Full Answer

Platform variance is not a problem to solve — it is a structural reality to understand. Even with perfect data collection on both sides, GA4 and Facebook will report different numbers because they are answering different questions about the same set of transactions.

With clean server-side tracking feeding both platforms the same events, the typical variance for WooCommerce stores ranges from 10-30%. Most of this gap is explained by view-through conversions. Facebook counts purchases that occurred after an ad impression within its view window. GA4 never sees these as Facebook-attributed sessions because no click occurred. This single factor accounts for the majority of the expected gap.

Without server-side tracking, the variance balloons to 40-80%. The additional gap comes from client-side data loss: ad blockers preventing GA4's JavaScript from firing, Safari ITP restricting cookie-based attribution, consent banners delaying or blocking tag execution, and page-load failures where the confirmation page did not fully render the tracking script. These losses affect GA4 disproportionately because it depends entirely on client-side execution, while Facebook's CAPI implementation sends data from the server.

The diagnostic approach is to track the variance percentage over time rather than obsessing over absolute alignment. If the variance holds steady at 20% week over week, your measurement infrastructure is stable and the gap is attributable to known model differences. If the variance suddenly jumps to 50%, something has changed — a new ad blocker list update, a consent banner configuration change, a broken tag, or a redirect stripping UTM parameters. Consistent variance is healthy. Volatile variance is a signal.