Most abandoned carts are lost to surprise costs, forced accounts and slow forms, not indecision. A step-by-step audit of your checkout flow, plus how to set up abandoned cart recovery that actually earns sales back.
Cart abandonment optimization means finding the exact step where buyers leave your checkout and removing the reason they left. In practice the biggest wins come from four fixes: show the full price early, allow guest checkout, cut the form to the fields you truly need, and offer express payment. Then add recovery emails to win back part of what still slips away.
Baymard Institute's running average puts cart abandonment at roughly 70 percent. Some of that is window shopping you cannot fix. The rest is friction you created, and friction is cheap to remove.
You cannot optimize a checkout you have not measured. Set up a funnel with these events, in this order:
Google Analytics 4 has standard ecommerce events for most of these (add_to_cart, begin_checkout, add_shipping_info, add_payment_info, purchase). Shopify, WooCommerce and most hosted carts can send them for you. Our analytics setup guide walks through the basics.
Look for the biggest drop between two neighbouring steps. That step is where you start. A store losing most people between "checkout started" and "shipping details entered" has a form or cost problem. A store losing them at payment has a trust or payment-method problem. The fix differs, so do not guess.
Also watch real sessions. A tool like Microsoft Clarity or Hotjar will show you people rage-tapping a coupon field or scrolling up and down looking for the shipping cost. Ten recordings usually teach you more than a week of dashboards.
Unexpected extra costs are the reason shoppers cite most often for leaving. The problem is not the shipping price itself, it is learning the price at the last step.
Fixes, from easiest to hardest:
Taxes are harder to show early in regions where they depend on the address, but you can still say "taxes calculated at checkout" in the cart so nobody feels ambushed.
Every field is a small chance to quit. Go through your checkout and ask of each field: do we need this to ship the order or charge the card?
| Field | Keep? | Notes |
|---|---|---|
| Yes | Needed for receipts and cart recovery. Put it first. | |
| Full name | Yes | One field is fine; split only if your carrier needs it. |
| Company name | Usually no | Make optional or hide behind a link. |
| Address line 2 | Collapse | Show as "Add apartment, suite" link. |
| Phone | Depends | Only if the courier requires it. Say why. |
| Billing address | Default to shipping | A single "same as shipping" checkbox, ticked by default. |
| Create password | No | Move to the thank-you page. |
Then add the helpers:
Express wallets let a returning shopper pay in a couple of taps because their address and card are already stored. At minimum, offer:
Put express buttons at the top of checkout and in the cart, not only on the payment step. Check current fees with each provider before adding them, because they vary by region and change.
A shopper who has never heard of you is about to type a card number. Help them.
Most store traffic is on phones, and a checkout that is merely fine on desktop is often painful on a small screen. Test your own checkout on a mid-range Android phone over mobile data, not on office Wi-Fi. Look for:
Remove every pop-up from checkout. An email capture modal on the payment step is one of the most self-defeating things a store can do.
Once the checkout is clean, recover part of what still leaves. Because email is the first field, you can reach anyone who got past it.
A simple, effective sequence:
| Timing | Message | Discount? |
|---|---|---|
| 1 to 4 hours | "You left something behind," with product image and a one-click link back to the filled cart | No |
| About 24 hours | Answer objections: shipping time, returns, reviews, a real person to contact | No |
| 3 days (optional) | Last reminder, stock or price note if true | Small, if margins allow |
Rules that separate good recovery from spam:
SMS recovery works too, but only with explicit opt-in, and it should be one message, not three.
Go through your own store as a new customer on your phone and tick off:
Any unticked box is your next task. Change one thing at a time and compare the step-to-step drop in your funnel for a couple of weeks before moving on, so you know which change helped.
If you are building a new store, pick a setup where these basics come built in rather than bolted on. With We.Inc you can describe the store you want in chat and get an online store with product pages, cart and checkout, then adjust the flow visually or in code, with hosting and SSL included. Our guide to making a website to sell stuff covers the wider setup.
Baymard Institute's long-running average, compiled from dozens of studies, sits at roughly 70 percent. Many of those shoppers were only browsing, so the realistic goal is not zero. It is removing the reasons people who wanted to buy still left.
The most cited reasons in shopper surveys are extra costs shown late (shipping, taxes, fees), being forced to create an account, a checkout that feels long or complicated, not trusting the site with card details, and slow delivery. All of these are fixable on your side.
It works in three ways: it shortens the form (address autocomplete, saved details, express wallets like Apple Pay and Google Pay), it shows you where people drop off with step-level analytics, and it recovers carts after the fact with timed reminder emails or texts. The software only helps if the underlying causes, like hidden fees, are fixed too.
Yes. Let people buy with just an email, then offer to create an account on the confirmation page using the details they already typed. You keep the customer record without putting a password wall in front of the payment.
A common, sensible pattern is two or three: one within a few hours as a plain reminder, one about a day later answering objections (returns, shipping, reviews), and an optional last one a few days later. Only add a discount to the final message, or you train customers to abandon on purpose.
How we research, test and update this page: our editorial policy. We.Inc is our own product.
We.Inc is an AI-powered website builder you can resell under your own brand. Launch a branded client dashboard, bill on Stripe Connect, and deliver AI-generated websites in minutes. White-label plans start at $99 a month for 25 client sites, with a 7-day free trial and no per-site fees.
