Bid2X: Revealing Dynamics of Bidding Environment in Online Advertising from A Foundation Model Lens

2025-10-27

AI Summary

Existing automated bidding systems are typically scenario-specific, exhibiting limited generalizability. This paper introduces Bid2X—the first foundation model designed for generic bidding environments—unifying the modeling of advertising performance across diverse bid values. Methodologically, Bid2X employs a unified sequence embedding to encode heterogeneous, multi-source data and optimizes via a joint classification-regression loss. Its key contributions include: (1) a dual-attention mechanism capturing both cross-variable and temporal dependencies; (2) a variable-aware fusion module aligning semantic representations of heterogeneous features; and (3) a zero-inflated projection module jointly modeling the sparsity and continuity of performance metrics. Evaluated on eight real-world datasets from Taobao, Bid2X significantly outperforms state-of-the-art baselines. Online A/B tests demonstrate improvements of +4.65% in GMV and +2.44% in ROI, leading to full-scale deployment in production.

Technology Category

• Machine Learning: Online Learning & Bandits
• Search and Optimization: Learning to Search
• Multiagent Systems: Adversarial Agents

Application Category

• User Modeling, Personalization and Recommendation: User modeling for targeted and personalized online advertising
• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Auto-bidding is crucial in facilitating online advertising by automatically providing bids for advertisers. While previous work has made great efforts to model bidding environments for better ad performance, it has limitations in generalizability across environments since these models are typically tailored for specific bidding scenarios. To this end, we approach the scenario-independent principles through a unified function that estimates the achieved effect under specific bids, such as budget consumption, gross merchandise volume (GMV), page views, etc. Then, we propose a bidding foundation model Bid2X to learn this fundamental function from data in various scenarios. Our Bid2X is built over uniform series embeddings that encode heterogeneous data through tailored embedding methods. To capture complex inter-variable and dynamic temporal dependencies in bidding data, we propose two attention mechanisms separately treating embeddings of different variables and embeddings at different times as attention tokens for representation learning. On top of the learned variable and temporal representations, a variable-aware fusion module is used to perform adaptive bidding outcome prediction. To model the unique bidding data distribution, we devise a zero-inflated projection module to incorporate the estimated non-zero probability into its value prediction, which makes up a joint optimization objective containing classification and regression. The objective is proven to converge to the zero-inflated distribution. Our model has been deployed on the ad platform in Taobao, one of the world's largest e-commerce platforms. Offline evaluation on eight datasets exhibits Bid2X's superiority compared to various baselines and its generality across different scenarios. Bid2X increased GMV by 4.65% and ROI by 2.44% in online A/B tests, paving the way for bidding foundation model in computational advertising.

Problem

• Developing a unified bidding function for cross-scenario generalization
• Modeling complex dependencies in heterogeneous bidding environment data
• Addressing zero-inflated data distribution through joint optimization objective

Innovation

• Uniform embeddings encode heterogeneous bidding data
• Dual attention mechanisms capture variable and temporal dependencies
• Zero-inflated projection models unique bidding data distribution