Short answer
Compare step completion rates in GA4 and Shopify, check your tracking first, and use the funnel to find the shopping step worth improving.
The answer in detail
GA4 provides two ways to read a conversion funnel: the fixed Purchase journey report and a Funnel exploration with up to ten steps. Read the proportion of people progressing from one step to the next. Use that to find the biggest loss, rather than trying to reproduce Shopify's totals.
The benchmark
The Purchase journey report sits under Reports > Life cycle > Monetization. It follows session_start, view_item, add_to_cart, begin_checkout, and purchase. The funnel is closed by default; “Make open funnel” permits entry at any step. Its steps cannot be customized.
Funnel exploration supports ten steps and four segments, but each user enters only once. Shopify's Conversion rate breakdown counts sessions. These English help labels and the source's product observations are dated October 2026.
| Comparison | GA4 Purchase journey | Shopify Conversion rate breakdown |
|---|---|---|
| Counting unit | Users | Sessions |
| Checkout step | begin_checkout, when checkout opens |
A click or input during checkout |
| Entry | Closed by default; open-funnel toggle | Open Funnel toggle |
| Bots | Known bots excluded automatically | Identified bots filtered by default since September 2026 |
Google maps begin_checkout to Shopify's checkout_started, which fires when checkout opens. Shopify's report waits for an interaction. Cart-to-checkout progression can therefore look higher in GA4. Shopify's September 21–23, 2026 session-measurement change also shifted the baseline; see the dedicated explanation.
The rule
Compare step completion rates and put absolute volumes aside. A 15% difference between Shopify sessions and GA4 users, for example, does not diagnose the site. A mobile cart-to-checkout drop visible in both tools identifies a journey worth investigating.
For a closed funnel with consistent counting, a step cannot retain more than 100% of the preceding population. A separately calculated ratio above that level warrants checking missing or duplicated events and incompatible definitions. In a January 2026 Club audit, cart-to-checkout ratios reached 175% on tablets and 1,400% for visits without an identified device. Tracking needed repair first.
Keep the funnel closed for journey diagnosis; open it when checking missing earlier steps near purchase. More purchases than checkout starts can expose a measurement problem. The overall conversion rate is a separate measure.
The steps
- Open the report. Use Reports > Life cycle > Monetization > Purchase journey. An editor can add a missing report through Library and Edit collection. Record four weeks of step rates by device.
- Check coverage. Divide GA4 purchases by Shopify online orders for the same dates. Above 100%, investigate duplication and reconciliation scope. Below 70%, repair measurement before interpreting the final step. That 70% is the author's working threshold from the full funnel method, not a Google standard.
- Build an exploration. Recreate the five steps and add
view_item_list,add_shipping_info, andadd_payment_info. Google's Shopify event mapping saysadd_shipping_inforequires Checkout Extensibility. - Inspect time. Add device category as the breakdown and enable elapsed time. A long cart-to-checkout delay warrants checking when shipping charges appear.
- Inspect the next action. Use a page dimension under Next action to see the five most frequent actions following each step.
- Inspect the segment. Right-click the largest-loss step to create a segment, then examine relevant session recordings.
The pitfall
A dramatic final step often describes tracking before it describes customers. Shopify's checkout_completed fires once per checkout, on the thank-you page or first post-purchase offer. If that page never loads, the purchase event may never be sent. Cookie refusal can remove another share where tracking depends on consent.
Duplication also happens. In April 2026, a fashion brand recorded 1,026 GA4 purchases against 568 Shopify orders. A server-side tracking app and the original Google & YouTube app each sent purchase. Before removing an integration, compare what each suspected sender transmits with what GA4 receives.
Do not modify GA4 merely to resemble Shopify. Choosing the management reporting number is a different decision. A sudden step collapse after a migration or app installation often indicates an event that stopped firing.
What to tell your team
Use your own figures; this example is fictional:
“On mobile, 38% of people who add to cart open checkout, compared with 52% on desktop. Shopify and GA4 show the same gap, so this month we are improving the mobile cart.”
