GA4 & server-side conversion tracking for Shopify

Why do GA4, Meta and Shopify all report different numbers?

Each platform counts a different thing. Shopify counts orders. GA4 counts purchase events that reached it from a browser that allowed analytics. Meta and Google Ads count conversions they could attribute to an ad, under their own attribution windows. Some gap between them is normal.

The problem is when nobody knows how big the gap should be. A store with several leftover pixels from old apps can double-count purchases in one platform while losing them in another, and the ad platforms then optimize toward a distorted signal. Bidding algorithms are only as good as the conversions they are fed.

How does tracking work on Shopify today?

Shopify manages storefront and checkout tracking through Customer Events, the pixel manager in the Shopify admin. According to Shopify’s pixel documentation, there are two kinds:

• App pixels, installed by marketing and analytics apps, run in a strict sandbox (a web worker) without access to the page’s DOM.
• Custom pixels, added manually, run in a lax sandbox inside an iframe, which lets older pixel code work but not everything a script tag in the theme could do.

Both subscribe to standard events such as page_viewed, product_viewed, product_added_to_cart, checkout_started and checkout_completed. In regions that require consent, Shopify runs pixels only when the visitor has granted the permissions the pixel is configured to need, which is one legitimate reason tracked purchases fall below orders.

What does the reconciliation actually check?

For a fixed date range, we pull Shopify orders and compare them with tracked purchases in each platform, matched on order ID where the platform stores it.

Finding	Typical cause
Order in Shopify, missing in GA4	Event never fired, consent not granted, or tag blocked
Same order counted twice	Theme snippet and app pixel both sending purchase
Purchase in GA4, no Shopify order	Test orders, or a purchase event sent more than once
Revenue mismatch on matched orders	Tax, shipping or discounts treated differently

Every gap gets a cause and a size. That turns “tracking feels off” into a short, ranked list of fixes.

Where does server-side tagging fit?

Server-side Google Tag Manager runs a tagging container on infrastructure you control, such as your own Google Cloud project. The browser sends data to your server container, and you decide what gets forwarded to GA4, Google Ads or other destinations.

For Meta, the server-side path is the Conversions API. When the Pixel and Conversions API both send a purchase, Meta deduplicates them using a matching event name and event ID, so both sides must be built to share that ID.

Server-side tagging is a tool, not a default. We add it when the reconciliation shows it recovers real signal or when you need to control and shape data before it reaches third parties.

How do I keep tracking from breaking again?

Tracking rarely fails all at once. It degrades when a new app adds its own pixel, a theme update drops a snippet, or someone edits a custom pixel without testing it. Nobody notices until a campaign review weeks later.

The handover includes a reconciliation query you can run on any date range and a short checklist for every theme or app change: place a test order, confirm one purchase event per platform, and check the transaction ID matches the Shopify order. If store data is already flowing into BigQuery, the same comparison can run daily and alert when the gap between tracked and actual orders moves outside its normal range.

Who is this not for?

If you spend little on paid acquisition and only need GA4 to show broad traffic trends, the native Shopify integrations are probably enough. This service is for brands where ad platforms optimize on your conversion data and a wrong number costs money every day.

When your ad platforms, GA4 and Shopify have drifted far enough apart that nobody trusts any of them, an Asabisoft tracking diagnostic is a reasonable next step.