Google Analytics 4 for ecommerce - events, reports, and common gaps
Search Console tells you how people find you; GA4 tells you what happens next. For ecommerce, GA4's value lives in its ecommerce events - and so do its problems, because those events only exist if something sends them correctly.
The event chain that is your funnel
GA4's ecommerce model is a chain of events: view_item → add_to_cart → begin_checkout → purchase (plus view_item_list for collection views). Every ecommerce report is built from these. On Shopify, they're typically sent by the Google & YouTube channel app, a tag manager setup, or your theme - and "typically" is doing heavy lifting. Stores routinely run with view_item missing item IDs, purchases double-firing, or a checkout upgrade having silently broken an event. Rule of thumb: before trusting any GA4 ecommerce report, compare a month of GA4 purchase revenue with Shopify's own numbers. Within a few percent is normal (attribution and timing differ). Off by 20% means your data layer needs attention before your analysis does.
The reports worth an ecommerce manager's time
Monetization → Ecommerce purchases: performance per item. Cross-reference with your catalog: products with high views and low purchases are conversion problems (price, imagery, data gaps like a missing size guide). Products with almost no views are visibility problems - GA4 shows the symptom, GSC usually holds the diagnosis.
Funnel exploration: build the view_item → purchase funnel once, save it. Where the drop-off concentrates tells you where to work - a weak add-to-cart rate is a product page problem, a weak checkout completion is a friction problem.
Traffic acquisition: which channels drive purchases, not sessions. Sessions flatter; purchases pay rent.
GA4 and GSC answer different questions - use both
GSC sees the search results page (impressions, position). GA4 sees your site (behavior, conversion). Neither replaces the other, and the seam between them is productive: a collection page GSC shows winning clicks but GA4 shows converting nobody is a landing-experience problem - often, in our experience, a product data problem wearing an analytics costume: the page promised something the products on it don't clearly deliver.
Two setup notes that prevent future grief: extend data retention from the 2-month default to 14 months (Admin → Data settings), and mark your key events properly so purchase-based reporting works everywhere. Deeper measurement - server-side tracking, consent mode, attribution - is its own rabbit hole; what's here covers the daily reading.