A client ran regular A/B tests on landing pages. They used GA4 to compare performance — but got three different answers:
• One report said Variant A won
• Another said B performed better
• A third said they were about equal
Why? Each report used a different metric: Conversions, User Conversion Rate, and Session Conversion Rate.
They look similar — but they’re fundamentally different.
GA4 offers three key conversion metrics:
• Conversions — total number of conversion events
• User Conversion Rate — % of users who converted at least once
• Session Conversion Rate — % of sessions with at least one conversion
And in reality:
• A user can have multiple sessions
• One session can include multiple conversions
• A user may convert later — not during the first session
If you don’t know which metric you're looking at, it’s easy to misinterpret the data. It looked like data-driven testing — but was really metric roulette.
Now, I always clarify:
• Are we comparing users, sessions, or event count?
• I compare the same metric across all test variants
• I include formulas under charts in BI dashboards
• And I always try to make all the calculations in BigQuery
GA4 is powerful — but it won’t warn you when you’re mixing metrics and drawing the wrong conclusion from the right-looking chart.
Want to get all my top Linkedin content? I regularly upload it to one Notion doc.
Go here to download it for FREE.


