-
Unidentified values linked to session start events in custom dimensions
Hey all!
So here’s the story. We’ve been working on our website and decided to add GA4 tracking, pretty cool right? We also set up some custom parameters on our events and turned them into custom dimensions for our reports. But here’s the kicker, we started to see “(not set)” values showing up when we break out by a custom dimension. These appear with 0 views, but interestingly, about double the total sessions.
We then dove deeper and filtered for “(not set)” on the custom dimension, and found mostly “session_start” and “first_visit” events recorded. But when we filtered out “(not set)” for the custom dimension, the number of sessions for the page_view event just doesn’t match up with the total. Pretty odd, right? Especially since we don’t even have an app or any app tracking set up.
Our detectives checked the API results and they sync with this. Here’s the code on our site, by the way, and we’ve been following the gtag guide:
[Insert Code Here]
So here’s what we’re trying to understand:
– Could we be missing something in our implementation that’s causing these (not set) values to turn up on custom dimensions?
– Are we messing up somewhere and not linking session_start and first_visit events to page_view events within the session?
– Any thoughts on why the Total Sessions are more than the count of sessions with a page_view event?
– It seems like multiple parameters within the same session are counted twice when breaking out by these parameters (custom dimensions) – is this part of the GA4 package or did we miss something in our set up?Just so you know, we’re running analytics.js in parallel, we’ve upgraded to 360 and we’ve turned off the “Collect UA events in GA4” setting in our admin config. Our GA4 property is also connected to GAM, and some of our advertisers use gtag.js too. But after checking over and over, it seems our specific G- measurement ID only shows up in that code I shared above.
Here’s an example page to give you some context: [Insert Link Here]
We’re excited to hear your thoughts! Looking forward to some enlightening conversations. Thanks a bunch!
-
2 Replies
-
Hey there!
Just to clear up the situation a bit – it’s totally normal that your custom parameters are appearing as ‘(not set)’ for the ‘session_start’ and ‘first_visit’ events. These are auto-collected events and unfortunately, they’re just not designed to capture custom parameters. It’s kind of like they’ve got their own thing going on, y’know?
And hey, before you ask – yes, this could have some impact on your reports. Here’s how it works. If you base your reports on the ‘views’ metric solely, and you’re always setting your custom parameter, you’ll never lay eyes on a ‘(not set)’ value in your report. But if you’re a ‘users’ metric kind of person, be prepared to see a bit more ‘(not set)’ action. This is simply because ‘users’ includes those fun-lovers, the ‘session_start’ events, which as we know, don’t play ball with your custom parameters.
So, don’t fret. You’re not missing anything, and you’re not messing up. It’s just the way things flow in the GA4 world! Hope we’ve managed to illuminate the situation for ya. Let us know if you need more info. Keep up the good work with your data and happy analyzing!
-
It sounds like you’re experiencing some interesting quirks while setting up Google Analytics 4 (GA4) on your website. It’s normal to see (not set) values when dealing with custom dimensions. This usually indicates that the parameter you’re trying to pull the value from wasn’t set when the event was triggered. With “session_start” and “first_visit” events, it’s possible these aren’t properly connected to your “page_view” events, which might explain why your total session counts are off. As for the doubling of parameters, GA4 counts events differently than previous versions and will track multiple parameters in a single session separately.
Remember that when running GA4 parallel with analytics.js, keep a close eye on your configurations to not overlap or interfere with each other. And as always, dive deep into the data, filter, and examine various possibilities to root out the cause of these discrepancies. Your code implementation also plays a key role, so a double-check never hurts. With GA4, the learning curve is steep, but it’s a powerful tool once mastered.
Log in to reply.