GA4 funnel exploration for e-commerce: step by step

Check your ecommerce events first

A funnel can only show steps GA4 receives. Google says its recommended ecommerce events are not sent automatically, because they need extra context to be meaningful. Your store, theme or tag setup has to send them. Each step below is one event.

Source: Google Analytics Help: Recommended events and the developer reference for recommended events, accessed October 31, 2026. The developer reference also marks currency and value as required, with a note that currency is required whenever you set value.

Two checks before you build anything. First, confirm that purchase carries a transaction_id. Second, compare GA4 purchases with your store's orders for the same dates and time zone. If GA4 shows more, the event may be firing more than once. Other causes of gaps are in our conversion rate drop guide.

Build the funnel in GA4, step by step

Menu labels below come from Google's funnel exploration help page and its checkout funnel example. GA4's interface changes, so if a label has moved, look for the nearest match.

1. In the left navigation, click Explore.
2. Under Start a new exploration, select the Funnel exploration template.
3. In Settings, remove any pre-filled steps you do not need with the icon next to each step. Then click Edit next to Steps (the help page calls it the edit pencil).
4. Give step 1 a clear name, such as "View product". Open the Add new condition drop-down and choose view_item.
5. Click Add step, name it, and repeat with add_to_cart, begin_checkout, add_payment_info and purchase. A funnel can have up to 10 steps.
6. For each step after the first, choose how it follows the previous one. is indirectly followed by allows other actions in between. is directly followed by means the step must come immediately after. Use Within to set a time limit.
7. Click Apply.
8. Drag Device category from Dimensions into Breakdown (add it in Variables first if it is not listed). Mobile and desktop rates differ a lot, so one blended funnel hides the picture. See checkout completion rate by device. Google notes that a user is counted only under the first breakdown value that applies, so a shopper who starts on mobile and buys on desktop shows up under mobile in every step.

Then decide open or closed with the Make open funnel toggle. In a closed funnel, users must enter at step 1. In an open funnel, users can enter at any step. Google's own checkout funnel example suggests an open funnel when a shopper adds to cart in one session and comes back later to buy.

Optional settings: Show elapsed time shows the average time between steps. Next action fills a "Top 5 Next Actions" card for each step. Visualization switches between a standard (stepped) and a trended (line chart) funnel. You can apply up to 4 segments.

Save it as a funnel report

Open the exploration and click Save as a report in the Library (top right). Only Editors and Administrators see the button. The saved report keeps your steps, filters, funnel type and breakdown, but not elapsed time or next action. A trended funnel cannot be saved, and neither can one with segments applied, more than 5 filters, or a dimension, metric or filter that reports do not support. Source: Google Analytics Help: Create a custom funnel report.

Limits to know

Source: Google Analytics Help, Funnel exploration, Segment builder, Get started with Explorations, Create a custom funnel report and Data retention, accessed October 31, 2026. The retention setting only affects explorations and funnel reports. Large properties are limited to 2 months.

Google also says explorations are subject to data thresholds to protect user privacy, for example when they include demographics or Google signals data. When sampling or thresholds apply, the icon in the right corner of the exploration turns from green to yellow, and its tooltip explains why. Check it before you trust a small row.

Funnel exploration counts users, not sessions

Google's funnel exploration page talks about users throughout and never mentions sessions. A user is counted only in the steps completed in the specified order, and only the first sequence per user is reported. A shopper who visits three times and buys on the third visit is one user who completed the funnel. The Purchase journey report (Reports, then Monetization under the Life cycle collection) also counts users, per Google.

Littledata's benchmarks work the other way. Each of its rates is a share of sessions.

Source: Littledata average website performance, 436 Shopify stores with a connected GA4 property, 90 days from July 27 to October 24, 2026, figures generated October 28, 2026, accessed October 31, 2026. Conversion rate also on its conversion rate benchmark page.

Count sessions instead

To get rates you can set beside the benchmarks, count sessions with a free form exploration and session segments. Google's segment builder offers session segments, which it describes as subsets of the sessions that occurred on your site or app.

1. In Explore, start a Free form exploration.
2. In Variables, add the Sessions metric and drag it into Values. Add Device category as a dimension to split the rows.
3. Click the plus symbol beside Segments, create a custom segment, and choose the session type.
4. Click Add new condition, choose the event add_to_cart, and set the scope to Within the same session. Name the segment, then click Apply.
5. Repeat for begin_checkout and purchase. Then build two segments with two conditions joined by AND, still within the same session: add_to_cart and purchase, and begin_checkout and purchase. An exploration holds up to 10 segments, but you can apply only 4 at a time, so swap them in and out.
6. Read Sessions for each segment, then divide.

Compare with the benchmarks

Use the same window and device split as the benchmark. Littledata uses 90 days. It publishes mobile and desktop medians for add-to-cart rate, conversion rate and checkout completion rate, but not for cart abandonment. Then read the pages that cover each step: add-to-cart rate, checkout completion rate and cart abandonment rate. Our Checkout Leak Calculator has a Google Analytics 4 option. You enter sessions, sessions with add_to_cart, sessions with begin_checkout, purchases and average order value, and it compares each step with the median Shopify store.

Common problems

• A step shows zero. The event is not being sent, or it is sent under a different name. Check the event list for the exact names above.
• GA4 purchases exceed store orders. The purchase event may be firing more than once, for example when a confirmation page reloads. Check transaction_id for repeats.
• The numbers may be sampled. Google says explorations may use sampled data when a query includes more than 10 million events. The icon in the right corner turns yellow when that happens. Shorten the date range or add a filter.
• You cannot go back far enough. On standard properties, event data retention is 2 months by default and can be raised to 14 months. Explorations cannot see older data.

Frequently asked questions

Sources

1. Google Analytics Help: Funnel exploration. Menu labels, steps, open and closed funnels, breakdown attribution, visualization, users counted. Accessed October 31, 2026.
2. Google Analytics Help: Ecommerce exploration solutions. Checkout funnel example, menu labels for editing steps, open funnel suggestion. Accessed October 31, 2026.
3. Google Analytics Help: Create a custom funnel report. Saving an exploration as a report, what is not saved. Accessed October 31, 2026.
4. Google Analytics Help: Get started with Explorations. Explore navigation, limits, default retention, sampling, data thresholds. Accessed October 31, 2026.
5. Google Analytics Help: Segment builder. Session segments, condition scope, segment limits. Accessed October 31, 2026.
6. Google Analytics Help: Data retention. 2 and 14 month options for standard properties. Accessed October 31, 2026.
7. Google Analytics Help: Recommended events. Ecommerce event descriptions, not sent automatically. Accessed October 31, 2026.
8. Google for Developers: Recommended events. Required parameters for ecommerce events. Accessed October 31, 2026.
9. Google Analytics Help: Purchase journey report. Menu path, five steps, users counted. Accessed October 31, 2026.
10. Littledata: Shopify and Ecommerce Benchmarks (2026). Metric definitions, medians, sample of 436 stores. Accessed October 31, 2026.
11. Littledata: Average Ecommerce and Shopify Conversion Rate (2026). Conversion rate definition, sample criteria, date range, mobile and desktop medians. Accessed October 31, 2026.

Related

• Shopify add-to-cart rate by categoryThe first step of the funnel, with the median and the best 20%.
• What is a good checkout completion rate?The last step, by percentile, device and industry.
• Why did my Shopify conversion rate drop?A step-by-step way to find which part of the funnel moved.
• Checkout Leak CalculatorEstimate what your funnel gaps cost each month. Free, no sign-up.