People put an item in the cart and left. The shop owner’s first thought: “The ads bring the wrong people.” Sometimes that is true. But more often the person reached the cart, which means the ads did their job, and something on the site pushed them away afterwards. So before you change campaigns or add budget, walk the buyer’s path yourself.
What “70 %” really means
The Baymard Institute, which studies the usability of online stores, collects dozens of cart abandonment studies in one table. As of 2025–2026 the average there is about 70 % (70.19 % and 70.22 % in different versions of the table, from 50 studies). In other words, roughly seven shoppers in ten do not complete the purchase.
Two caveats so the number does not scare you. First, it is a worldwide average over many years, not your own rate. You will see yours in your analytics. Second, not every abandoned cart is a lost sale. In the same table, 42 % of surveyed US shoppers admitted they abandoned a cart because they were “just browsing” and not ready to buy. For many people the cart is a bookmark or a way to add up the total.
Why people leave: what shoppers themselves say
Baymard asks shoppers what made them leave checkout (excluding the “just browsing” group). The most common answers:
- extra costs (shipping, tax, fees) were too high — 40 %;
- delivery was too slow — 20 %;
- they did not trust the site with their card details — 19 %;
- the site wanted them to create an account — 18 %;
- checkout was too long or complicated — 17 %;
- site errors or crashes — 17 %;
- unsatisfactory returns policy — 13 %;
- they could not see the total cost up front — 12 %;
- the card was declined — 10 %;
- not enough payment methods — 9 %.
The percentages do not add up to a hundred: each person could name several reasons. Figures from other roundups may differ: depending on the country, year and sample, the shares shift by a few points. So read this as an order of magnitude, not an exact formula. The picture is stable, though: most people are put off by unexpected costs and extra steps, not by the fact that they saw your ad.
Three checks on your own site (20 minutes)
Take your phone, open the site the way a shopper does, add an item to the cart and go to the payment step. Do not enter a card, just look.
1. Can you buy without registering? If the site demands an account before payment, some people will leave. In Baymard’s data that is 18 % of answers. The simplest fix: a “Buy as a guest” button, or creating the account after payment, when the details are already entered.
2. Is the full price with delivery visible before payment? This is the most common reason, and another 12 % of shoppers complain they could not see the total up front. Check whether the product page and the cart show the delivery cost or at least the rule (“free above a certain amount”). If you show the delivery price only on the last step, the shopper feels led to the payment and leaves.
3. Do Apple Pay and Google Pay work? One-tap payment saves typing the card number and address. Baymard does not publish a separate figure for these wallets (only that 9 % named too few payment methods and 17 % a too long checkout), so this is our reasoning: fewer fields means fewer places where a person gives up. Check on two devices whether the button appears at all and whether the payment goes through. It is also the easiest way to notice a broken payment you may not know about.
Trust, returns and errors: what else matters
Beyond the three checks, there are three more items from Baymard’s list that an owner can judge without a developer. Trust: 19 % did not dare to enter card details on the site, so the payment step should clearly show the company name, contacts and terms. Returns: 13 % left because of an unsatisfactory returns policy, so show the return terms before payment rather than hiding them in the footer. Errors: 17 % named site crashes. Go through checkout in two browsers and on a phone: if a button does not respond somewhere or a promo code “breaks” the total, you will find it in five minutes.
“Everything on one page” is not always the answer
A common piece of advice among practitioners: make checkout a single page and conversion will grow. Baymard is more cautious. Their conclusion: one page is not always better than several steps, because the problems shoppers run into are usually not in the number of screens but in what they have to do at each step. They separately call the step count a “red herring”: the number of form fields matters more. In their 2024 research the average checkout had 5.1 steps and 11.3 form fields, although most sites need only about eight.
One nuance: these Baymard conclusions partly rest on older material (their article on one-page checkout came out back in 2011, and they later wrote about accordion-style checkouts). So it is the researchers’ position, not a fresh measurement on your audience. Practical takeaway: before rebuilding checkout, remove unnecessary fields (one “Name” field instead of two, a hidden “Address line 2”, the promo code behind a link), and then test.
How to see where exactly you lose people
If ecommerce is set up in your Google Analytics 4, look at how many people reach each step: checkout start (the begin_checkout event), shipping details (add_shipping_info), payment details (add_payment_info) and purchase (purchase). These event names are given in Google’s developer documentation. Where the percentage drops the most, look for the cause there. If you do not know whether it is set up, ask whoever built the site or the analytics; see also our Shopping page.
It also helps to simply ask: write to five recent buyers and five people who abandoned a cart (if they left an email or phone) and ask what they disliked. A direct answer from a shopper often tells you more than guessing from numbers.
When the ads are to blame after all
Do not pin everything on the cart. Tell two situations apart:
- few people reach the cart while clicks are plentiful: then look at the ads, the search terms and the pages they lead to;
- many people reach the cart and then vanish: then the problem is on the site, and more budget will only increase the number of people who leave.
For a few other places to look before you blame the ads, read Sales dropped: 7 places to check. And for why shoppers go to a competitor already at the product stage, see the article on price in Shopping ads.
Summary
- Roughly seven carts in ten are not completed on a worldwide average (Baymard), but some people were just browsing.
- The most common reason is unexpected extra costs: show the full price with delivery as early as possible.
- Let people buy without registering.
- Check that Apple Pay and Google Pay work on a phone.
- Remove unnecessary fields, not only steps; “one page” is not a guarantee by itself.
- Fix the cart first, then add budget.
How we set up advertising for online stores is in the Shopping and Performance Max sections.
Want a free audit of your Google Ads account? Send us a request — we will look at your account and tell you whether money is lost in the ads or already on the site.