Analytics
GA4 for ecommerce stores
GA4 records the ecommerce events a store sends it, from item views to purchases, and builds its reports, funnels and audiences from those events. Most store properties we open have the platform's default integration switched on, the purchase event arriving without item details or refunds, data retention at two months and nobody sure why revenue does not match the store. We configure the property so its numbers can be tied back to orders and used for decisions about traffic, pages and channels.
GA4
Analytics · by Google. A text wordmark, not a logo: we are not affiliated with the vendor and make no claim to its trademarks.
What we do with it
The work inside GA4
Ecommerce events with complete item data
The recommended events are view_item, add_to_cart, begin_checkout, add_payment_info and purchase, each carrying an items array with id, name, category, price and quantity. We implement them through the platform's integration, a tag manager or a custom pixel so every event fires once and the purchase carries transaction id, value, tax, shipping and currency.
Property settings that fit a store
Data retention extended to the maximum for the property type, the reporting timezone and currency matched to the store, internal and developer traffic excluded, key events limited to the actions that matter, and Google Signals and consent settings decided for the markets the store sells to.
Checkout on hosted platforms
Shopify's checkout accepts tracking only through its own integration or a custom pixel, and other platforms have their own constraints. We choose the method that captures the full funnel through to purchase and confirm cross-domain measurement so a session does not split at checkout.
Reports built around the growth model
Standard reports rarely answer a store's questions directly. We build explorations and Looker Studio views for traffic by channel, conversion rate by landing page and device, average order value by source, and returning customer share, so the four levers of revenue are visible in one place.
Attribution and channel grouping
GA4 defaults to data-driven attribution and its own channel definitions. We set custom channel groups so paid social, paid search, email, SMS and affiliates are separated correctly, tag every campaign consistently, and choose the attribution comparison the store will use when judging channels.
Reconciling revenue with the store
GA4 revenue will differ from the store's because of consent, blockers, refunds, currency conversion, test orders and missed events. We measure the gap each month, explain it by cause and keep it stable, so the trends GA4 shows can be trusted even though the totals differ.
When it fits
GA4 is the right tool when
Every store that runs paid traffic or SEO
Channel performance, landing page conversion and device splits need a session-level tool. GA4 is free, connects to Google Ads and Search Console, and exports to BigQuery.
Stores on Shopify, WooCommerce or BigCommerce
Each has a documented way to send ecommerce events, from native integrations to plugins and custom pixels. The setup differs, but the event model is the same.
Merchants who want raw data later
The BigQuery export gives event-level data without sampling, which becomes valuable once the store wants customer-level analysis or a reporting stack of its own.
Teams comparing what ad platforms claim
GA4 sees every channel on one attribution basis. It is the neutral reference when Meta, Google and email each report the same order.
What to watch
Where accounts drift
Facts about defaults and costs, not criticism. These are the settings we check on every account.
Data retention defaults to two months
Event-level data used in explorations is kept for two months unless the setting is changed to the longer option, and the change is not retroactive. Set it on day one.
Thresholds and sampling hide detail
Reports with Google Signals enabled can withhold rows to protect user privacy, and large explorations may be sampled. Small stores see thresholds in demographic reports; larger stores see sampling in long date ranges.
Consent mode changes the numbers
Where consent is declined, GA4 may model conversions and sessions or record nothing, depending on the consent mode setup. Reports in those markets are estimates, and the property notes should say so to anyone reading them.
Duplicate purchases from two integrations
A native platform integration and a tag manager setup both sending purchase events doubles revenue. One method should own each event, and a monthly transaction count against orders catches the overlap.
What is an acceptable gap between GA4 and store revenue?
A stable one that can be explained. The gap comes from declined consent, blockers, refunds, currency and missed events, and its size depends on the store's markets and traffic mix. We aim for a consistent gap month to month rather than a specific figure, and investigate whenever it moves.
Should a Shopify store use the native GA4 integration or a tag manager?
The native integration through the Google channel is the fastest correct setup and covers checkout. A tag manager setup gives control over event detail and other tags but needs a custom pixel for checkout. Many stores use the native integration for GA4 and a tag manager for everything else, with care that purchase is sent once.
Which GA4 reports should an ecommerce team look at weekly?
Sessions and conversion rate by default channel group, landing page conversion for the top pages, the checkout funnel by device, average order value by source and returning customer share. Everything else is for investigation, not routine.
Do you set up the GA4 BigQuery export?
Yes, when the store wants event-level data or is close to the limits of the interface. The export is switched on in the property, lands in a Google Cloud project the store owns, and we build the first queries and views so the data is usable from the start.
Does GA4 replace the store's own analytics?
No. The platform's analytics are the record of orders and revenue. GA4 explains where sessions came from and what they did before ordering. Use the store for totals and GA4 for behaviour and channels, and keep them reconciled.
Delivered through
Works alongside
Find the leak.
A free Growth Analysis ranks what your store should fix first, by revenue at stake.