Search papers, labs, and topics across Lattice.
This paper introduces OCARM, a two-stage distillation-aligned framework designed to enhance user retention modeling in real-time bidding advertising systems by leveraging post-conversion onboarding content without causing feature leakage. The method first trains a hierarchical encoder on onboarding content to generate teacher representations, which are then used to guide a user encoder through a distillation process, allowing the model to predict future user engagement based solely on observable features. Extensive experiments show that OCARM significantly improves retention predictions, demonstrating its effectiveness in real-world applications.
Retention models can now harness the power of post-conversion content without risking feature leakage, leading to more accurate predictions of user engagement.
User retention is a key metric to measure long-term engagement in modern platforms. In real-time bidding (RTB) advertising system for user re-engagement, the retention model is required to predict future revisit probability at bidding time, before the user converts and consumes any content. Although post-conversion content, termed Onboarding Content, provides highly informative signals for retention prediction, directly using it in training causes severe feature leakage and creates a gap between training and serving. To address this issue, we propose OCARM, a two-stage distillation-aligned framework for Onboarding Content Augmented Retention Modeling, enabling the model to implicitly capture future content using only observable features during inference. In the first stage, we deliberately expose onboarding content to train a hierarchical encoder that produces teacher representations. In the second stage, a user encoder is aligned with the frozen teacher through distillation, allowing the model to approximate the inaccessible onboarding signals without leakage. Extensive offline experiments and online A/B tests demonstrate that our framework achieves consistent improvements in a real-world growth scenario.
