How Grammarly figured out if their YouTube ads actually work
Original: Measuring Marketing Effectiveness in a Cookie-Less World
Summary
This article discusses the technical approaches and considerations for resolving user identities across devices and touchpoints, including anonymous-to-user joins, deterministic vs. probabilistic linking, and managing identity hierarchies and conflict resolution logic. It covers the role of customer data platforms (CDPs) in stitching customer profiles and how to ensure data quality, privacy, and compliance in these identity-matching processes.
Who This Is For
Data Scientists
Performance Marketers
Marketing Analysts
Key Takeaways
- Learn why traditional attribution models fail for top-of-funnel channels like video and brand advertising
- Understand how to set up geo experiments by dividing markets into control and test groups to measure true incrementality
- Discover how to balance statistical power with business disruption when deciding experiment size and duration
- See how geo experimentation reveals 'halo effects' - the impact one channel has on performance of other channels
Tools & Technologies
YouTube advertising platform Designated Market Areas (DMAs) Geo-experimentation frameworks Statistical analysis tools
Topics Covered
identity-resolution cross-device customer-data-platform user-tracking
Ready to dive deeper?
Read Full Article on www.grammarly.com
