Dealing with spill-over effects

• What is it? The effect of one campaign influences another, making it difficult to isolate the true effect of each campaign.
• Experimental Approaches.
  • (Of course) RCTs with Holdout Groups.
  • GeoLift (Geographical Testing).
• Statistical & Econometric Approaches.
  • Difference-in-Differences (DID)
  • Marketing Mix Modeling (MMM)
  • Time-Series Models with Decay Factors
• Attribution Modeling
  • Multi-Touch Attribution (MTA)
  • Causal Inference (Synthetic Controls)
• Practical Adjustments