by Trivas.aiA founder checks Shopify Analytics at 9am. Revenue for yesterday: $14,200. Same founder opens GA4 at 2pm to pull a channel report, and ecommerce revenue for the same day reads $11,800. Nothing got refunded. No orders got cancelled. The numbers just don't match, and now there's a Slack thread asking which dashboard is broken.
Neither one is broken. This is the most common version of the google analytics vs shopify analytics question we hear, and it's not a bug in either tool. It's a structural difference in what each one is actually built to measure.
This piece isn't another feature-list comparison. We pulled an original data breakdown from aggregated, anonymized account patterns across Shopify stores we work with, to show where GA4 and Shopify Analytics actually diverge and by how much. If you're already comfortable using both tools but can't decide which one to trust for which call, this is written for you.
Shopify Analytics is transaction-first. It counts completed orders, checkout completions, and revenue straight from the Shopify order database. When someone pays, it shows up. That's it.
Its attribution model is last-click and tied to the Shopify session itself. That means if a customer clicks an ad, leaves, comes back three days later through a bookmark, and buys, Shopify will often credit "direct" instead of the channel that actually did the work. It undercounts influence from anything that doesn't touch the Shopify domain right before purchase.
What it's genuinely good at: being correct about money. Revenue, refunds, and inventory-linked reporting all read straight from the order ledger, the actual source of truth for what got sold and for how much. If you want to know what you made yesterday, this is the number that's real.
Its blind spot is everything upstream of checkout. Shopify Analytics has no meaningful visibility into on-site behavior before someone reaches checkout, no funnel drop-off data, and no way to tell you ad spend or ROAS by channel. It knows what happened at the register. It doesn't know what brought someone to the store.
GA4 is behavior-first. Pageviews, sessions, funnel drop-off, multi-channel attribution models like data-driven, last-click, or first-click, all of it lives here. This is the tool built to answer "how did people get to my site and what did they do once they arrived."
GA4 builds its own event-based revenue tracking through the ecommerce tracking snippet fired on your site. That's a parallel system to Shopify's order database, not a mirror of it. GA4 is reconstructing a purchase event from JavaScript firing in a browser. Shopify is reading a row from its own order table. Two different methods, two different numbers.
This is why the drift happens. Ad blockers stop the snippet from firing at all. Cookie consent rejections (especially common on EU and UK traffic) mean GA4 never sees the session in the first place. Session timeout rules can also split or merge events in ways that don't match how Shopify defines a single checkout. None of this is GA4 "failing," it's just the tradeoff of tracking behavior client-side instead of reading a server-side ledger.
Where GA4 earns its keep is upper-funnel visibility, traffic sources, landing page performance, audience behavior, the stuff Shopify Analytics doesn't even attempt to track. If you want to understand GA4 funnel data alongside your store, this is the half of the picture Shopify simply can't give you.
Here's the original data breakdown. Across connected Shopify accounts we've seen, GA4-to-Shopify revenue variance isn't random. It clusters by traffic source, and the pattern is consistent enough to plan around.
Direct traffic is where the two tools agree most, usually within a few percentage points, because there's less tracking infrastructure between click and purchase for anything to break. Paid social and email are where it falls apart. Consent banners routinely block GA4 from ever seeing the session, and in-app browsers (Instagram, TikTok, Gmail) handle cookies differently than a standard browser, so GA4 undercounts a real chunk of traffic that Shopify still records as a completed order.
Neither number is wrong. They're answering different questions: "what did we actually sell" versus "what did our tracking see." Reconciling the two means building a third layer that maps both back to the same order IDs, which is exactly the step most teams skip because it means manually exporting from both tools and matching rows by hand.
Simple rule: use Shopify Analytics for anything finance-facing. Actual revenue, refunds, AOV for your P&L. It reads from the order ledger, so it's the number your accountant should trust and the one you should report upward.
Use GA4 for anything channel- or campaign-facing. Which landing page converts best, which traffic source drives repeat visits before someone finally buys, where your funnel drop-off actually happens. That's GA4's job and it does it well.
Don't use either tool alone to calculate true ROAS. Neither one natively joins ad spend data to Shopify order data. GA4 can tell you a session came from a paid campaign, but it's not pulling in what you spent on that campaign from Meta or Google. Shopify doesn't know about ad spend at all. Calculating real ROAS from either tool solo means manual exports and a spreadsheet, every time.
This is exactly the gap unified ecommerce dashboards exist to close, pulling order data, funnel data, and ad spend into one place instead of three separate logins. For teams running this reconciliation manually every week, that's hours spent matching numbers that a connected system handles automatically. If you're the one responsible for explaining the gap between dashboards to your team, this breakdown for founders and CEOs covers the decision-making side of it.
Trivas pulls Shopify order data, GA4 funnel data, and ad platform spend into one Redshift-backed dashboard, so the reconciliation happens automatically instead of someone eyeballing two tabs and guessing which number to trust.
The Wingman AI layer sits on top of that and flags when GA4 and Shopify numbers diverge beyond a normal range for your account. If your paid social revenue suddenly shows a 40% gap instead of the usual 20%, that's not noise, it's usually a sign something broke, like a consent banner misconfiguration cutting off tracking. Catching that on day one beats finding out after a week of decisions got made on bad data.
If you're running Shopify and want this reconciled view, the Shopify integration setup is the first step, and you can also find the app directly on Trivas AI on the Shopify App Store.
If you just want to see what your own GA4-vs-Shopify gap actually looks like before changing anything, that's worth mapping out on its own. No pressure to overhaul your stack to get that clarity.
Why do GA4 and Shopify Analytics show different revenue for the same day?
They track revenue through two separate systems: GA4's event-based tracking versus Shopify's order ledger. Consent banners, ad blockers, and attribution windows cause the two to diverge, usually most on paid social and email traffic.
Which tool should I trust for my actual revenue?
Shopify Analytics. It reads directly from the order database instead of reconstructing revenue from tracked browser events.
Can I use GA4 to calculate ROAS accurately?
Not on its own. GA4 shows ad-driven traffic and conversions but doesn't natively pull in ad spend, so ROAS calculated purely in GA4 is an estimate, not a reconciled number.
Does switching to GA4's data-driven attribution fix the gap with Shopify?
It narrows attribution differences between channels inside GA4 itself, but it doesn't close the gap with Shopify Analytics. That gap comes from tracking loss and session definitions, not the attribution model.
Is there a way to see both numbers side by side automatically?
Yes. Connecting Shopify and GA4 into a unified dashboard maps both data sources to the same order IDs, so the variance is visible and explainable instead of hidden in two separate tabs. That's the core of what BI reporting is built to do, and the data dictionary is a useful reference for exactly how each metric gets defined along the way.
explore more insights
3 min read
3 min read
3 min read
.d53b12e5.png&w=3840&q=75 "Trivas.ai")