Cart abandonment: why shoppers leave right before they pay

MyDeko sells handmade soap-flower bouquets. Since March we’ve raised the price of the bestseller with the owner, rebuilt the website and rebuilt the ads. The store’s conversion rate jumped from 1.6% to 3.0%. And still: out of 100 people who add something to the cart, 56 start checkout and 18 buy. Fewer than one in five.

Each of those 82 people had already clicked an ad, read a product page and decided “I want this.” You paid for that visit. When they walk away from the cart, it doesn’t just push up your CPA, the cost per order. It drags down the conversion rate that everything else is calculated from, including how much you can afford to pay per click. Below: why shoppers abandon carts, where your store leaks, and what to fix before you spend more on ads.

What cart abandonment is, and how much of it is normal

Cart abandonment is a purchase a shopper started and didn’t finish: they added items to the cart, maybe began checkout, and left without paying. The cart abandonment rate is the share of all carts that end this way. Baymard Institute puts the average at around 70%, based on 50 studies.

Seventy percent sounds like a disaster. It isn’t. Many carts were never going to become orders. People add items to see the price with shipping, to compare with another store or to save them for payday. In the same Baymard Institute research, 42% of US online shoppers say they have abandoned a cart because they were “just browsing” and not ready to buy.

So the goal is never zero. It’s to separate people who didn’t want to buy from people who did and got stuck. The second group is money almost in the till.

Be careful comparing numbers. Every tool counts abandoned carts differently: one by sessions, another by users, a third only from the start of checkout. The 70% average is a rough figure pooled from many sources, not a bar for your store. What matters is how your number moves over time and at which step people drop off.

Why shoppers abandon carts: the 10 most common reasons

Take out the people who were just browsing, and almost every remaining reason sits on the store’s side. Ranked by Baymard Institute’s survey of US shoppers who abandoned an order during checkout in the previous three months:

1. Extra costs were too high (40%). Shipping, handling, payment fees.
2. Delivery was too slow (20%). Especially for gifts and anything needed now.
3. They didn’t trust the site with their card (19%). Unknown store, no reviews, no contact details, no clear company behind it.
4. The site wanted them to create an account (18%). For a first-time customer an account is a hurdle, not a perk.
5. Checkout was too long or complicated (17%). Too many fields, too many steps, no idea where you are.
6. The site crashed or showed an error (17%). A dead button, a pickup-point map that never loads, a form that won’t submit.
7. The returns policy wasn’t good enough (13%). Or they couldn’t find it.
8. They couldn’t see the total up front (12%). The final amount appears only on the last step.
9. The card was declined (10%). Payment gateway, bank verification, card limit.
10. Their payment method was missing (9%). Usually the one they’re used to from other stores.

This is US data, and shoppers elsewhere have their own habits: pickup points, bank transfers, cash on delivery. The logic still holds: more people are put off by a surprise on the last step and by friction in the form itself than by the price of the product.

Cart, checkout, payment: where exactly people drop off

Cart abandonment isn’t one number. It’s three different exits with different causes. Someone who leaves the cart hasn’t decided yet. Someone who leaves a half-filled checkout form had decided, and something stopped them. Someone who leaves at payment wanted to pay and couldn’t.

From cart to checkout

This is where most people leave, and most of them for legitimate reasons: they were looking, comparing, saving for later. The rest are put off by what they see in the cart for the first time: the shipping cost, no free-shipping option, the delivery date. Broken buttons belong here too, and a cart that “escapes” under the cookie banner on mobile.

In the checkout form

These are people who made up their minds. Every exit here is friction: forced sign-up, a required phone number for a parcel going to a pickup point, a pickup-point picker that won’t load on mobile, or validation that wipes the address you just typed. These are the most expensive losses.

At payment

The payment gateway, a declined card, verification in the banking app, a redirect back to the store that fails. You don’t see this exit in the store as such — you see it as “unpaid” orders or as sessions that never came back from the gateway. Check what happens when a payment fails: does the shopper get a second try, or a blank page?

Enter your store’s numbers and see where to start. Two percentages from GA4 or Clarity are enough.

The default values are a model. There’s no built-in norm for your industry, because there isn’t one.

How to measure the cart abandonment rate

Before you change anything, you need a funnel you trust. The formula is simple: cart abandonment rate = 1 − purchases ÷ sessions with an add-to-cart. If 1,000 sessions with a cart produce 250 purchases, the abandonment rate is 75%. For diagnosis, though, it’s more useful to split it by step.

GA4: ecommerce events

Google Analytics 4 has recommended ecommerce events that follow each other: add_to_cart, begin_checkout, add_shipping_info, add_payment_info and purchase. When your store sends all of them, you can build a funnel exploration in GA4’s Explore section and see how many people drop off between every two steps, separately for mobile and desktop, separately by traffic source.

A common problem: the store template only sends add_to_cart and purchase. Then you know people leave, but not where. Before you draw any conclusions, check in GA4 which events actually arrive.

Clarity: why they leave

GA4 tells you where, Microsoft Clarity shows you why. In session recordings you filter visits that went through the cart and never reached the thank-you page, and you watch. Clarity flags “dead clicks” (a click that gets no response), “rage clicks” (rapid repeated clicks on the same spot) and “quick backs” (someone leaves for another page and comes straight back). Ten recordings from the leaking step beat a week of reports.

A cart is not a conversion

Measuring the cart has one trap. When add-to-cart reaches Google Ads as a primary conversion, the campaigns learn to find people who like adding to carts. Buying is optional. At Papírnictví VojTech, 81% of reported conversions were just add-to-carts, and in March 2026 only 8% of the conversions the account reported were real purchases.

We measure the cart to know where people leave. We train campaigns only on purchases with a value.

The rule we use to set up tracking in every account

GA4, Google Ads and Meta will never see the same number of purchases. Each system uses different attribution and different windows. Why they never quite agree is covered in GA4 vs Google Ads conversions (and Meta): why the numbers never quite match. For a cart funnel it doesn’t matter. You compare steps within one tool, not tools with each other.

Calculator: what abandoned carts cost you

Enter how many carts you get a month, what share of them end in a purchase, your average order value and your margin. The calculator shows how many orders and how much margin a few percentage points of better completion bring — and how many extra ad visits you’d need to buy the same result.

It uses margin on revenue excluding shipping and assumes the average order value stays the same. It’s an order-of-magnitude estimate, not a promise — but usually enough to see whether fixing the form pays more than more ad budget. Download your numbers as a CSV on the left.

How to reduce cart abandonment: 10 checkout fixes

Roughly ordered from least work to most. Change one thing, watch it in the funnel for a week, then move on.

1. Show the total earlier. Put the shipping cost and the free-shipping threshold on the product page and in the cart. The most common reason to leave is a surprise, not the price.
2. Make guest checkout the default. Offer an account on the thank-you page. You can create it from the details they already entered in one click.
3. Drop the fields you don’t need. Baymard Institute counted more than 23 form elements in the average US checkout and puts the ideal at 12–14. Every extra field is a reason to leave.
4. Run through checkout on mobile. All of it, up to payment, with a real address and a pickup point. Not in an emulator — on a phone.
5. Test the pickup-point picker. A map that won’t load, a search that can’t find a small town, a window that won’t close. It’s a third-party component in the middle of your form — and when it fails, the order can’t be finished.
6. Put trust next to the button. Store rating, contact, the returns policy in one sentence. Not in the footer.
7. Offer the payments your customers know. Card, instant bank transfer and whatever is most common in your category. Choose by who buys from you, not by what your gateway happens to offer.
8. Remove distractions. Checkout doesn’t need the main menu, banners or links to the blog. Logo, steps, cart summary.
9. Show error messages at the field, in plain words, right away. “Enter your ZIP code as 12345,” not “The form contains errors” at the top of the page.
10. Never lose what they typed. Coming back from the payment gateway, switching shipping or a typo in the ZIP code must not wipe what the shopper already filled in.

Mobile and the Facebook in-app browser

One thing most guides skip. If your traffic comes from Meta ads, a large share of people don’t open your site in Chrome or Safari, but in the browser built into the Facebook or Instagram app. At MyDeko, Clarity showed 57% of all sessions over 30 days in the Facebook in-app browser, and 81% in mobile browsers overall. That browser behaves differently: with logins, with autofill and with redirects to the payment gateway and back.

So when you test checkout, test it the way most of your customers arrive: tap your own ad in the Facebook app and finish a purchase there. You won’t find a bug that most of your visitors see on an office desktop.

The cart as a place to sell

The cart isn’t only where people leave. It’s also where they add. MyDeko has five add-ons in the cart, from 19 to 591 CZK. A “thank you for careful packing” add-on is in 20% of orders, express shipping in 12%. The average order in September was 59% higher than in March — together with a price increase that can’t be separated from the add-ons. One condition: an add-on offers, it doesn’t block. A checkbox, not another step.

Once they’ve left: email and retargeting

Some people will leave however good your checkout is. For them there are two tools: abandoned-cart emails and retargeting ads.

Abandoned-cart emails

They only work for people whose email you have and who you’re allowed to contact. Watch out for built-in platform features: according to its help center, the abandoned-cart feature of Shoptet, a common Czech store platform, only collects data on registered customers who were logged in when they left the cart. For everyone else you need an email tool connected to the store, such as Ecomail, which supports abandoned-cart automation. At Klenoty Mahdal, the abandoned cart is one of the retention automations alongside the welcome series, abandoned product and post-purchase emails.

We wouldn’t put a discount in the first email. Remind them what’s in the cart, answer the most common objection (shipping, returns, delivery date) and give them one link back to the cart. A discount in the very first email teaches customers that walking away and waiting pays off.

Retargeting

Ads aimed at cart abandoners usually look great in reports. Part of that is an illusion: many of those people would have come back anyway, and the ad just claims their purchase. How many orders ads actually added is a question of incrementality, which we cover in detail in When 400% ROAS beats 500% ROAS. Keep cart retargeting on a small budget and a short window, and judge it by the store’s total revenue, not by the campaign’s ROAS.

Both are a second line of defense. Every cart you save by email, you first lost in checkout. Fixing checkout works on every customer; email only works on the ones whose address you have.

Data from our accounts

Three accounts where we worked on the cart: through tracking, the website or automation. Month-by-month details are in the case studies.

MyDeko also showed how price, website and ads affect each other. The conversion rate jumped in the very first month after the rebuild, to 4.0% in April, and has held up even with much more cold traffic from Meta. Meanwhile nobody has tuned the cart yet. That’s why it’s in the next section.

What we got wrong

Two things we’d do differently today. Both have to do with the cart.

We left the cart for last

At MyDeko, since March, we worked on three levers: price, website and ads. Margin after ad spend went from −25% in March to +37% in September. But today fewer than one in five people who add to the cart complete the order. Checkout sat in the queue behind bigger changes the whole time. The order was deliberate, because a better checkout can’t save an order that loses money. But it means that every month more than four in five people we paid to bring in leave from the cart. It’s next on the list.

For ten weeks we steered campaigns by carts

At Papírnictví VojTech, for ten weeks after taking over the account, we ran on tracking that counted add-to-cart as a purchase. April ended at a total ACoS of 13.7%, above the 12% target. We fixed the tracking on 11 May. Today we check what the account counts as a conversion on day one, before we touch a single campaign.

Checkout checklist

Go through it with a phone in your hand. The full 24-point checklist and a calculation with your numbers are on the left — no email, no form.

Selling on Shoptet and want to know where shoppers drop out of your checkout? Funnel setup, Clarity recordings and cart fixes are part of the Shoptet stores we build and tune. Or just send us an enquiry below.

FAQ

Jiří Kopejska

Co-founder, Marketing ASAP

Co-founded Marketing ASAP. Before adding money to ads, he checks Clarity for where people leave checkout — because a paid visit that ends in an abandoned cart is the most expensive visit a store gets.