She has been on the site for eleven minutes, and the jacket is already in her cart. Size medium. Her card is on the desk next to her phone. She taps through to checkout and starts filling it in. Email, name, address. Then the order summary updates, and a shipping line appears that she has not seen before: $14.95.
She scrolls back up to read it again. Nothing on the product page mentioned $14.95. She opens a new tab, searches the brand name and the words returns policy, and lands on a page that says returns are accepted within 30 days. It does not say who pays the postage.
She goes back to the checkout tab. Her thumb sits over the card field for a second or two. Then she closes it.
Nothing on that site was broken. She was not price shopping and she was not just browsing. She wanted the jacket, she had her card out, and she still left, because she reached the end of the checkout still carrying a question, and leaving was the cheaper way to answer it.
Here is the same forty seconds described from the other side of the screen, by a founder:
"We're at 86 grand a month across Meta and TikTok. Checkout sessions are up 31 percent since we rebuilt the product page. But only 44 percent of the people who start checkout finish it, and that number hasn't moved in nine months."
Read the last number again. If 44 percent finish, more than half of the people who start his checkout do not. He is paying to send them there, the product page is working well enough to grow that traffic by 31 percent, and the money is going out at the final screen.
Why Does A Buyer With Her Card Out Still Walk Away?
The framework I keep coming back to for this is the LIFT Model, built by Chris Goward at WiderFunnel in 2009. It reads a page as six factors. The Value Proposition is the thing being weighed, and it gets carried by Relevance, Clarity and Urgency. Two factors pull against it: Anxiety and Distraction.
Anxiety is the one that owns a checkout, and Goward's definition is more specific than the everyday word suggests. He defines it as a function of the credibility you have built with your visitor and the trust you are asking them to have in you.
That is the part worth holding onto. Anxiety is not a weak argument, it is an unanswered question. You do not beat it with a better offer. You beat it by answering the question in the place the buyer is asking it.
If the weighting of one framework feels like a matter of taste, there is a second one that arrived at the same place independently. MECLABS built a Conversion Sequence Heuristic, C = 4m + 3v + 2(i - f) - 2a, where a is anxiety and it carries a minus sign. Two different CRO lineages, both putting anxiety on the negative side of the equation. That agreement is the point, not the arithmetic.
Baymard Institute reports the average documented cart abandonment rate at 70.22 percent, calculated across 50 studies conducted from 2006 to 2025. That is the cart-level figure, measured across the whole journey rather than at the checkout stage.
A number that stable over nearly twenty years is not a run of bad luck, it is structural. The thing driving it is what I call the Anxiety Anchor: the last unanswered question a buyer carries into checkout. It is what they leave on, and it is almost never the price itself.
What Is She Actually Asking At That Screen?
Baymard also asked people who abandoned during checkout why they did it. Set aside the 42 percent who said they were simply browsing and not ready to buy, and what is left is the part a site can do something about. These are self-reported reasons from buyers who abandoned during checkout, and respondents could pick more than one, so they do not add up to 100.
Extra costs too high, at 40 percent, which is the shipping reveal: a cost that was never mentioned upstream becoming real on the one screen where the buyer has already spent effort.
Delivery too slow, at 20 percent, although most checkouts I audit never state a date at all, so what the buyer ends up rejecting is her own worst guess at it.
Not trusting the site with credit card information, at 19 percent, which is a judgment about you formed on the exact screen where you ask for the most sensitive thing you will ever ask for.
Returns policy not satisfactory, at 13 percent, and the version I see far more often is a policy that is perfectly fair but never says who pays the postage, so the buyer prices in the worst case.
Could not see or calculate the total order cost up front, at 12 percent, where the buyer is not refusing to pay the number, she is refusing to agree to a number she cannot see yet.
Every one of those is the same shape. The buyer has a question, the answer exists somewhere in the business, and it is not on the screen where the question arrived. Under LIFT they all sit on the same side of the equation. None of them are really price objections. They are questions the page left open.
Where Does Your Returns Policy Actually Live?
When a founder hears this, the audit that usually follows is a checklist of existence. Do we have a returns policy? Yes. Do we have trust badges? Yes, three of them. Do we publish shipping rates? Yes, there is a page for it. Everything on the list is present, the checklist comes back clean, and nothing changes.
The more useful question is a question about location. When the doubt arrives, where is the buyer standing, and is the answer on that screen with her. The answer belongs where the doubt appears, not where the policy lives.
That turns a content problem into a placement problem, and placement is the cheaper of the two to fix. You are not writing a new returns policy. You are moving one sentence of it to within a thumb's reach of the button she is about to press.
Run This Audit Before You Touch The Checkout Design
Two lists, one afternoon.
First, open your helpdesk and read the last 60 days of inbound. Gorgias, Zendesk, Shopify Inbox, live chat, Instagram DMs, wherever your buyers actually write to you. Pull 30 pre-purchase messages, meaning people who had not bought yet, and tally what they asked. You are not looking for the clever questions. You are looking for the one that keeps repeating.
Then run your own checkout on a phone with a real card and mark the exact screen where each of the questions from the first list goes unanswered.
If the messages come back as post-purchase "where is my order", you pulled the wrong window. Filter to pre-purchase only. If nothing surfaces there either, the doubt is upstream on the product page, and checkout is not where your money is leaking.
Anxiety is the cheapest thing to remove at checkout. It costs you a line of copy. Every other lever you have costs margin.
This Week's Free Resource
I put together a free six-slide deck, The Checkout Doubt Map, showing exactly where the cost, delivery, returns and security answers go on a product page and a cart page. It also covers the badge-soup anti-pattern, which is why three generic trust seals underperform one specific answer.
Thanks for reading!
Conal Brady
Ecom Luxx
P.S. Learn more about Ecom Luxx and our GrowthOps™ Partnership below:
