:ladybug: Additionally, we have a problem with the...
# ask-questions
i
🐞 Additionally, we have a problem with the activation metric. We have the experiment set up as before, and it seemed ok. But after adding the activation metric, it shows no results, although the activation metric returns users to be included in the analysis. Where do we start debugging this?
f
A good start is to 'view queries' for that experiment
grab one, and then you'll see the section that adds the activation metric
i
Ok, I'll inspect that.
Meanwhile, I was wondering, if we chose our own metric for bucketing users into different variants, let's call it "custom_anonymous_id" do all our metrics need to use that one as "anonymous_id" or can we use the default Segment anonymous_id. I'm not sure how GB connects all of the events. Does the bucketing metrics influence that or not at all?
And by bucketing I mean the one where it says "split users by"
f
they don't have to match
but keep in mind, that if the id's have different times to live, you can end up with multiple exposure warnings
i
Alright. Thanks a lot.