Answering data modeling question for attribution with Google Analytics 4

GA4 + BigQuery for Attribution: Consent Mode, Omnichannel Journeys, and Data Modeling Choices These questions came up in my longer data modeling session: https://youtu.be/7oSazVTUEE0 In this video I am trying to answer four follow-up questions on using GA4 data for customer journey analysis, attribution, and data modeling. I explain how consented GA4 data includes identifiers (session ID, pseudo ID/client ID, and optional user ID) and works for journey analysis, while non-consent data is largely anonymized “pings” with null IDs and limited usefulness beyond gauging untracked volume or rough aggregated hints. I discuss omnichannel attribution by combining online, virtual touchpoints (e.g., post-purchase “how did you hear about us?”), and offline data via identity mechanisms like loyalty cards, catalogs, or discount codes. I weigh Google’s advanced consent mode modeling as a black box mainly for feeding Google Ads, recommending custom modeling only with strong testing, pipelines, and monitoring, especially at higher ad spend. Finally, I frames Kimball, data vault, and one-big-table as design patterns driven by use cases, favoring activity schema for event-based GA4 modeling and sequence analysis. 00:00 Intro and Setup 00:41 Consent Data vs Pings 03:38 Non Consent Workarounds 05:41 Omnichannel Attribution Basics 06:51 Offline Touchpoints and IDs 09:56 Advanced Consent Mode Models 12:25 Custom Modeling in BigQuery 14:41 Kimball vs Data Vault 17:30 Activity Schema for GA4 20:00 Wrap Up and Next Steps