Roughly seven in ten online shopping carts are abandoned, and the reasons are unglamorous. People do not leave because the checkout was insufficiently beautiful. They leave because something appeared that they did not expect, or because completing the purchase asked more of them than they were willing to give. Nearly every improvement worth making follows from those two sentences.
The single largest cause is surprise cost
Shipping, taxes and fees revealed at the final step are consistently the biggest driver of abandonment. The reason is not the money — it is the sequence. A shopper who has mentally committed at one price and is shown another feels the change as a bait-and-switch, even when the total is perfectly reasonable.
The fix is to show costs as early as you possibly can. State shipping on the product page. Put a threshold for free delivery where it is visible before the cart. If tax varies by destination, say so before checkout rather than adding it silently at the end. A shopper who sees the real total early either accepts it or leaves early — and leaving early costs you far less than an abandoned checkout.
Do not force account creation
Requiring registration before purchase is the second reliable killer, and it exists for the shop’s benefit rather than the customer’s. Guest checkout should be the default path, with account creation offered after the order is placed, when it is genuinely useful — the customer now has an order to track, and creating an account costs them one password rather than one decision.
What a good checkout does
| Practice | Why it matters |
|---|---|
| Ask for as little as possible | Every optional field is a chance to reconsider. If you do not act on it, do not ask for it |
| Show progress | People will finish something with a visible end; an open-ended form feels endless |
| Validate as they type | An error revealed only on submit means re-reading the whole form |
| Keep the cart editable | Forcing someone out of checkout to change a quantity is where they leave |
| Offer the payment methods they use | Varies enormously by market; the wrong set is an invisible barrier |
| Never clear the cart | A cart that survives a closed tab or a switch to a phone recovers sales by itself |
| Show delivery dates, not durations | “Arrives Thursday” answers the question. “3–5 business days” makes them do arithmetic |
Design it for a thumb
Most browsing and an increasing share of purchasing happens on a phone, where a checkout designed at a desk becomes hostile. The specific failures repeat: fields too small to tap accurately, forms that do not summon the right keyboard for a number, a card field that fights autofill, and a total that scrolls out of sight at the moment of decision.
Two things fix most of it. Set the correct input types so the numeric keypad appears for numbers and postcodes, and support the platform wallets, which replace the entire form with one authentication. A checkout that can be completed with a fingerprint removes almost every reason to abandon.
Trust is a checkout feature
People part with card details on the basis of signals they read in about a second. A visible returns policy, a real contact route, a delivery estimate, and the absence of anything that looks improvised all contribute. So does restraint — countdown timers, artificial stock warnings and pop-ups arriving mid-checkout read as pressure, and pressure reads as untrustworthy. The most effective trust signal is a checkout that looks like it belongs to a business that expects you to come back.
Recovering what you lose anyway
Some abandonment is unavoidable, and a proportion of it is recoverable. A single well-timed reminder to someone who entered an email address recovers a meaningful share; a sequence of increasingly urgent ones mostly generates unsubscribes. Capture the email early in the flow, send one useful message, and include a direct link back to the populated cart rather than the homepage.
Frequently asked questions
Why do people abandon shopping carts?
Predominantly unexpected costs revealed late, followed by forced account creation, checkouts that are too long, and payment methods they do not use.
Should I require customers to create an account?
No. Offer guest checkout and invite account creation after the order, when it has an obvious benefit to them.
How many steps should a checkout have?
As few as the order genuinely requires. What matters more than the count is that progress is visible and nothing unexpected appears late.
Do trust badges increase conversions?
Less than a visible returns policy, a real contact route and a clear delivery date. Badges help most where they are recognised; substance helps everywhere.
Should I show shipping costs before checkout?
Yes, as early as possible. Late shipping costs are the single most cited reason for abandonment.
Are abandoned cart emails worth sending?
One is. A long escalating sequence produces unsubscribes rather than orders, and it needs to link back to the populated cart.
Related reading
See also what makes a homepage work and why the mobile version is the one that matters.
