search

GoogleCloudPlatform / analytics-componentized-patterns

Apache License 2.0
173 stars 97 forks source link

which user identifier to use #65

Closed Joshuacourse closed 3 years ago

Joshuacourse commented 3 years ago

In the notebook, I saw "In our analysis, we used user_pseudo_id as the user identifier. However, ideally, your app should send back the user_id from your app to Google Analytics." I have a few questions: Is user_pseudo_id automatically enabled? If I am using Google Analytics data only (no other 1st party data) for model building and audience activation is via Google Ads, DV360, or SA360 (no CRM), user_id is not necessary, am I assuming correctly?

minhaz commented 3 years ago

user_pseudo_id is always available in the GA4 dataset. Here's more info on this.

Joshuacourse commented 3 years ago

Hi Minhaz, thank you for sharing. The link you shared does not mention user_pseudo_id. I am trying to connect the user_pseudo_id column in the BigQuery data to the UI concept. Did I miss anything?

minhaz commented 3 years ago

In GA4 BigQuery dataset, the user_id column is what you(i.e. developer) sends back to GA as user id with each hit. If you don't send back anything, this will be blank. user_pseudo_id is generated by GA based on user identity and is automatically filled in. The link above explains how the user identity is determined.

Joshuacourse commented 3 years ago

Hi Google team, if I would like to identify high-intent users in Google Analytics and activate them via DV360 custom bidding. Should I use user data import or a GA goal?

Joshuacourse commented 3 years ago

Hi Google team, GA already has a conversion probability report under Audience > Behavior > Conversion Probability. What is the added value of doing modeling with raw data?