You can spend ages polishing product pages, fine-tuning ads and hunting for new traffic channels — but all of it is worth little if the customer turns back at the checkout.
This is an especially painful loss, because at this point we're no longer talking about an uncertain browser. The visitor has chosen a product, added it to the cart, and in many cases already entered their shipping or billing details. Their intent to buy is far stronger than on an average product page visit. Checkout optimization is therefore anything but a minor UX detail you can get to next year.
Based on the Baymard Institute's aggregation of more than fifty cart abandonment studies, the average online cart abandonment rate sits slightly above 70 percent. At first glance that's an alarming number, but it needs to be read correctly. A large share of cart abandoners simply aren't ready to buy yet. In Baymard's survey, 42 percent said they were just browsing or didn't actually intend to purchase. Among the remaining reasons, however, there are plenty that the webshop can directly influence.
Too-high extra costs were cited by 40 percent of respondents, 20 percent turned back because of slow delivery, 19 percent didn't trust the site enough to enter their card details, and 18 percent were put off by mandatory registration. A checkout process that was too long or complicated was a problem for 17 percent. Further shoppers left because they couldn't see the total amount payable up front, or couldn't find the payment method they wanted to use.
These numbers show clearly that checkout optimization is rarely about one big overhaul — in most cases, it's about removing several smaller obstacles.
It's not the checkout's beauty that decides whether people finish buying
When designing a checkout, it's easy to slip into perfecting the visual appearance above all. A tidy, professional interface matters, of course, but at this point most shoppers aren't looking for spectacular design — they want to get through the process as fast as possible.
The root cause can be many things. A redundant field that's annoying to fill in on a phone. A shipping option with an incomprehensible name. A handling fee that only shows up at the last step, or simply the feeling that the webshop is asking for more information than it genuinely needs. Individually, these little things don't necessarily look like serious problems. Stacked one after another, though, they can create just enough uncertainty for the customer to say “I'll come back later” — and in many cases, they never do.
A one-step checkout isn't necessarily better
In recent years, the one-step checkout has become almost a marketing concept of its own. It's easy to assume that if everything fits on one screen, buying becomes simpler and the happy customer will hastily toss two more products into the cart. Usability research, however, paints a more nuanced picture.
According to Baymard's checkout research, the number of steps matters less in itself than the total number of fields, questions and decisions the shopper has to deal with. A well-structured two- or three-step checkout is often easier to use than a single long, crowded page. And the best part: this isn't just a theoretical distinction — one of the biggest marketing dogmas of recent years has officially fallen.
Based on Baymard's current benchmark data, an efficient checkout can be built from 12–14 form elements, including as few as 7–8 actual data fields. The US webshops they examined averaged considerably more: close to 15 fields. So the question isn't how to cram the same long form onto one page — it's whether every piece of requested data is truly necessary.
If a piece of information isn't needed to fulfil the order, it has no place in the checkout. Typical examples are the unnecessary second phone number, automatically displayed company fields, date of birth, and various marketing questions that put the customer's patience through a thorough stress test.
You have to decide: is the registration or the transaction more important?
From the merchant's perspective, the desire for registered users is entirely understandable. A customer with an account returns more easily, reordering is simpler for them, and the webshop gains more customer data. But it doesn't follow that registration should be made a condition of purchase.
In Baymard's survey, 18 percent of respondents abandoned an order because the site demanded account creation. What's more, according to the institute's 2025 checkout benchmark, 62 percent of the webshops examined still failed to make the guest checkout option properly clear. For a new customer, “Register to continue” means something completely different than it does to the merchant. The webshop sees extra data and a longer-term relationship; the user sees another task, another password and another commitment they simply don't want to make.
That's why many have concluded it's wiser to let the purchase finish first and only then invite the user to create an account. Since most of the required data is already available after the order confirmation, a single password may be all it takes — and the guest shopper becomes a registered customer without ever putting the order at risk.
Keep the shipping cost clean and clear
One of the most sensitive points of the checkout is the final total — and it's also one of the most common blunders. If the customer saw a HUF 19,990 price on the product page and only discovers in the checkout that with shipping and other fees they'll actually pay around HUF 24,000, they can easily feel the webshop promised one thing and delivered another. It's no accident that too-high extra costs top Baymard's list: 40 percent of non-browsing respondents abandoned for this reason. Another 12 percent said they turned back because they couldn't see or calculate the order's total cost in advance.
A webshop doesn't necessarily have to show a to-the-forint shipping fee on every product page. In many cases it depends on the address, the parcel size or the delivery method. But the expected costs should be clarified as early as possible. If shipping is free above HUF 20,000, for example, that shouldn't be announced on the checkout's final screen. If shipping starts from HUF 1,590, that too can be useful information before the checkout even begins.
And at the end of the checkout, no mental arithmetic should be required. The customer must clearly see the value of the products, any discount, the shipping cost and the final total. All of it spelled out plainly, without convoluted labels or exceptions to the exceptions. In fact, this isn't just a conversion issue — European consumer protection rules also require the total price and related costs to be communicated clearly.
For shipping, the arrival time matters more than the service's name
As an e-merchant, it's easy to assume everyone has a formed opinion about every courier company. But the customer rarely cares what internal product name a given service runs under — what they really want to know is when the parcel arrives, where they can have it delivered, and how much it costs.
“3–4 business days” can also raise questions in the user's mind: do they deliver on Saturday, does that count as the weekend? It's far clearer to communicate something like: “Home delivery – expected arrival: September 25–26 – HUF 1,490.”
Baymard's usability studies reached the same conclusion: a concrete or estimated delivery date is easier to interpret than generic labels like “standard” or “2–3 day” shipping. In the 2025 benchmark, 48 percent of the webshops examined still handled this poorly.
You don't need many payment methods — you need the right ones for your users
Card payment has effectively become a baseline requirement, but that doesn't mean every shopper is looking for the same payment method. What feels natural to a twenty-something mobile shopper may differ completely from a corporate buyer or someone who regularly orders cash on delivery.
In 2025, Stripe ran a large-scale randomized experiment examining how different payment options affect conversion. In transactions where at least one relevant payment method was displayed beyond cards, they measured on average 7.4 percent higher conversion and 12 percent higher revenue. The study covered more than fifty payment methods and shoppers in over two hundred countries.
This result shouldn't be read as three new payment buttons automatically lifting every webshop's orders by 7.4 percent. The checkout needs to work well with the payment methods your target market's shoppers actually use. Too many options can clutter the interface just as surely as too few can scare the customer away.
For a Hungarian webshop, the place to look first is therefore your own order data. What's the share of card payments? How many choose cash on delivery? Is there a payment option many people select but that fails more often than average? These tell you more than any generic international top list.
On mobile, users punish you even harder
Completing a checkout on a phone is an entirely different experience than on a desktop.
On a laptop, fixing a mistyped postcode is an almost invisible operation. On a phone, the keyboard pops up, covers part of the screen, you have to find the right field again, then scroll back to continue.
That's why something only mildly inconvenient on desktop can easily become a serious obstacle on mobile.
In Baymard's 2025 benchmark, 63 percent of the mobile checkouts examined received only a mediocre rating or worse. That's notable because mobile commerce stopped being a secondary traffic channel a long time ago. On mobile, improvements that look trivial at first glance count for a great deal:
Show a numeric keyboard for the phone number field.
For the email address, a keyboard where the @ symbol is easy to reach.
Make sure the browser's autofill works.
Make the fields large enough to be comfortably used with a finger.
Baymard's research also found multi-column forms problematic. On mobile, it's especially easy to misread what order two side-by-side fields belong in. A simple, single-column form layout running top to bottom is generally less ambiguous.
Trust doesn't grow from five padlocks next to the pay button
A sense of security in the checkout matters, of course. In Baymard's research, 19 percent of shoppers abandoned an order because they didn't trust the site enough to enter their card details. Many webshops respond with visual fixes: shields, padlock icons, “100% secure” labels and assorted logos crowd around the payment fields.
On their own, these don't solve the trust problem.
What counts far more is the quality of the webshop as a whole. Make it clear which business the customer is buying from. Don't hide the company details. Make the return and shipping process understandable. And the name of the payment provider shouldn't be a mystery either.
If the customer is redirected to another page for card payment, it's worth signalling that in advance. For the user, even an unexpected domain change can feel suspicious — even when the service behind it is technically perfectly secure.
Don't shove the coupon field in the customer's face either
Coupons can be a good sales tool, but in the checkout they can easily achieve the opposite. Picture the customer who has already decided to spend HUF 25,000 on something. At the checkout, a big, prominent field greets them: Coupon code!
Except they don't have a coupon. From there it's only one step to opening a new browser tab and searching whether there's a 10 percent discount somewhere. They might find an old coupon aggregator site, an expired offer, or even a competitor's webshop. Or end up on a lifestyle magazine's site, where the autumn fashion pieces steal their attention entirely.
Not every checkout problem calls for a redesign
When optimizing, it's tempting to change everything at once: a new layout, a shorter form, a different shipping module, new payment options and new button copy all shipping in the same update. If conversion improves, great — but we'll have no idea what actually worked.
It's better to test concrete hypotheses. If we see many mobile shoppers leaving at the data entry stage, a logical hypothesis is that the form is too long. Then the redundant fields can be removed and the results compared. If the drop-off happens at shipping selection, it's more likely worth looking at the prices, the delivery options or the expected delivery dates.
A/B testing is truly useful when we're not just pitting two designs against each other but seeking the answer to a clear question — and it's wise to measure more than checkout conversion alone. A change might bring more orders while significantly increasing the number of unclaimed cash-on-delivery parcels. Another tweak might lift conversion while the average basket value drops.
The ultimate goal isn't the maximum number of “Place order” clicks, but orders that can actually be fulfilled — and that turn a profit at the end of the day.