Hierarchical model decomposes pure and global causal lifts in overlapping customer journeys via multiplicative factors and Monte Carlo estimation on 3M users, finding larger pure lifts and modest positive synergies.
Improving the sensitivity of online controlled experiments by utilizing pre-experiment data
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Hierarchical Causal Uplift Modeling in Overlapping Customer Journeys
Hierarchical model decomposes pure and global causal lifts in overlapping customer journeys via multiplicative factors and Monte Carlo estimation on 3M users, finding larger pure lifts and modest positive synergies.