More sales from the traffic you already have
The average checkout asks for almost twice as many form elements as it needs. Here is the exact sequence I walk through when an owner says people add to cart but never pay.
When an owner tells me “people add to cart but nobody buys”, I do not open the homepage, the ads account, or the theme editor. I add a product to the cart on my phone and I start counting taps. Checkout is the one place where every visitor has already said yes to the product, so every loss here is a loss of your warmest, most expensive traffic. This entry is the sequence I walk through, in order, with the published numbers that explain why each check earns its place.
Two published figures frame every checkout audit I do. Littledata’s benchmark of 2,800 ecommerce sites puts the average checkout completion rate for Shopify at 45%. Not the site conversion rate, the checkout completion rate: of the people who begin entering their details, more than half walk away. The best 20% of stores complete above 59%, the best 10% above 66%.
And Baymard Institute, after years of large scale checkout testing, estimates that the average large ecommerce site can gain a 35.26% increase in conversion rate through better checkout design alone. You will not capture all of that in an afternoon, but the direction of the estimate matters: the bottom of the funnel is usually the highest density of recoverable money on the whole site.
The first thing I do is the most mechanical: I count every input, dropdown, checkbox, and radio button a new customer sees by default, from cart to payment. Baymard’s benchmark gives the reference point, and it is brutal. The average US checkout flow contains 23.48 form elements shown by default. Their testing puts an ideal flow at 12 to 14 elements, which is 7 to 8 actual form fields. In other words, the average checkout asks for nearly twice what it needs, and 18% of US shoppers in their survey had abandoned an order specifically because the checkout was too long or complicated.
On Shopify the payment page itself is standardized, so the excess usually hides earlier: an optional company field nobody needs, a second address line shown by default, a phone number demanded with no explanation, a mandatory account. Every one of these gets a note in my audit file with the same question next to it: what would break if this disappeared?
Baymard’s abandonment survey is the clearest ranking of self inflicted wounds I know. Excluding people who were just browsing, the top reason for abandonment, named by 39% of US shoppers, is extra costs that were too high: shipping, tax, fees. Another 14% abandoned because they could not see or calculate the total order cost up front.
So the second check is a stopwatch question: how many seconds pass between a visitor wanting the product and knowing the full price of getting it? I add the product to the cart and watch for the moment the number changes. If shipping shows up as a surprise on the last step, the store is running a bait and switch it never intended, and the analytics will show it as a cliff between the shipping step and payment.
19% of shoppers in Baymard’s survey had abandoned an order because the site wanted them to create an account. This one takes ten seconds to check and is still worth its own line, because I keep finding stores where account creation is switched to required, sometimes by accident, sometimes because someone wanted to grow “membership”. Guest checkout first, account offer after payment. The order confirmation page is a fine place to ask, the doorway to payment is not.
The checkout inherits whatever mood the cart created, so I audit the handoff too. Can a visitor buy quickly from a collection page, or must they visit every product page? Does the cart open in place, or does it teleport them away from browsing? Is the checkout button visible when the cart is long, or does it scroll away? These sound like small mechanics, and individually they are, but they are also exactly the kind of thing that separates themes built by people who watch session recordings from themes that just look good in screenshots. If you are choosing or changing a theme, look for these mechanics as listed, built in features rather than app add ons. For example, Ultra’s feature list in the Shopify Theme Store ships quick buy, a slide out cart, and a sticky cart as standard checkboxes, which is the pattern I want: the boring conversion plumbing present out of the box, with no extra app weight slowing the store down.
The remaining big abandonment reasons in Baymard’s ranking are about nerve: 19% did not trust the site with their credit card information, 15% were put off by an unsatisfactory returns policy, and 10% left because there were not enough payment methods. My check here is a simple inventory of the screen where the card number gets typed. Are the payment logos real and current? Is the returns policy one tap away, written by a human? Are the express options a distracted phone user actually has, Shop Pay, Apple Pay, Google Pay, switched on?
Nobody abandons a checkout because it lacked beauty. They abandon it because it asked too much, cost more than promised, or felt like a stranger’s hand reaching for their wallet.
The output of this walk is a short ranked list, and the ranking rule is fixed: cost surprises first, forced friction second, missing trust third, cosmetics last. Then I compare the store’s checkout completion rate against the 45% average and watch the same number, and only that number, after each fix ships.
The reason this entry sits under Diagnosis rather than Experiments is that most of what I find at this stage does not need an A/B test to justify removal. A surprise fee reveal or an account wall is not a hypothesis, it is a leak with published research behind it. Fix those first. The experiments come later, once the checkout has stopped actively arguing with the people trying to pay you.
Littledata's benchmark of 2,800 ecommerce sites puts the average Shopify checkout completion rate at 45%. More than 59% would put you in the best 20% of stores, and more than 66% in the best 10%. If you complete under 40%, checkout is almost certainly your cheapest place to recover revenue.
Less than people think on standard plans, and that is partly good news: the payment step itself is standardized and fast. Most of the leaks I find live in what you do control, the cart, the information you ask for, the shipping options and their prices, the moment costs are revealed, and the trust signals around the flow.
Yes. In Baymard's survey of abandonment reasons, 19% of US shoppers had abandoned an order because the site wanted them to create an account. Let people pay first. You will get their email in the order anyway, and account creation can happen after the money.
Baymard's testing puts an ideal checkout at 12 to 14 form elements, which is 7 to 8 actual fields. If you cannot say why a field exists in one sentence, it is a candidate for deletion. Every extra field is one more place for a tired thumb to give up.