Abandoned shopping trolleys and the checkout: where exactly do orders fall by the wayside?

An abandoned cart is not one problem but the sum of several: unexpected costs at the last step, mandatory registration, a missing preferred payment method and too many fields to fill in. The practical approach is not to chase the overall abandonment rate, but to measure the drop-off across three separate transitions: from product page to cart, from cart to checkout, from checkout to payment. The biggest gap between two adjacent stages shows where to work first. Then you check which of the four causes applies to the specific store. Free delivery is one of the possible measures, not a default solution. Winning back the people who did not finish is a separate task and does not replace fixing the actual reason for the drop-off.

My name is Hristiyan Panov — a performance marketing consultant with over 5 years’ experience and over €2 000 000 Managed advertising budgets for over 90 companies on Google and Meta. Below is an explanation of how to interpret the conversion funnel and which adjustments yield results most quickly.

How do you measure leakage correctly?

Step by step, not overall. Four events are needed: product view, add to cart, checkout start, purchase. The ratios between them show where the hole is. The overall abandonment rate does not point to a place to work; the three ratios do.

Measure unique users, not events. One person who adds to the cart three times counts as one. Look at the same period for all four steps and split the data by device and by source, because mobile checkout and paid traffic usually have a different drop-off profile. UTM parameters must survive to the last step, otherwise the purchase is attributed elsewhere.

If these events are missing or duplicated, every calculation after them is a guess. The typical defects are easy to recognise: an add-to-cart that fires on page load; a checkout start that is recorded again at every step of a multi-step form; a purchase that is counted again when the thank-you page is reloaded. That is why the tracking check comes before the checkout work.

What are the most common reasons for rejection?

• Delivery cost shown only at the very end. The most common reason in almost every shop I visit.
• Registration is compulsory. Ordering as a guest is one of the few changes with an almost guaranteed effect.
• Payment method missing. In Bulgaria, cash on delivery still accounts for a significant proportion of orders among certain customer groups.
• Far too many fields. Any field that you do not need to fulfil the order is a waste.
• Lack of trust. No contact details, return policy or visible telephone number.
• Uncertain delivery time. „Up to 5 working days“ is better than saying nothing at all.

What should you fix first?

Sorted by the ratio of effect to effort:

• Show the delivery price already in the cart, and even better, on the product page. If the price depends on the address, provide a calculator or a range instead of silence until the last step.
• Allow ordering without registration. Registration can remain as a tick box after the order is confirmed, when the details have already been entered.
• Remove every field you do not use. The test is simple: if the field does not go into the waybill, the invoice or the delivery message, it is redundant.
• Show the payment methods and delivery times before the last step. Card icons and cash on delivery tell the person that their way of paying is available, before they have got that far.
• Put a visible contact on every step of checkout. A phone number and opening hours, not just a message form.

Before you touch anything, record the current ratios between the steps. Without a baseline there is no way to tell whether the fix worked.

Does free delivery work?

Usually, yes, but it isn’t free for you. Treat it as a reduction in your margin and set it above a threshold that increases the average order value. The threshold should be chosen based on your current average order value, not on what your neighbour is doing.

Look at the distribution of orders, not just the average. If most orders cluster just below the chosen threshold, it is too high and people give up instead of adding more. If the majority is already above it, you are giving away delivery that they would have paid for anyway. A useful threshold sits slightly above the bulk of orders: enough that adding one more item is worthwhile, but not so much that it requires a second purchase.

Communicate the threshold where it matters — in the cart and at the start of checkout, with the amount remaining to reach it. A change in margin also changes your break-even threshold, so recalculate it every time you reduce the margin with delivery, a discount or a promotion.

How do people who haven’t completed their studies return?

Using three channels, in this order: email or SMS reminders, retargeting with the specific product via catalogue advertisement, and finally a discount — only if the margin allows it.

The discount is deliberately offered last. If you offer it first, you’ll get people into the habit of waiting for a discount before they order.

What should you avoid doing?

Don’t test five changes at once — you won’t know which one made the difference. Don’t compare yourself with industry averages; Compare your sales funnel to how it was a month ago. And don’t tinker with the checkout in the middle of peak season, when the cost of a mistake is highest.

FAQ — Frequently Asked Questions about Abandoned Trolleys

What percentage of dropouts is considered normal?

Average figures are of little use because they depend entirely on the product and the price. The only useful comparison is with your own history.

How many reminder emails should I send?

Usually two or three, with the first one being sent within a few hours. More than three start to become annoying and damage the sender’s reputation.

Does a pop-up window appear when you try to leave the page?

Sometimes, when traffic is slow. For people who are already in the checkout, it tends to get in the way more often than it helps.

A single-page or multi-step checkout?

The number of fields is more important than the number of screens they’re spread across. The multi-step approach provides better data on exactly where people drop out.

How can I tell whether the problem is with the checkout or with the network?

Compare the conversion funnel by source. If organic traffic is performing well but paid traffic is stalling at the same stages, the problem lies in the expectations created by the advert.

Conclusion

The checkout is where your most valuable traffic slips away unnoticed. The difference between a well-functioning and a poorly functioning sales funnel usually lies not in the design, but in a few decisions regarding transparency — price, delivery time and how much trust customers place in you.

If you are paying for traffic that isn’t converting into sales, start by getting an independent review.

Request a quote for a digital audit →

Related articles: website speed and conversions · catalogue adverts · ROAS threshold