- An online shop that gets visitors but few orders has a problem at one specific stage of the buying path. Find that stage first, then look for the cause.
- Check the source of your traffic before you change the shop. A campaign that does not match the page lowers the conversion rate on its own.
- The average documented cart abandonment rate is 70.22% (Baymard Institute). The biggest reason, extra costs that are too high, is named by 40% of shoppers.
- Treat every suspected cause as a hypothesis. Collect behavior evidence, then A/B test the fix and measure it in your own GA4.
An online shop does not convert when visitors drop off at one stage of the buying path: landing page, category, product page, cart or checkout. The cause can be traffic that does not match the page, unclear total cost, weak product information, missing trust or a hard checkout. Find the stage with the largest drop, check its likely causes, and test one fix at a time.
This page works like a diagnosis. It starts by separating traffic problems from shop problems, then follows the buying path stage by stage with one question per stage. A diagnosis table links each symptom in Google Analytics 4 to a likely cause and a first test. The end of the page shows how to turn suspects into decisions with behavior data and A/B tests.
First check: is it a traffic problem or a shop problem?
Before you change anything in the shop, find out whether the visitors are the right ones. A shop can look weak in the numbers when the real issue is that the campaign promises something the page does not deliver.
Open your analytics tool and compare your main traffic sources on three numbers: conversion rate, bounce or engagement, and order value. Then read the result like this:
- One source converts far worse than the others. Check the ad or link that feeds it. The wording, the offer and the landing page should match.
- Every source converts poorly. The problem is most likely in the shop itself, and the rest of this page applies.
- Visitors leave within seconds. Compare what they clicked with what they see.
- Returning visitors buy, new visitors do not. New visitors lack trust or information. Look at the first screen, the shipping information and the return terms.
If the source and the page do not match, fix that first. For the surrounding methods, see the guide to conversion optimization.
Find the stage where buyers drop off
The fastest way to a diagnosis is a funnel: the number of visitors who reach each step of the buying path, from the landing page to the purchase. In Google Analytics 4 you can build it in the Explore section as a funnel exploration. Use the standard e-commerce events: view_item_list, view_item, add_to_cart, begin_checkout and purchase.
Read the funnel from top to bottom and look for the step with the largest drop compared with the one before. Split the funnel by device, because a desktop shop that works well can fail on phones. Then match the weakest step to the table below. Use it as a starting point: each row names a likely cause that you confirm with data in the next steps.
| Stage | Symptom in GA4 | Likely cause | First test |
|---|---|---|---|
| Landing page or home | High share of single-page sessions, few clicks into categories | Message does not match the source, unclear offer, weak first screen | Rewrite the headline and the main call to action; test one variation |
| Category | Many view_item_list events, few view_item events | Hard to compare products, filters missing, slow page | Add clear filters or a sort order; test the product grid |
| Product page | Many view_item events, few add_to_cart events | Total price or delivery unclear, weak images, no reviews, slow on phones | Show shipping and delivery on the page; test the price block |
| Cart | Many add_to_cart events, few begin_checkout events | Surprise costs, no guest option, no clear next step | Show the total early; test the cart summary |
| Checkout | Many begin_checkout events, few purchase events | Forced account, too many fields, missing payment methods, errors | Offer guest checkout; test form length and payment order |
If the events are missing or look wrong, fix the measurement first. A funnel built on incomplete data sends you to the wrong page. The documentation on tracking setup check explains how to verify what arrives.
Product and category pages: the 6 most common causes
Most lost sales in a shop happen before the cart. These six causes appear again and again on product and category pages. Check them in this order:
- Price and shipping are unclear. A visitor who cannot see what the order will cost in total leaves to look elsewhere. Show shipping costs and delivery time on the product page, close to the price.
- The images do not answer questions. One small photo is not enough for most products. Show the product in use, show details, and show sizes or scale.
- There are no reviews. Visitors look for proof from other buyers. Place ratings near the title and let people read full reviews on the page.
- The page loads slowly. Heavy images and scripts delay the first screen, especially on a mobile connection. Measure on a phone connection.
- The mobile page is awkward. Check tap targets, readability and whether the buy button is reachable without searching.
- Search and filters fail. A visitor who types a product name and gets no result rarely tries again. Review the search terms with no results.
All six are hypotheses until you test them on your own shop. The page on product page optimization goes through the elements one by one.
Cart and checkout: why buyers leave at the last step
A visitor who has put something in the cart wanted to buy. When that person leaves at the last step, the reason is usually something the shop can control. The Baymard Institute reports an average documented cart abandonment rate of 70.22%, based on 50 studies. Its list of reasons (US shoppers, retrieved October 20, 2026; shoppers could name more than one) looks like this:
- Extra costs (shipping, tax, fees) too high: 40%
- Delivery too slow: 20%
- No trust in the shop with card details: 19%
- Forced to create an account: 18%
- Checkout too long or complicated: 17%
- Errors or crashes: 17%
- Return policy not satisfactory: 13%
- Total cost not visible up front: 12%
Source: Baymard Institute, 50 Cart Abandonment Rate Statistics, retrieved October 20, 2026. Shares are US shoppers who abandoned a cart and named the reason.
Cost and clarity head the list. Compare it with your own funnel: if the drop sits between cart and checkout, extra costs and a forced account are the first suspects. If it sits between checkout and purchase, look at form length, payment methods and errors.
The guide to abandoned cart best practices covers ten fixes and ways to bring buyers back. Checkout steps get their own, more detailed treatment in a separate guide on checkout optimization.
Trust: when visitors do not believe the shop
Even with the right product and price, a visitor may doubt that the shop is safe or reliable. Trust comes from many small signals, and one missing signal at the wrong moment can end a visit.
- Who you are. A named company, an address, a working contact route and an imprint that is easy to find.
- What happens after payment. Delivery times, tracking and return conditions in plain words, before the checkout.
- Proof from others. Real reviews with dates.
- Payment safety. Familiar payment methods and recognizable security notices next to the payment fields. Baymard lists missing trust with card details as a reason for abandonment (19%).
Place trust signals where the doubt arises, and test whether a given note helps your visitors. For badges and seals, see whether trust badges increase conversions.
From cause to fix: test the hypothesis
A diagnosis gives you suspects. To turn a suspect into a decision, collect evidence about behavior, then test the fix.
Step 1: watch behavior. Heatmaps and session recordings show where visitors click, how far they scroll and where they hesitate. Varify.io connects to Microsoft Clarity, Hotjar and Mouseflow, so you can filter recordings by experiment and variation. See the documentation for Microsoft Clarity, Hotjar and Mouseflow.
Step 2: get a second opinion. The AI CRO Audit takes the URL of a page, analyzes it and returns CRO insights, concrete recommendations and ten test ideas you can use directly as A/B tests.
Step 3: test the fix. Describe the change in plain language with Prompt-Based Experimentation (PBX). The AI builds the variation, writes the code and sets up the experiment, without a developer, and you can refine the variation as often as you like. See PBX examples. Then run it as an A/B test. The result is measured in your own GA4, so test numbers and analytics numbers match. Varify works without cookies. More on cookieless A/B testing.
Change one thing per test, so the result has one explanation. Before you start, estimate the runtime with the free test runtime calculator.
Diagnosis checklist
Work through the list in order. Stop at the first stage where you find a clear problem and fix that before you move on.
Measurement and traffic
- The purchase event arrives in your analytics tool and matches the orders in your shop system.
- Conversion rate is compared by source, device and new and returning visitors.
- Each main campaign leads to a page that repeats its promise.
Product and category pages
- Total price, shipping and delivery time are visible on the product page.
- Images show use, detail and scale.
- Reviews or other proof from buyers appear near the title.
- The page is fast and easy to use on a phone.
- Search returns a result for the most common terms.
Cart and checkout
- The total cost is visible before the last step.
- A guest checkout is available.
- Payment methods match what your visitors use, and the form has no needless fields.
- Return terms are easy to find.
Trust
- Company details and contact routes are easy to find.
- Security and return notes sit next to the buy button and the payment fields.
Next step
- Pick the weakest stage, write one hypothesis, and run one A/B test.
Turn your diagnosis into a tested fix.
Describe the change, let the AI build it, measure it in your own GA4 and see whether your shop converts better.