Google GA4 Dungeons And Dragons
summary
Describes concrete GA4 fixes: stop sending user_id as the identity field and instead push a CRM_ID variable so blended identity works and reduced not_set revenue from €100k to ~€1–5k/day; persist UTM into localStorage and read it via GTM to join with BigQuery and store UTM in localStorage to join Google Ads + BigQuery costed data; always send client_id/user_id/session_id when using Measurement Protocol; warn that a rogue BigQuery query can hit 4 PB and cost ~€6k; shows a working online+offline data lake with loyalty-card alerts that raises AOV and lowers churn.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 2:14Setting the scene in Dataland
- 3:42Lord Tyros and data pitfalls
- 5:09Lonely Data Tavern heroes
- 8:21First quest: Data depth issues
- 12:38Not set on Ads attribution
- 16:12Not set on GTM dimensions
- 17:17Measurement protocol pitfalls
- 20:11Defeating HyperLogLog trap
- 22:41Scroll of Insights and tracking
- 23:02BigQuery sandbox optimization
- 24:47AI tools for BigQuery
- 25:10Case study: Defeating errors
- 27:44Summary
- 28:07Q&A: Conversion tracking
- 29:54Q&A: UA to GA4 switch
- 31:39Q&A: Audience size differences
- 33:25Q&A: User path mapping
- 34:49Q&A: GA4 vs PPC choice
related talks
48m3 Very Practical Applications of Google Analytics 4
45mBy Enrico Pavan
38mMeasure performance with web-vitals.js, Google Analytics, and BigQuery
47mWhy We Need to Understand Data & SEO in 2024
16mMeasure and debug performance with Google Analytics 4 and BigQuery
13mPower Up Performance – How To Maximise Performance in a Privacy-First Advertising Landscape.
42mGoodbye Tracking, Hello Privacy
38m"Performance Impact, Part Two: More Findings from the Front Lines..."