Learn the techniques to accurately measure the impact of your marketing and advertising efforts.
Overview
Syllabus
Introduction
- Measuring marketing performance
- Last-click attribution: The default model
- Time decay and conversion lags
- Linear attribution: Treating all touches equally
- First-click models: From awareness to acquisition
- Position-based models and assigning credit
- Data-driven attribution and machine learning
- Click windows and view-through conversions
- Before and after an event: Trend analysis
- Linear regression with a single variable
- Variables with positive and negative correlations
- Multivariable regression: Building your marketing mix model
- Feature transformation with diminishing returns and adstocks
- Statistical tests to validate your model's accuracy
- Forecasting future scenarios for planning
- A/B testing for statistical significance
- Bandit testing: Optimizing for results over accuracy
- Geo and lift testing to prove incrementality
- "How did you hear about us?": Surveys and panel studies
- Working with multiple attribution methods
- Continuing to improve your model accuracy
Taught by
Michael Taylor
