Kapital - AI Adoption Masterclass - AI-Powered Analytics in GA4

Overview of the embedded and extended methods to leverage AI within the Google Marketing Platform and Google Cloud Platform

1. AI -Powered Analytics in GA4 Masterclass: Marketing in the era of AI Yasen Lilov Partner, Head of Data & Analytics

2. All time high!

3. All time high!

4. AI is not only this year’s hype.

5. AI is not only this year’s hype.

6. AI is not only this year’s hype.

7. DATA COLLECTION DATA ANALYSIS REPORTING DATA ACTIVATION App - Firebase Web - gtag.js / GTM Offline - Measurement Protocol Standard Reports Exploration Reports Data API Google Ads Search Ads 360 Display & Video 360 BigQuery … Google Looker Studio BigQuery Google Analytics 4 AI

8. Embedded AI Improves performance at scale through AI functionalities embedded in platforms. Google AI supports effective performance through best-in-class machine learning algorithms. Applied AI Augments marketing strategy by incorporating Cloud-based solutions and leveraging an advertiser’s first-party data to harvest unique insights that can be subsequently integrated into campaigns.

9. Embedded AI in GA4 Modeling: Observed data where possible, modeled data where necessary Regulatory Changes Regulations such as GDPR and ePrivacy Directive are impacting how data can be collected and used 3rd Party Cookies Use cases for measuring and reaching users with pixels are limited and will continue to decrease over time Browser Updates Heightened controls are impacting traditional data collection

10. Embedded AI in GA4 Modeling: Observed data where possible, modeled data where necessary Reported data 2016 100% 0% 2022 Modelled Observed Modelled Observed Future Modelled Observed

11. Embedded AI in GA4 Modeling: Observed data where possible, modeled data where necessary Attribution modeling looks at how credit for conversions is assigned to advertising touchpoints. Behavior modeling uses machine learning to predict the behavior of users who opt out of Analytics cookies, based on the behavior of similar users who don’t opt out. Conversion modeling estimates online conversions that can’t be directly observed by looking for trends between directly observed and unobserved conversions.

12. Embedded AI in GA4 Modeling: Enabled by [Google Consent Mode] Attribution modeling looks at how credit for conversions is assigned to advertising touchpoints. Behavior modeling uses machine learning to predict the behavior of users who opt out of Analytics cookies, based on the behavior of similar users who don’t opt out. Conversion modeling estimates online conversions that can’t be directly observed by looking for trends between directly observed and unobserved conversions.

13. Embedded AI in GA4 Google Consent Mode Observed data (with Cookies) Cookieless pings for Modelling Cookiebot (example) Modeling: Enabled by [Google Consent Mode]

14. Embedded AI in GA4 Google Consent Mode Observed data (with Cookies) Cookieless pings for Modelling Cookiebot (example) Modeling: Enabled by [Google Consent Mode]

15. Embedded AI in GA4 Behavioral modelling kicking in blending observed and modelled data. Behavior modeling uses machine learning to predict the behavior of users who opt out of Analytics cookies, based on the behavior of similar users who don’t opt out.

16. Embedded AI in GA4 Without Consent Mode With Consent Mode, Firing cookieless pings for ad_storage or analtics_storage = denied users Modeled Conversions Observed Conversions Observed Conversions 80 % of lost conversions recovered Actual Conversions Data exported from the CRM Total conversions Recovered conversions Reported conversions Average opt-in rate 76% Conversion modeling estimates online conversions that can’t be directly observed by looking for trends between directly observed and unobserved conversions.

17. Embedded AI in GA4 Modeling: Available out-of-the box. Attribution modeling looks at how credit for conversions is assigned to advertising touchpoints. Behavior modeling uses machine learning to predict the behavior of users who opt out of Analytics cookies, based on the behavior of similar users who don’t opt out. Conversion modeling estimates online conversions that can’t be directly observed by looking for trends between directly observed and unobserved conversions.

18. Embedded AI in GA4 30% 1 5% 20 % 35 % Attribution modeling looks at how credit for conversions is assigned to advertising touchpoints. [Available out-of-the box]

19. Embedded AI in GA4 How to take advantage? Implement and adopt Google Analytics 4 Configure Advanced Google Consent Mode 1 2

20. Applied AI in GA4 Marketing data is a very rich source that tells us about user behavior and product behavior. Applied AI can make use of this data by leveraging custom-built solutions to deliver additive value relative to embedded solutions .

21. Applied AI in GA4 Native integration Your entry to the Cloud . CRM Data Product Data Ads spend Data …

22. Applied AI in GA4 Forecasting User Segmentation and Clustering Propensity modeling

23. Applied AI in GA4 User Segmentation and Clustering Group similar customers for personalized marketing and product development. Rejuvenation Cream = high interest Foot Care = high interest Lipstick = high interest Nail Polish = medium interest Make Up = high interest Hair Care = low interest

24. Applied AI in GA4 Propensity modeling Identify customers most likely to convert or churn, enabling targeted marketing campaigns and resource optimization. Machine Learning Mobile Device, 1 Page View 5% likelihood to convert Mobile Device, 5 Page Views 30% likelihood to convert Desktop Device, 7 Page Views, 1 Add to Cart 75% likelihood to convert

25. Applied AI in GA4 Forecasting Leverage historical data to decide on actions to take in the future. The decisioning model relies heavily on accurate and timestamped data. ACTIVATION

26. Applied AI in GA4 Search Ads 360 Campaign Manager / Display Video 360 GA4 Meta Ads LinkedIn Ads …. Activation

27. Applied AI in GA4 How to take advantage? Integrate and adopt Google BigQuery Explore Google Cloud Platform and its capabilities 1 2

28. Forecasting Leverage historical data to decide on actions to take in the future. The decisioning model relies heavily on accurate and timestamped data. User Segmentation and Clustering Group similar customers for personalized marketing and product development. Propensity modeling Identify customers most likely to convert or churn, enabling targeted marketing campaigns and resource optimization. Attribution modeling looks at how credit for conversions is assigned to advertising touchpoints. Behavior modeling uses machine learning to predict the behavior of users who opt out of Analytics cookies, based on the behavior of similar users who don’t opt out. Conversion modeling estimates online conversions that can’t be directly observed by looking for trends between directly observed and unobserved conversions. Applied AI Embedded AI

29. Technical Marketer

30. Thank you!

31. Yasen Lilov Partner, Head of Data & Analytics VertoDigital /in/yasenlilov/ [email protected] Scan to connect!