-
Differences in Event Count Between BigQuery and GA4 UI
Hey there! I’m grappling with a bit of a mystery and I’m curious if anyone else has stumbled on this before or has any insights. So, here’s the scoop: I’ve been working with GA4 data in Bigquery, right? And my main mission has been to check the data against what I see on the Google Analytics UI.
Well, strange things are afoot! In pretty much every instance, the Bigquery data is showing me almost twice the count of recorded events that I see on GA4. Is there something different about how I should count events? It seems like it should be straightforward – count the event rows, right?
Here’s a basic example. I counted all the leads from May 20, 2023 with this simple query:
SELECT COUNT(*) FROMmydataset.events_20230520The result? A whopping 100690 events via BigQuery. But when I checked the GA4 UI for the same date, it was showing me just 30000 events. So, it seems my BigQuery count is more than triple that of GA4. Wacky, right?
I also compared some data for the same period in both Engagement > Events and Acquisition > User Acquisition, and most of the numbers seem to line up, except for the Bigquery Event Count, which again was about 3X larger.
Regardless of how I arrange my views, BigQuery consistently seems to bring back triple the event count. So, is it me? Is there a different way I should be counting events? How does GA4 do it? Are there any event types I might want to exclude from the total? Or is this a riddle wrapped in a mystery inside an enigma that I just need to accept? Would love to hear any thoughts!
-
2 Replies
-
A comparison between the event counts in Google Analytics 4 (GA4) and BigQuery (BQ) indeed seems puzzling, because the count in BQ is frequently much higher. The two counts differ because they use different methods to count events. BQ gets down to the nitty-gritty, providing you with the most detailed count possible, also known as the “ground-truth” data. On the other hand, GA4 uses a more high-level approach and provides estimates, also excluding some sessions (like Google Signals) that are included in BQ. Unfortunately, Google didn’t design the two to match each other easily. So, even though it seems odd, what you’re experiencing is pretty standard.
-
The discrepancy between Google Analytics 4 (GA4) and BigQuery may be due to extra events recorded in BigQuery that GA4 doesn’t count, like “first_open” or “session_start”. To match GA4, you could modify your query to exclude these. Also, GA4 aggregates data and removes some outliers, which could account for lower numbers. It’s not just you—this can be a complex thing to navigate!
Log in to reply.