Introduces a framework that converts sparse incrementality experiment lifts into daily attribution corrections under structural constraints, reducing calibration error and measured cannibalization rate by ~15pp in TikTok deployments.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Attributed, But Not Incremental: Cannibalization-Corrected Attribution for Large-Scale Advertising
Introduces a framework that converts sparse incrementality experiment lifts into daily attribution corrections under structural constraints, reducing calibration error and measured cannibalization rate by ~15pp in TikTok deployments.