Abandoned Carts: Why Customers Leave Without Buying

A customer reaches your website, looks at a product, chooses a variation and adds it to the cart. For the person running the store, the hardest part seems almost finished. Then the order never arrives. The immediate response is often to find more traffic, increase the advertising budget or prepare a discount code.

Before doing that, I would pause to ask what happened between interest in the product and payment. Abandoned ecommerce carts can tell very different stories. Some shoppers are comparing options. Others genuinely want to buy but encounter an unexpected charge, an unclear requirement or a step that does not work.

Understanding that difference helps us decide where to act. A campaign can bring interested people to a store; if the buying journey makes things difficult, sending more visitors into it can amplify the same problem. Let us look at cart abandonment and checkout improvement in practical terms, with an approach that matters before Black Friday and throughout the year.

Cart abandonment and checkout abandonment: what are we measuring?

A cart is created when someone adds a product. Checkout is the process of completing the order, providing the necessary information and choosing delivery and payment. A shopper can leave before starting checkout or stop after beginning that process. These are different points in the journey.

That distinction matters because the questions change. If few shoppers move from the cart to checkout, I would look at the information available in the cart and the expectations set earlier. If many start entering their details but stop at payment, I would investigate those later steps. A single overall abandonment rate cannot show us which explanation fits.

Before choosing a solution, we need a clear definition of what the store records as a cart and a checkout start. Otherwise, a dashboard can look precise while grouping together situations that call for different decisions.

Cart abandonment benchmarks need context

The Baymard Institute overview, updated on 28 October 2026, reports an average cart abandonment rate of 70.22%, compiled from 50 studies published between 2006 and 2025. This is not a new measurement of Italian stores. Baymard also reports that 42% of surveyed US shoppers had abandoned a cart because they were browsing or were not ready to buy.

A cart can serve as a reminder, a comparison tool or a temporary shopping list. Treating every unfinished cart as a sale we should have won creates unrealistic expectations. That does not mean ignoring abandonment. It means distinguishing ordinary exploration from avoidable obstacles for people who were trying to purchase.

I would use an industry benchmark to frame the question, then examine the store's own history, products, devices and acquisition channels. An expensive product that requires consideration will not necessarily behave like a familiar item bought repeatedly.

Before trying to recover every cart, I would find out which customers were actually trying to buy and what stopped them.

Make the real price clear before payment

Imagine a product advertised at €39, with €8 in shipping charges appearing only at the end. This is a hypothetical example, but the decision is straightforward: the customer is now evaluating a different total from the one that first caught their attention. The delivery charge may be reasonable; discovering it late can still change the purchase decision.

Where possible, delivery costs or a clear way to estimate them should be available before the final payment stage. The same applies to promotion conditions, minimum spend requirements and differences between product variations. A discount should remain understandable when quantities or delivery destinations change.

Delivery information deserves the same attention. A product being available is not the same as it arriving by a particular date. If someone needs it for an event, a vague promise can leave too much uncertainty. I would prefer a realistic delivery estimate and clear conditions over a reassuring claim that the store cannot reliably meet.

The aim is to help shoppers understand what they will pay, what they will receive and when they can reasonably expect it. These are purchasing questions, not details to hide until someone has already invested time filling out a form.

Accounts and forms: ask for what the purchase needs

An account can make repeat purchases easier. For a first purchase, however, being asked to register and create a password adds a separate task. Where the business model allows it, guest checkout gives shoppers a way to buy without making account creation a condition of the order.

I would review each required field and ask what purpose it serves at that moment. Information needed for a business invoice does not automatically belong in every consumer checkout. Optional fields should look optional, and explanations should appear where a person might reasonably need them.

Error messages are part of the form experience. “Invalid information” leaves the customer to investigate; a message that identifies the field and explains the expected format helps them continue. Correcting one entry should not erase everything already completed.

This is not an invitation to remove information required to fulfil an order or to weaken security. It is a way to separate necessary requirements from requests that have accumulated over time without a clear reason.

Check trust and mobile usability in the actual buying journey

Before paying a store they do not know, shoppers need enough information to judge it. Who is selling the product? How can they get help? What happens if delivery goes wrong or they need to return something? Clear contact, delivery and returns information can answer questions that a collection of generic trust badges cannot.

Reviews should be genuine and relevant. The objective is to give customers useful evidence, not to decorate the checkout with reassurance that cannot be checked. Important information also needs to remain easy to find at the point where a person is making the decision.

I would complete the journey on a phone, from choosing a product variation to returning after a payment attempt. A keyboard may cover a button, a popup may hide the total, or an address field may be awkward to complete. Problems that seem minor on a desktop can become obstacles when someone is buying with one hand on a small screen.

Webita's free SEO analysis tool can provide an initial view of aspects such as metadata, page structure and performance. A checkout still needs its own functional checks: an SEO report alone cannot confirm that a customer can complete a purchase.

When payment is the obstacle

A failed payment is not the same as a customer deciding that an offer is unsuitable. The Stripe documentation on payment declines explains different reasons a payment may fail. That is why I would examine payment outcomes alongside the behaviour recorded in the store.

Repeated attempts also need careful interpretation. Five failed attempts by one customer do not necessarily represent five different people or five lost orders. Understanding which shoppers and purchases are involved makes it easier to assess the size of the problem.

Where possible, the customer should receive a useful explanation and a practical next step, such as checking their details or trying another supported method. Their cart should remain available. The solution should preserve the provider's security and fraud controls rather than bypass them to make a number look better.

Payment methods should fit the audience, market and economics of the store. Adding every available option creates its own costs and complexity. I would start by checking whether customers can use the methods they reasonably expect and whether those methods work reliably.

Measure abandonment without confusing events and people

Google's GA4 ecommerce documentation includes events such as add_to_cart, begin_checkout, add_shipping_info, add_payment_info and purchase. These events help describe the journey, but event totals alone do not automatically give us a reliable abandonment rate.

A person can add several items, revisit checkout or make repeated payment attempts. Dividing unrelated event counts can therefore produce a misleading result. We need a consistent unit of analysis, whether carts, sessions or users, and a coherent observation window that allows us to connect the relevant steps.

Consider a hypothetical matched group of 1,000 distinct carts. If 650 begin checkout and 400 become orders, overall cart abandonment is 60%. Among the 650 that reached checkout, 250 did not complete it: checkout abandonment is about 38.5%. Both figures describe the same example, but they answer different questions.

I would also reconcile recorded purchases with the orders in the store's management system. Duplicate events, missing purchases, consent choices, changes of device and customers returning later can limit what analytics tells us. A measurement system should make those limits visible rather than turn incomplete tracking into certainty.

A better rate does not automatically mean a better business result

Abandonment changes with the visitors we attract. A campaign that reaches people earlier in their decision may bring more exploratory carts. A campaign aimed only at customers already ready to buy may improve the rate while reaching fewer people. The percentage needs to be read alongside completed orders and acquisition activity.

Discounts add another complication. Two months could have the same acquisition cost, while the second generates more orders through deeper promotions. If discounts and delivery subsidies consume the extra margin, the higher order count does not by itself demonstrate a better result.

I would distinguish the potential value of abandoned carts, completed recovered orders and the economic contribution of those orders. Discounts, shipping and other relevant costs belong in that assessment. The amount sitting in an unfinished cart is not revenue the business has already earned.

Cart recovery emails: understand the reason before offering an incentive

As the Shopify guide to abandoned cart emails explains, these messages can remind shoppers of an unfinished purchase and help them return to their cart. Whether that reminder helps depends on why the shopper left.

If delivery timing was unclear, useful information or access to support may be more relevant than a discount. If payment failed, fixing the payment issue comes first. An email cannot compensate for a checkout that presents the same obstacle when the customer returns.

I would keep the message focused: recognizable products, a clear route back to the cart and a way to get help. Links should avoid exposing personal details. The sequence should respect applicable consent requirements and customer preferences, stop after purchase and account for stock availability and message frequency.

Automatically discounting every abandoned cart can reduce margin and encourage shoppers to wait for the next incentive. Where traffic allows, comparing a recovery sequence with a suitable control group can help distinguish additional purchases from orders that would have happened anyway. A shopper clicking an email and buying does not prove that the email created the entire sale.

What to check before Black Friday

Black Friday falls on 27 November 2026, making this a timely moment to review the journey. More traffic puts promotions, stock, delivery and payment under pressure together. I would rather arrive with a tested checkout and reserve the final days for focused checks.

The main scenarios should reflect the store: a mobile order, a product with variations, a promotion code, a delivery address with different conditions and a payment that fails on the first attempt. Each check needs an expected outcome so that problems can be recognized before a customer reports them.

For payments, I would use the provider's test environment. Any real transaction checks should be planned and authorized, with control over orders, charges and stock movements. Improvised tests can create operational problems of their own.

I would also compare the promise in the advertisement with the product page and cart. A promotion can work technically while still confusing people if they cannot find the conditions they expected. The customer experiences one journey; difficulties often appear between the teams responsible for campaigns, content and the store.

Where I would start to improve ecommerce conversions

I would begin with a few changes supported by clear evidence. An unexpected cost, a reproducible error or a confusing requirement gives us a specific problem to investigate. For each change, I would record what we are altering, which step should improve and what result we will observe.

With enough traffic, a controlled test can help compare alternatives. For smaller stores, I would examine comparable periods and include customer feedback and direct observation, without treating two extra orders as definitive proof. Keeping a record also prevents the team from losing track of what has changed.

This work connects website design and marketing. Through Webita Site, we can examine and develop the store journey, from product pages to integrations. With Webita marketing consulting, we connect those priorities with campaigns, the offer and commercial goals.

Before paying to bring more people to checkout, I would make sure those already there can buy without avoidable obstacles.

Not every cart will become an order. We can still make the offer easier to understand, the journey easier to complete and the results easier to interpret. That is where I would start improving ecommerce conversions: understand where customers stop, address the reason and check that the improvement benefits the business as well.