- A GA4 funnel exploration shows how many users move from one step to the next, such as product view to add to cart to checkout to purchase, and lets you break that down by device, channel or product.
- It only works if your ecommerce events fire reliably and once per action. Check the events before you trust the funnel, especially checkout steps.
- Use the funnel to find the biggest drop between two steps, then use breakdowns, next actions and session recordings to learn why. The funnel shows where shoppers leave, not the reason.
What is a GA4 funnel exploration and when should you use it?
A funnel exploration is one of the templates in the Explore section of GA4. You define a sequence of steps, and GA4 shows how many users completed each step and how many dropped out before the next one. Google’s help center says you can define up to 10 steps.
For an online store, the obvious funnel is the path to purchase: view a product, add to cart, start checkout, add shipping, add payment, buy. You can also build funnels for narrower questions, such as whether shoppers who use site search reach checkout more often than those who browse collections.
It is the measurement half of funnel optimization; the fix comes later.
Which events do you need before building an ecommerce funnel?
A funnel is only as good as the events behind it. GA4’s recommended ecommerce events map neatly onto funnel steps.
| Funnel step | GA4 event | What to check before you trust it |
|---|---|---|
| Lands on site | session_start or page_view |
Fires once per session, consent setup understood |
| Views a collection | view_item_list |
Fires on collection and search results pages |
| Views a product | view_item |
Fires once per product page view, not on every variant click |
| Adds to cart | add_to_cart |
Fires from the product page, quick add and cart drawer |
| Starts checkout | begin_checkout |
Fires when checkout opens, including express buttons if possible |
| Adds shipping details | add_shipping_info |
Present at all; many setups skip it |
| Adds payment details | add_payment_info |
Present at all; many setups skip it |
| Buys | purchase |
Fires once per order, with a transaction ID |
On Shopify, checkout events depend on how your tracking is installed, so confirm in GA4’s real-time or DebugView reports that each step fires. Duplicate or missing events create impossible jumps or false drop-offs. Our guide to GA4 ecommerce tracking covers the full event tracking setup.
How do you build a GA4 funnel exploration step by step?
- Open Explore and choose the Funnel exploration template.
- Set the date range. Use at least four weeks, and avoid mixing a major sale with normal weeks.
- Define the steps. Add one step per event, for example
view_item,add_to_cart,begin_checkout,add_shipping_info,purchase. Steps can use events or dimension values, but Google’s help center notes you cannot define steps based on metrics. - Choose open or closed. In a closed funnel, users must enter at the first step. In an open funnel, users can enter at any step.
- Set step relationships. Each step can be “indirectly followed by” (other actions allowed in between) or “directly followed by” (must happen immediately after). You can also add a “within” time limit between steps.
- Turn on elapsed time to see the average time between steps.
- Add a breakdown dimension, such as device category, to split each step.
- Add a next action dimension, such as event name or page path, to see the top five things users did after each step.
- Apply segments to compare groups side by side. Google’s help center says you can apply up to four.
Which funnel settings change what the numbers mean?
The same steps can tell different stories depending on a few settings.
| Setting | What it does | When to use it |
|---|---|---|
| Closed funnel | Users must start at step 1 | Clean path analysis, such as product view to purchase |
| Open funnel | Users can enter at any step | Seeing how many people reach checkout from any route, such as a direct cart link |
| Indirectly followed by | Other events allowed between steps | Default for most ecommerce funnels |
| Directly followed by | Next step must come immediately | Tight sequences, such as checkout sub-steps |
| Within (time limit) | Step must happen within a set time | Same-visit behavior, for example add to cart within 30 minutes of view |
| Standard vs trended | Stepped chart or line chart over time | Trended to check whether a fix moved a step |
| Breakdown | Splits each step by a dimension | Device, channel, landing page, item category |
Two details from Google’s documentation affect how you read the results. GA4 counts each user once in a funnel, and only the first sequence of steps is reported. And in a breakdown, users are attributed only to the first value of the breakdown dimension that applies to them. A shopper who browsed on mobile and bought on desktop may sit entirely in the mobile row.
How do you find why shoppers drop out?
The funnel tells you where the biggest loss is. Finding the cause takes more work.
Break the step down
Split the leaking step by device category, session default channel group, landing page and item category. A mobile-only drop usually points to layout or speed. A drop for one channel often means a traffic quality or message mismatch. A drop for one category can point to price or delivery.
Read the next actions
The next action card shows the most common things users did after a step. After add_to_cart, you might see view_cart, begin_checkout or a return to browsing. Google’s help center notes that “(no next action)” does not always mean the user left, particularly when the next action dimension is a page or screen, so read it with care.
Check elapsed time
Long gaps between add to cart and checkout suggest hesitation. Quick exits during checkout suggest a surprise, such as shipping cost.
Watch the sessions
Right-click a row in the funnel table to build a segment from it, then use a session recording tool such as Microsoft Clarity to watch visits that match the same pattern. Twenty to thirty recordings usually reveal a repeated behavior.
Worked example: a checkout drop on mobile
All numbers below are illustrative.
A store builds a closed funnel over four weeks:
| Step | Users | Continue to next step |
|---|---|---|
view_item |
50,000 | 8% |
add_to_cart |
4,000 | 50% |
begin_checkout |
2,000 | 70% |
add_shipping_info |
1,400 | 71% |
purchase |
1,000 |
- Look past the first step. Most users leave between product view and add to cart, but that step mixes browsers and buyers, so the store looks for a sharper signal first.
- Break down by device. On desktop, 80% of checkout starters add shipping details. On mobile, only 60% do. The gap sits at one step.
- Read elapsed time and next actions. Mobile users who leave spend little time on the checkout page and often go back to the cart.
- Watch recordings. Mobile sessions show shoppers scrolling to the shipping cost line and leaving. On desktop, the shipping estimate already appears in the cart, but the mobile cart drawer hides it.
- Fix and measure. The store shows the shipping cost, or the amount needed to reach its free shipping threshold, in the mobile cart drawer. A trended funnel over the next four weeks shows whether the mobile step improves without moving the drop to a later step.
For more checkout fixes, see our guide to Shopify checkout optimization.
What are the limits of GA4 funnel explorations?
- Data retention. Google’s documentation says the data retention setting affects explorations and funnel reports, but not standard aggregated reports. Standard properties can keep event data for 2 or 14 months. Check that yours is set to 14 months, or year-over-year funnels will be empty.
- Users, not orders. The funnel counts users, so its purchase step will not match your order count; our article on the GA4 vs Shopify revenue discrepancy explains the usual gaps.
- Consent and blocked tracking. Users who decline analytics cookies or block scripts may be missing from some steps.
- Event quality. A step that drops to near zero, or purchases far above checkouts in an open funnel, usually means a missing or duplicate event.
Checklist: before you share a funnel exploration
- [ ] Each ecommerce event fires once per action, confirmed in real-time or DebugView.
- [ ] Checkout steps (
begin_checkout,add_shipping_info,add_payment_info) are present, or the funnel skips them on purpose. - [ ] The date range covers at least four weeks and does not mix sale and non-sale periods.
- [ ] Open or closed funnel is chosen on purpose and labeled in the report.
- [ ] Step relationships and any time limits are noted.
- [ ] The leaking step is broken down by device, channel and landing page.
- [ ] Next actions and elapsed time are checked for the leaking step.
- [ ] Session recordings for the same segment have been reviewed.
- [ ] Data retention is set to 14 months.
FAQ
What is the difference between an open and a closed funnel in GA4?
In a closed funnel, users must enter at the first step to be counted. In an open funnel, users can enter at any step. Use closed funnels for a strict path, such as product view to purchase, and open funnels to see total traffic into a later step from any route.
Why does my GA4 funnel show more purchases than checkouts?
Usually because the checkout event is missing on some routes, for example an express checkout button, or because the funnel is open and users enter at the purchase step. Check the events in DebugView and confirm the funnel type.
Can I create an audience from a funnel step?
You can right-click a row in the funnel table to build a segment from it, which you can then compare or use to find recordings. For remarketing, build an audience in GA4 with the same conditions, such as users who added to cart but did not purchase, and check the cart abandonment rate it represents before you spend on it.