Your cart is a shortlist the shopper means to come back to, so it has to be there when they return
In November 2021 a startup called Carrot launched a browser extension for Chrome on desktop that collected the carts a shopper had left open across different webshops. It started by going through the shopper’s browsing history for shops they had been to, and pulled the abandoned carts from those shops into one overview. It kept carts from the previous 45 days, tracked price changes on the items in them and let users group products into collections they could share. Its co-founder and chief executive, Bobby Ghoshal, saw the abandoned cart as unfinished shopping rather than lost interest. Explaining to Glossy why shoppers leave, he put it simply: “you’re still shopping”. He also told PYMNTS that current approaches were not working for anyone, and the article noted that the usual merchant response, emails and display ads to pull shoppers back, was something consumers had learned to ignore.
The coverage dates from 2021 and the extension was a young product at the time, so read it as a signal about shopper behaviour, not as a market that took shape. The behaviour itself has not gone away. Baymard Institute’s latest survey of US online shoppers, on a page last updated in September 2025, reports that 42% have left a cart because they were only browsing or not ready to buy. Baymard links this kind of abandonment to window shopping, price comparison, saving items for later and gift ideas, and calls it largely unavoidable. As an abandonment, that is right. As a return visit, it is not settled at all.
A third party built the memory the cart did not have
Carrot’s premise was that each shop’s cart works alone. Ghoshal found it odd that shopping still happened one site at a time. A shopper who compares lamps in four shops over a week has to remember where each one was, and the cart in each shop may have emptied by the time they come back. The extension ran on desktop only. On a phone, TechCrunch reported, users could send products to a Carrot number by text message. Shoppers, the article noted, were already sending products to themselves that way. Shoppers were doing the remembering by hand.
The business model makes the cost of that gap visible. Carrot earned affiliate revenue: according to TechCrunch, the retailer paid Carrot for each customer who checked out after their first session. In practice that could mean a fee on an order for a product the shopper had already put in the retailer’s own cart. Whatever happened to the extension since, the lesson for your checkout stands: if your cart does not hold on to the shortlist, someone else will offer to, on their terms.
What a cart that remembers does
Start with how long a cart survives. Check how many days your platform keeps a guest cart on the same device, and what happens to the cart of a logged-in customer who switches from phone to laptop. A cart that empties when the session ends treats every shopper who compares as one who left. Give the guest cart weeks, not hours, and sync the logged-in cart across devices.
Next, let the shopper carry the cart. A button that sends this cart to an email address the shopper types in, or copies a link that reopens it, does what the texted items did, but with your prices and your stock. Because the shopper asks for it, it is not a reminder sequence, and it reaches the device they will buy on. The same link works for the shared lists Carrot users made for birthdays, weddings and other life events.
Then decide what the cart says when time has passed. A product that sold out should stay in the cart with its status and an alternative, not disappear. A price that changed should show the old and the new amount in the cart itself, where the shopper sees it before paying. TechCrunch reported that Carrot tracked price changes on saved items, and PYMNTS described alerts for price drops as a feature still being tested. It is better that they hear about a price drop in your cart than from an extension.
Finally, keep removal and postponement apart. A shopper who wants to narrow the shortlist should be able to move an item out of the order without deleting it, so the cart can hold both what they buy today and what they are still weighing.
A cart reopened after a week has to recheck itself before the payment step
A remembered cart brings its own risk. When the shopper returns after a week, the delivery date, the stock, a discount code with an end date and the delivery charge may all have changed. If the cart shows last week’s figures and the checkout shows today’s, the shopper meets the difference at the payment step, which is the worst place to discover it. Recalculate the cart on reopening and state what changed at the top of the cart, in one line, before the shopper starts the checkout. Then keep the route from that cart to the payment step as short as for a fresh order.
Count the orders that started in an earlier session
Pull last quarter’s orders and find the share whose first add to cart happened in an earlier session, and how many days lay between that moment and the order. That share tells you how many of your buyers already use the cart as a shortlist. Then run a simple test: add three products on a phone without logging in, close the browser, open the shop again a week later on the phone and on a laptop, and write down what is still there and what changed without explanation. Fix what the test shows before you add another reminder email.
Sources: pymnts.com, glossy.co, techcrunch.com and baymard.com
