Shopify · Analytics
Shopify analytics agency
Shopify is an analytics system in its own right, and it rarely agrees with GA4 or the ad platforms. The figures differ because each tool counts a different thing at a different moment. Analytics work on Shopify means choosing which number is the store's truth, then making every other report explain its gap rather than argue with it.
What Shopify changes
Platform facts that shape the work
These are true of Shopify today and they decide the approach before any strategy does.
The order ledger is the source of truth
Shopify records the order, the discount, the tax, the shipping and any later refund. That ledger is complete in a way a browser tag can never be, so other reports are judged against it rather than the other way around.
Marketing tags run in a sandbox
Customer events runs pixels away from the theme with a defined set of events covering page views, cart actions, checkout steps and purchase. Custom measurement is a pixel in that sandbox rather than a script pasted into theme files.
Shopify defines its own sessions
Sessions, channels and attribution follow Shopify's rules, which do not match GA4's. Comparing conversion rates between the two without stating whose definition you are using produces a disagreement that no amount of tag debugging will settle.
Refunds land after the fact
A refund reduces revenue against the original order in Shopify. GA4 and the ad platforms keep the original figure unless refunds are deliberately sent back to them, so the gap between systems widens as the month goes on.
Markets add currency to the problem
With Shopify Markets a shopper can pay in a local currency while the store reports in its own. Analytics has to fix a currency of record and a conversion approach, or channel performance will appear to move whenever exchange rates do.
The work
Item by item, inside Shopify
Measurement plan
One written definition for session, order, revenue, new customer and channel, agreed before any tag is touched. Every dashboard afterwards refers to that document, which is what stops the same question being re-litigated each quarter.
GA4 rebuilt on customer events
The ecommerce event set implemented through a custom pixel with consistent item ids, prices excluding or including tax by decision rather than accident, and checkout steps mapped so the funnel in GA4 matches the funnel shoppers walk.
Server-side tagging and consent
A server container for the events that matter most, consent signals passed correctly so measurement degrades in a controlled way, and enhanced conversions configured where the shopper has agreed to it.
Reconciliation reporting
A standing report that shows Shopify orders next to GA4 and each ad platform for the same window, with the size and cause of every gap named. Differences stop being alarming once they are expected and explained.
Attribution and channel view
Platform reported numbers set beside a last non-direct view and, where volume allows, a holdout or geo test, so budget decisions rest on something other than each platform grading its own homework.
Dashboards people use
A weekly view built around the growth levers: traffic, conversion rate, average order value and purchase frequency, each split by the few dimensions the team can act on, rather than a page of charts nobody opens.
What goes wrong
On Shopify, specifically
- Trusting a single number from each platform and adding them together, which counts the same order several times
- Comparing Shopify's conversion rate with GA4's without noting that the two count sessions differently
- Keeping legacy tags alongside a new pixel so purchases are recorded twice for months
- Excluding refunds from every report and then wondering why finance and marketing disagree
- Building a warehouse before anyone has agreed what a new customer means
Which number should a Shopify store treat as correct?
Shopify's order data, for anything about money. It knows discounts, taxes, shipping and refunds. GA4 and the ad platforms are better at questions about behavior and channel contribution, so use them for direction and use Shopify for the ledger.
Does a Shopify store need server-side tagging?
It helps when browser measurement is losing a meaningful share of conversions, when you run significant paid spend, or when you need consistent data across several platforms. It is not a fix for a badly specified event model, so the measurement plan comes first and the server container second.
Why does the Shopify dashboard disagree with GA4 on sessions?
The two define a session differently, treat bots and referrals differently, and start counting at different moments. Neither is broken. Pick one for trend reporting, note the typical gap between them, and stop comparing the absolute figures side by side.
Can Shopify data be sent to a data warehouse?
Yes. Orders, customers and products can be exported through the API or through a connector into a warehouse, which is where cohort analysis, contribution margin by product and lifetime value calculations become straightforward. It is worth doing once the questions have outgrown the built in reports.
How do you report new versus returning customer revenue on Shopify?
Shopify flags whether an order is a customer's first, which is the cleanest basis for the split. Ad platforms report differently because they see a device rather than a customer record, so we report acquisition from the order data and use the platforms only for in-flight optimization.
More on Shopify
Grow on Shopify.
A free Growth Analysis ranks what your store should fix first, by revenue at stake.