How to Get Ahead in Advertising
summary
Explains concrete multi-channel attribution methods implemented in R and released as a channel attribution R package, comparing Markov chain models (using a removal effect to measure lost conversions) and game‑theory Shapley allocations. Runs on session sequences (UTM-tagged touchpoints) inside Jupyter/R workflows, and finds surprisingly little difference between models on real data because most converting users have few touchpoints; reveals inflated Google clicks and concludes better data quality and tagging matters more than choosing a perfect model.
AI-written summary · about
chapters (6)
- 0:00Introduction and background
- 3:37Marketing attribution problem
- 8:15Attribution models and methods
- 12:03Model implementations and demos
- 18:20Practical lessons learned
- 20:30Conclusion and Q&A
related talks
29mMedia Mix Modelling
37mMulti-touch Attribution: What am I training for?
47mIncrementality Bidding and Attribution
31mMedia Mix Modeling:How to Measure the Effectiveness of Advertising
27mAI and analytics for better marketing performance
46mFrom correlation to causation: measuring ad effectiveness at scale
39mHow to Allocate Marketing Budget - Bayesian Approach to the Rescue
18mBuilding Models Using Ad Sales as an Example