How to audit checkout abandonment step by step

When a store owner tells us checkout conversion is "bad", the first thing we ask is where exactly buyers leave. Almost nobody knows. Analytics tools report a single abandonment rate, usually somewhere between 65 and 80 percent, and that number is close to useless on its own. A shopper who adds a product to the cart to check shipping costs is not the same as a shopper who types a card number, gets declined and gives up. Treating them as one population leads to redesigns that change a lot and fix little. This is the audit sequence we run before anyone opens a design file.

Instrument every step before you judge any of them

You cannot prioritize what you cannot see. The goal of the first week is a step-by-step funnel with real numbers, split by device and traffic source. On most platforms that means firing distinct events for each meaningful moment rather than relying on page views, because single-page and one-page checkouts hide the transitions.

• Cart viewed and checkout started, so you can separate browsing carts from buying intent.
• Contact submitted, shipping method selected and payment submitted as separate events.
• Payment failed, with the gateway's decline category attached, not just a generic error.
• Order completed, deduplicated by order ID so page refreshes do not inflate it.

If your analytics setup cannot do this cleanly, fix that first. A proper GA4 and analytics setup usually takes a few days and pays for itself on the first decision it informs. Collect at least two full weeks of data, and more if your store does fewer than a few hundred orders a week, so that a single promotion or a slow Monday does not skew the picture.

Read the funnel in the right order

Once the data is in, resist the urge to start at the top. Start at the bottom, where intent is highest and losses are most expensive.

Payment failures

Look at the ratio of payment submitted to order completed. On a healthy store this sits around 92 to 96 percent. We regularly see stores at 85 percent or lower, and the causes are usually mundane: an overly strict fraud rule, a missing 3-D Secure flow for European cards, or a wallet button that fails silently on certain browsers. Every point recovered here is revenue from people who already decided to buy.

Shipping and total cost surprises

Next, compare checkout started with shipping method selected. A sharp drop here almost always means buyers are seeing the real total for the first time and not liking it. The fix is rarely cheaper shipping. It is showing the cost earlier: an estimate on the product page, a threshold message in the cart, or a clear delivery date instead of a vague range.

Contact and account friction

Finally, look at the entry to checkout. Forced account creation, a coupon field that sends people off to hunt for codes, and long address forms without autocomplete all show up here. These are real losses, but they are also the ones most stores already know about.

A shopper who adds a product to the cart to check shipping is not the same as one whose card is declined. Treat them as one population and you will redesign the wrong thing.

Watch sessions with a question in mind

Quantitative data tells you where. Session recordings and a handful of moderated tests tell you why. The mistake is watching recordings aimlessly. Pick the step with the largest absolute loss and filter recordings to sessions that ended there. Twenty sessions is usually enough to see the pattern. In a recent audit for a DTC apparel brand, the recordings showed mobile users tapping the shipping selector repeatedly because the radio buttons had a 14-pixel hit area. The fix took an afternoon and lifted that step by four points.

Keep a simple log as you watch: timestamp, step, what the user tried to do, what happened. After twenty sessions you will have a ranked list of observed problems instead of opinions.

Prioritize by revenue, not by effort

With the funnel and the observations side by side, estimate the value of each fix. The math does not need to be precise. Take the number of sessions reaching the step each month, multiply by a conservative recovery rate, and multiply by average order value. A fix that recovers one percent of payment failures on a store doing 4,000 orders a month at a 90 dollar average is worth more than a full cart redesign that might lift add-to-checkout by a fraction of a point.

1. Fix payment failures and silent errors first. They are cheap and the buyers are already committed.
2. Move total-cost information earlier in the journey.
3. Remove or defer non-essential fields and account creation.
4. Only then consider layout and visual changes to the checkout itself.

On Shopify, much of this now lives in checkout extensions and payment customizations rather than theme code, which is a good constraint. Our checkout extensibility work usually focuses on exactly these items: delivery date messaging, conditional payment methods and trust content placed where hesitation shows up. On other platforms, a careful payment gateway integration review often uncovers the biggest single leak.

Validate the fix, then move on

Each change should ship with a clear expected effect on one step of the funnel. If you changed the shipping estimate, watch the checkout-started-to-shipping-selected ratio for two weeks. If it moved, keep going. If it did not, revert or rethink before layering on more changes, because stacking five simultaneous edits makes it impossible to know what worked. Stores with enough traffic can run a proper split test; smaller stores can use before-and-after comparisons, as long as they account for seasonality and promotions.

The whole audit, from instrumentation to the first round of fixes, typically takes four to six weeks. The result is not a prettier checkout. It is a checkout where you know which step is costing you money and by roughly how much, which is a far better foundation for any conversion rate optimization program that follows.

Want a second pair of eyes on your checkout?

We run checkout audits for Shopify, WooCommerce and custom stores, and we quote them at a fixed price before any work starts. Tell us about your store and you will have a scoped quote within 24 hours.