop, or how many of their abandoned carts were never completed at all. It does show that coming back to finish an order is ordinary behaviour, not an exception. The useful question for your checkout is therefore not only how many shoppers leave, but what the ones who come back find when they do.
The return visit should start where the first one stopped
A shopper who returns expects the basket they left. If that basket lives only in one browser session, the shopper who added products on a phone at lunch and comes back on a laptop in the evening starts again from the search box. Some will rebuild the order. Others will rebuild it at a competitor whose product page they also had open. Keep the basket on the device for a period that matches how long your customers take to decide, and where the shopper has made themselves known, through an email address given in the first checkout step or a login they chose, make the same basket available on their next device.
Do not make an account the price of a saved basket. An account wall at the moment of return is the same obstacle it was on the first visit, now placed in front of a shopper who had already decided to buy.
What the shopper entered should survive as well. A delivery address, a chosen delivery slot, a selected payment method: a shopper who stopped on the payment step and returns to an empty address form has been sent back two steps. Put the returning shopper on the step they left, with their details still in place, and let them change anything from there.
The terms should still be the same when they return
PYMNTS’s published figures also show how quickly merchants changed what shoppers were coming back to. Among the merchants PYMNTS tested, the share offering free delivery fell from 74% in the first quarter of 2021 to 63% in the second. In the consumer survey, 55% said that not being able to get free delivery could make them switch to another online merchant. PYMNTS advised merchants at the time to absorb delivery costs. Note what the survey measured. It asked what could make shoppers switch, not what they did after a merchant dropped free delivery, so the 55% shows how sensitive the subject is rather than how many orders were lost.
A returning shopper notices such changes more sharply than a new one, because they have seen the total before. If the delivery charge, the free-delivery threshold or the price of an item in the basket differs from what they saw on the first visit, the comparison is immediate and comes at the moment they had decided to buy. Terms sometimes have to change. When they do, show the new terms on the product page and in the basket rather than on the last step, and when the cost of a basket the shopper already put together has changed, say so in the basket, in one plain line, with the old and the new amount.
Measure the return, not only the exit
Abandonment is usually reported per session, and that hides the pattern PYMNTS describes. A shopper who leaves on Monday and orders on Thursday appears as one abandoned session and one converted session, with nothing linking the two. Link them. Count each week how many abandoned checkouts end in an order by the same shopper within seven days, and how many of those shoppers found their basket intact. Split it by device, because the move from phone to laptop is where a basket is most often lost.
Those two numbers tell you more about loyalty than the abandonment rate on its own. A rising share of returns that complete with the basket intact means the pause is working for you. A large group of returning shoppers who started from an empty basket means part of your abandonment is caused by the checkout forgetting, not by the shopper.
Then test it yourself. Put three products in your basket on a phone, enter a delivery address, stop on the payment step and close the browser. Come back the next day on a laptop and see what is left: the products, the address, the delivery choice and the terms you saw the day before. Every item that is missing is something a returning shopper has to redo before they can pay, and the first one to fix is the basket itself.
Sources: pymnts.com, Consumer Choice at the Checkout and pymnts.com, report findings
