Almost every article about cart abandonment opens with the same statistic: roughly 70% of online carts are abandoned. It's an accurate number and a misleading target, because most of those carts were never lost sales. People add things to carts to check delivery costs, to compare prices across sites, to save something for payday, or simply to keep a shortlist. Treating all of it as failure sends you chasing a number you can't move — while the genuinely recoverable losses, the people who wanted to buy and gave up, sit unexamined. This guide is about finding and fixing those.
What the 70% actually contains
Research behaviour — not fixable, and not a problem. Price comparison, checking shipping costs, saving for later, browsing, wishlist substitution. These people were never buying in that session, and no checkout change will alter that.
Genuine friction — fixable, and expensive. People who intended to purchase, entered the checkout, and abandoned because something went wrong, cost more than expected, or asked too much of them.
Your overall abandonment rate blends the two into a number you can't act on. The meaningful metric is drop-off after checkout begins.
That distinction reframes the whole exercise. Someone adding to cart is expressing interest; someone entering checkout is expressing intent. Optimising the flow is about protecting the second group, and everything below assumes that focus.
Diagnose your own flow first
Generic fix lists are why so much checkout work produces nothing — the biggest leak differs by store, and applying someone else's priority order means fixing whatever happens to be broken elsewhere. Instrument each step separately and let your own data direct the effort.
Track cart viewed, checkout started, contact details completed, shipping selected, payment entered, and order confirmed as distinct events, then calculate the conversion rate between each consecutive pair. One of those transitions will be visibly worse than the others, and that's your first project. Do the same split by device, because mobile and desktop patterns rarely match. Getting these events recorded accurately is the same discipline covered in setting up conversion tracking you can trust — a step-level view is useless if the events themselves are unreliable.
The leading cause isn't cost — it's timing
Across essentially every study of why people abandon a purchase they meant to make, unexpected extra costs come top: shipping, taxes and fees appearing later than the shopper expected. It's tempting to read that as "our prices are too high" and reach for a discount. That's usually the wrong conclusion.
The real objection People rarely abandon because the total was too high. They abandon because it became too high later than they expected — and surprise reads as bad faith.
The distinction matters because it relocates the fix. A shopper who knows shipping is £6 before adding to cart is deciding whether to buy at £6. A shopper who discovers £6 at the payment step has been recalculating trust as well as price. The same cost produces very different outcomes depending on when it appears.
Which means the highest-return change here isn't a checkout change at all — it's disclosing costs earlier: on the product page, in the cart, and in the header rather than in a final summary. Free-shipping thresholds shown with progress ("you're £12 away") work for the same reason. This is really an extension of product page optimisation: the checkout inherits whatever expectations the product page set.
Stamped
Reviews, ratings, UGC and loyalty platform for e-commerce brands, integrated with Shopify and Klaviyo.
Best for: E-commerce brands collecting reviews, UGC & loyalty
The friction inventory
| Cause | Fix |
|---|---|
| Forced account creation | Offer guest checkout; invite account creation after purchase |
| Too many form fields | Remove anything not needed to fulfil the order; use address lookup and autofill |
| Can't see the total upfront | Persistent order summary with the running total visible at every step |
| Limited payment options | Add digital wallets — they also bypass most manual entry |
| Payment security doubts | Place trust signals at the payment step specifically, not the homepage |
| Unclear delivery or returns | State delivery dates and the returns policy before payment, plainly |
| Errors and validation | Inline, forgiving validation; accept spaces in card numbers |
| Slow loading | Treat checkout speed as a revenue metric, not a technical one |
Two of these deserve extra attention. Guest checkout is consistently among the highest-return changes available, because forcing account creation asks someone to commit to a relationship when they only wanted a transaction. The productive compromise is to let them buy as a guest, then offer an account on the confirmation page with their details pre-filled — you get the account without risking the sale.
And trust signals belong at the payment step. Security badges on your homepage reassure nobody, because doubt doesn't arise until someone is typing card digits. Put the reassurance where the anxiety is.
The mobile gap
Mobile abandonment consistently runs higher than desktop, and it's where most stores have the largest recoverable loss. The causes are mundane and specific: small tap targets, forms that demand precise typing, the wrong keyboard appearing for numeric fields, layout that shifts as the page loads, and a total that requires scrolling to find.
The single most effective mobile intervention is usually adding digital wallets, because they replace the entire form-filling ordeal with an authentication step. Beyond that, test your own checkout on a phone, on mobile data, with one hand — the way customers actually use it, rather than on a desktop simulator. The gap between those two experiences is where a surprising amount of revenue disappears, and closing it is genuinely web development work as much as marketing.
One page or several?
This debate absorbs more energy than it deserves. Neither format wins universally, and studies supporting each are easy to find because the format isn't the variable that matters.
What consistently helps is reducing the total number of fields regardless of how they're arranged, showing clear progress so people know how much remains, and eliminating surprises at every stage. A well-signposted multi-step checkout routinely beats a cramped single page; a clean single page beats a confusing multi-step flow. If you have enough traffic to test it properly, test it on your own customers. If you don't, spend the effort on field reduction instead — that one reliably helps either way, in much the same spirit as the clarity principles behind good landing page design.
Capture the email early
One structural change pays off twice. Ask for the email address at the start of checkout rather than the end. It's a low-commitment field people expect to provide, and it means that if someone abandons later you can actually reach them — which converts an anonymous loss into a recoverable one.
That's the bridge to recovery: even a well-optimised checkout will lose people to interruptions, second thoughts, and comparison shopping, and a sequence of well-timed reminders recovers a meaningful share of them. The mechanics are covered in cart recovery flows that convert, and the value of that captured address compounds well beyond the single order, as we argue in building an owned audience.
Worth being clear about the order of operations, though: recovery emails treat the symptom. Fixing the flow prevents the loss. Do both, but fix the flow first — otherwise you're paying to re-acquire people your own checkout drove away.
Why this matters more in 2026
Checkout optimisation has always been sensible. It's now close to essential, for two reasons.
First, traffic costs more. With CPMs up substantially year over year — the picture we set out in rising CPMs and how advertisers are adapting — every visitor you lose to a fixable friction point is more expensive to replace than it was. Conversion rate improvement is one of the few levers that reduces your dependence on buying more attention.
Second, expectations have hardened. Shoppers arriving from fast, frictionless environments — and increasingly from AI-assisted product discovery, as covered in how AI search is changing product discovery — bring the standards of the best checkout they've used to yours. Tolerance for a clumsy form is lower than it was.
Where to start
If you do nothing else, work through this order. Instrument your steps and find your own worst transition. Disclose shipping costs before the cart. Enable guest checkout. Add digital wallets. Capture email at the start. Then test your own checkout on a phone and fix whatever irritated you.
That sequence front-loads the changes with the highest ratio of impact to effort, and it's deliberately biased toward removing things rather than adding them. If you want to look wider than checkout — because a conversion problem is sometimes really a traffic-quality or expectation problem originating much earlier — the method in an end-to-end funnel audit will locate it.
The bottom line
Stop treating the 70% figure as a target. Most abandoned carts belong to people who were browsing, comparing, or saving for later, and no amount of checkout work converts them. The money sits with the smaller group who genuinely intended to buy and hit something in the way — usually a cost that appeared later than they expected, an account they didn't want to create, a form that asked too much, or a phone experience nobody tested. Instrument your steps to find your own weak point, disclose costs early, remove fields ruthlessly, and capture the email at the start so the ones you still lose remain reachable. Fix the flow first; recover second.
Losing sales at the last step?
We find exactly where your checkout leaks, fix the friction, and turn abandoned carts into completed orders.
Explore E-commerce Marketing →Frequently asked questions
What is the average cart abandonment rate?
Research consistently places the average around 70%, though the figure varies by industry and device and is higher on mobile. The important caveat is that this headline number is misleading as a performance target. A large share of those carts belong to people browsing, comparing prices, checking delivery costs, or saving items for later — behaviour that was never going to convert in that session. The meaningful number is not your overall abandonment rate but your drop-off rate once someone has actually entered checkout.
Why do people abandon their carts?
Among people who genuinely intended to buy, unexpected extra costs are consistently the leading cause — shipping, taxes and fees appearing later than the shopper expected. Other recurring reasons include being forced to create an account, a checkout that's too long or complicated, not being able to see the total cost upfront, delivery being too slow, concerns about entering card details, website errors, an unsatisfactory returns policy, and too few payment options. Most of these are fixable; the underlying pattern is unwelcome surprises late in the process.
Does guest checkout reduce cart abandonment?
Yes — forcing account creation is one of the most reliably cited reasons for abandoning an otherwise completed purchase, and offering guest checkout is among the highest-return changes most stores can make. The productive compromise is to let people buy as guests and then offer account creation after the order is confirmed, when you can pre-fill their details and the value to them is obvious. You get the account without risking the sale.
Is one-page checkout better than multi-step?
Neither wins universally, and the debate matters less than practitioners suggest. What consistently helps is reducing the total number of fields, showing clear progress so people know how much remains, and avoiding surprises at any stage. A well-designed multi-step checkout with visible progress often outperforms a cramped single page, while a clean single page can beat a poorly signposted multi-step flow. Test it on your own traffic rather than adopting someone else's conclusion.
How do you measure checkout performance?
Instrument each step of the flow separately — cart viewed, checkout started, contact details completed, shipping selected, payment entered, order confirmed — and calculate the conversion rate between consecutive steps. The single overall abandonment rate tells you almost nothing actionable, whereas a step-by-step view shows exactly where people leave. Segment by device as well, since mobile drop-off patterns typically differ significantly from desktop and often reveal the largest recoverable losses.
