The Complete Guide to Cart Abandonment

Cart abandonment is the single biggest leak in e-commerce, and most store owners have never seen the real number for their own store broken down by cause. This guide pulls together the actual published research on why shoppers abandon carts, what it costs a typical South African store, and a complete, prioritised playbook for fixing it, section by section, rather than guessing at a redesign.

What cart abandonment actually is, and why the average number is misleading

Cart abandonment is any case where a shopper adds a product to their cart and leaves without completing the purchase. The commonly cited global average, according to the Baymard Institute‘s aggregation of 50 separate studies, is 70.22%. That means for every 100 people who add something to a cart, roughly 70 never pay for it.

That single number hides more than it reveals. Abandonment rates vary enormously by industry: grocery and pet care sit closer to 50 to 61%, apparel is higher, and travel, finance, and B2B commonly run above 80%, since these are considered, comparison-heavy purchases rather than impulse buys. The number also varies enormously by device, with mobile abandonment consistently running higher than desktop.

South Africa’s own numbers run above the global average. According to World Wide Worx’s 2026 Online Retail in South Africa report (produced with Mastercard, Peach Payments and Ask Afrika, drawing on 23,910 consumer interviews and Stats SA retail data), the local online retail market reached an estimated R159 billion in 2026, growing 22.5% and crossing 10% of total retail turnover for the first time. That growth is real and accelerating, but it also means more South African stores are now competing for shoppers who have every reason to hesitate: unfamiliar checkout flows, unclear delivery timelines, and card-not-present fraud concerns that are more acute in emerging payment markets. The report specifically flags that “fast authentication, dependable payment processing and a checkout that works well on a smartphone” are now decisive for revenue, not optional polish.

The real reasons shoppers abandon their cart

Baymard’s research is the most rigorously documented source on this question, because it separates people who were never going to buy (just browsing) from people who had real intent to purchase and were stopped by something specific. Once you exclude the roughly 42% who were simply browsing, the remaining reasons, in order, are:

• Extra costs too high (shipping, tax, fees): 40%
• Delivery was too slow: 20%
• Didn’t trust the site with card details: 19%
• Site required account creation: 18%
• Checkout was too long or complicated: 17%
• Website errors or crashes: 17%
• Returns policy was unsatisfactory: 13%
• Couldn’t see or calculate the total cost upfront: 12%
• Card was declined: 10%
• Insufficient payment methods: 9%

Every one of these is fixable without a full redesign, and most of them are fixable without spending anything beyond a developer’s time. That is the encouraging part: cart abandonment is not primarily a traffic problem or a pricing problem. It is a long list of small, specific trust and friction gaps, most of which nobody has ever audited end to end.

What abandonment is actually costing your store

The maths is simple once you have your own numbers. Take your monthly sessions that reach checkout, multiply by your abandonment rate, then multiply by your average order value. A store with 5,000 checkout sessions a month, an 80% abandonment rate, and a R900 average order value is losing roughly R3.6 million a year in carts that were started and never finished. Even recovering a modest 10 percentage points off that rate, without spending a single rand more on traffic, is worth R450,000 a year to that same store.

The problem is that most store owners have never measured their own abandonment rate broken down by stage and by cause. They know the industry average exists, but not where their own checkout specifically loses people, which means every fix is a guess. If you want to skip the guessing, run a free scan of your store and get a prioritised list of exactly where your checkout is leaking, instead of applying every tactic in this guide at once and hoping something sticks.

The complete fix playbook, by cause

1. Extra costs (the single biggest cause, at 40%)

The fix is not to eliminate shipping costs. It is to stop hiding them. Show a shipping estimate on the product page or the cart page, before the shopper has invested time filling in a checkout form, using a postcode-based calculator or a flat-rate estimate. If you cannot show an exact figure that early, show a realistic range. The goal is that nobody reaches the final payment screen and sees a number they have not already mentally budgeted for.

2. Slow or unclear delivery (20%)

State an estimated delivery date, not just a courier name, as early as the product page. “Arrives by Thursday” converts better than “Delivered via courier” because it answers the question the shopper is actually asking: will I have this in time. If delivery genuinely varies by region, show a simple postcode-based estimate rather than a vague range that covers every possibility.

3. Trust and security concerns (19%)

This is rarely fixed by adding more badges. It is fixed by removing the specific things that make a store look unfinished: a missing SSL padlock, inconsistent branding between the product pages and the checkout, a checkout hosted on an unfamiliar domain with no explanation, or the complete absence of a visible returns policy, contact details, or customer reviews. Trust is accumulated from many small, consistent signals, not manufactured by a single badge in the footer.

4. Forced account creation (18%)

Always offer a genuine guest checkout option, positioned at least as prominently as the account-creation option, ideally more prominently. You can still invite the shopper to create an account after the purchase is complete, when the cost of saying yes is zero and the benefit (saved details for next time) is obvious.

5. Long or complicated checkout (17%)

Audit your checkout form field by field and ask, for each one, whether the order can be fulfilled without it. Fields that exist “because we might want that data one day” belong in a post-purchase survey, not in the path to payment. Address autocomplete, sensible input types (a numeric keypad for phone numbers on mobile), and clear inline validation all reduce the perceived length of a form even when the field count stays the same.

6. Website errors or crashes (17%)

This is the one cause that has nothing to do with psychology and everything to do with testing. Test your checkout on the actual devices and browsers your traffic uses, not just your own laptop, and test it after every plugin, theme, or payment-gateway update, not just at launch. A checkout that worked perfectly for six months can break silently after a routine update nobody thought to re-test.

7. Unsatisfactory returns policy (13%)

Show your returns policy on the product page, not buried in a footer link or only visible after purchase. A shopper deciding whether to buy something they cannot try on or inspect in person is running a risk calculation, and an unclear or hard-to-find returns policy reads as a bigger risk than it probably is.

8. Total cost not visible upfront (12%)

Wherever technically possible, show a running total, including estimated shipping and any mandatory fees, from the cart page onward, not only at the final checkout step. Shoppers tolerate a cost they can see coming. They abandon a cost that appears to jump at the last second, even when the final number is the same either way.

9. Declined cards and insufficient payment methods (10% and 9%)

In the South African market specifically, offering only card payments leaves out a meaningful share of shoppers who prefer instant EFT, a digital wallet, or a buy-now-pay-later option. The World Wide Worx report notes that pay-by-bank services, instant transfers, and BNPL products now carry “meaningful checkout volume” in the local market, not a marginal share. If your checkout only supports one payment rail, you are structurally excluding shoppers who would otherwise have paid.

Recovering the carts you couldn’t prevent

Even a well-optimised checkout will not get every abandoner to convert on the first visit. A cart recovery sequence, email at minimum and SMS or WhatsApp where you have consent, recovers a meaningful share of these sessions if it is timely and specific (referencing the actual product left in the cart, not a generic reminder). This is a large enough topic to deserve its own dedicated guide rather than a summary here.

How to measure whether your fixes are actually working

Track abandonment rate by funnel stage (product page to cart, cart to checkout start, checkout start to payment), not just as one blended number, since a single blended rate hides exactly which stage improved and which one didn’t. Change one variable at a time wherever practical, and give each change at least one to two weeks of real traffic before judging it, since day-to-day noise in a small store’s numbers can easily be mistaken for a real trend.

A quick self-audit checklist

• Is a realistic shipping cost visible before the final checkout step?
• Is an estimated delivery date shown on the product page?
• Is guest checkout available and at least as visible as account creation?
• Does the checkout form only ask for fields genuinely required to fulfil the order?
• Has the full checkout flow been tested on a real mobile device this month?
• Is the returns policy visible on the product page, not just in the footer?
• Does the cart page show a running total including estimated shipping?
• Are local payment methods beyond card (EFT, wallets, BNPL) available?

Your fastest next step

Every fix in this guide matters, but they do not all matter equally for your specific store, and applying all nine at once makes it impossible to know which one moved the number. The fastest way to find out which of these gaps actually exists on your checkout, ranked by how much revenue each one is likely costing you, is to run a free scan of your store. It checks your live site against this exact list and gives you a prioritised report instead of a generic checklist to work through blind.

Related guides

• The Complete Guide to E-commerce Checkout Optimization
• The State of E-commerce Conversion in South Africa (2026)
• The Complete Guide to Trust Signals That Actually Increase Conversions
• The Complete Guide to Mobile Checkout Optimization

Frequently asked questions

What is a good cart abandonment rate?

Anything meaningfully below the 70.22% global average (per Baymard) is good, but the more useful benchmark is your own store’s trend over time, since acceptable rates vary enormously by industry and by how considered the purchase is.

Is South Africa’s cart abandonment rate higher than average?

Industry benchmarking data has put South African abandonment rates above the global average in recent years, consistent with a market where mobile checkout friction, payment method availability, and delivery trust concerns are more pronounced than in more mature e-commerce markets.

Can I fix cart abandonment without a redesign?

Yes. Most of the highest-impact fixes (showing shipping costs earlier, adding a guest checkout option, showing a delivery estimate) are configuration or copy changes, not structural redesigns.