Once Google Analytics 4 is connected, Cooee creates a GA4 audience for every variation of a live experiment. Each audience holds the visitors who saw that variation, so you can put the variations side by side in any GA4 report and check the result against revenue tracked by your own GA4 setup.
Before you compare
Wait 24–48 hours after the audiences are created. GA4 needs that long to process a new audience.
An audience collects visitors from when Cooee creates it, and GA4 also adds visitors it saw in up to 30 days before that. Cooee tells GA4 a visitor's variation on every page they view, so anyone who joined the experiment earlier is added on their next visit.
Compare variations in a report
Open any GA4 report. Reports snapshot and Monetization overview are good starting points.
Set the date range to start on the day the experiment went live.
Click Add comparison at the top of the report.
Set the dimension to Audience name, choose the first variation's audience (for example Cooee #42 · Homepage hero · Control) and click Apply.
Repeat for every other variation. You can remove the default All users comparison. GA4 shows up to four comparisons at once.
Compare per-visitor numbers, not totals
A variation that gets 20% of your traffic will always show roughly a quarter of the revenue of one that gets 80%. Compare rates instead: purchase revenue per user, the share of users who purchased, and average order value. Revenue per user is the hardest metric to fool.
Why every assigned visitor is counted. Cooee tells GA4 a visitor's variation as soon as they land in it, whether or not they go on to click or buy. Each audience holds everyone that variation was shown to, so the comparison stays like-for-like. Comparing only the visitors who engaged would favour keen shoppers and overstate the lift.
Break results down by variation in Explore
For funnels, or a breakdown by variation in your own explorations, register the variation as a custom dimension in GA4:
In GA4, go to Admin → Data display → Custom definitions and click Create custom dimension.
Name it (for example Cooee variation), set the scope to Event and the event parameter to
exp_variant_string, then save.
GA4 fills a custom dimension only from the day you create it, so create it before the experiments you want to analyse this way go live.
Why GA4 won't match Cooee exactly
GA4 only sees visitors its own tag can track. Visitors who decline analytics cookies or block Google Analytics are missing from your GA4 audiences.
GA4 can hide small numbers (data thresholds) and takes up to 48 hours to process data.
Audiences start collecting when Cooee creates them, while Cooee's own results count from the moment the experiment went live.
GA4 attributes revenue using its own cookie and rules, which differ from Cooee's and Shopify's.
GA4 does not tell you whether a difference is statistically significant. Use GA4 to confirm the direction of the result in your own analytics, and Cooee's results for the significance call.
For more on why counts differ between tools, see Why session and visitor counts differ between Cooee and Shopify.