Follow the TRACE: Exploiting Post-Click Trajectories for Online Delayed Conversion Rate Prediction

TL;DR

This paper introduces TRACE, a novel method for online CVR prediction that models post-click feedback trajectories to improve delayed conversion rate estimation without waiting for final outcomes.

Contribution

The paper formalizes feedback trajectory evolution and proposes TRACE, which dynamically refines posteriors and includes a reliability-gated retrospective completer, enhancing existing CVR systems.

Findings

TRACE outperforms state-of-the-art baselines in experiments.

The retrospective completer improves model accuracy and robustness.

The method is model-agnostic and adaptable to various systems.

Abstract

Delayed feedback poses a core challenge for online CVR prediction, forcing a trade-off between label accuracy and data freshness. Existing methods address this through delay modeling or sample reweighting, yet neglect how post-click behaviors evolve over the observation period. To overcome this limitation, we formalize this evolution as feedback trajectory and propose TRACE. Instead of forcing hard labels on unrevealed samples, our method evaluates how well the accumulated feedback status aligns with conversion versus non-conversion, dynamically refining posteriors without waiting for final outcomes. To counteract early-stage trajectory sparsity, we further design a reliability-gated retrospective completer that leverages full-lifecycle data to provide adaptive posterior guidance for unrevealed samples. Extensive experiments validate TRACE's superiority over state-of-the-art baselines and confirm the retrospective completion module as a model-agnostic enhancer for existing systems. Our code is available at https://github.com/LunaZhangxy/TRACE.