Optimal Auction Design in the Joint Advertising

2025-07-10

AI Summary

In online advertising, existing combinatorial auction mechanisms suffer from suboptimal allocation and revenue due to their neglect of ad bundle structures. This paper addresses both single-slot and multi-slot combinatorial advertising settings: we first propose the first theoretically optimal auction mechanism for the single-slot case; second, we design BundleNet—a neural network framework that jointly models bundle structures while enforcing incentive compatibility and individual rationality constraints. BundleNet explicitly captures advertiser synergies via end-to-end learning—achieving near-optimal revenue in the single-slot setting and significantly outperforming baseline methods in multi-slot settings. Experiments demonstrate substantial improvements in platform revenue and allocation efficiency. Our work establishes a new paradigm for combinatorial advertising mechanism design, bridging theoretical optimality with practical implementability.

Technology Category

• Machine Learning: Online Learning & Bandits
• Search and Optimization: Learning to Search
• Multiagent Systems: Mechanism Design

Application Category

• Economics, Online Markets and Human Computation: Advertising auctions, pricing, markets, and exchanges
• User Modeling, Personalization and Recommendation: User modeling for targeted and personalized online advertising
• Search and Retrieval-Augmented AI: Ad search and search for Web retail

Abstract

Online advertising is a vital revenue source for major internet platforms. Recently, joint advertising, which assigns a bundle of two advertisers in an ad slot instead of allocating a single advertiser, has emerged as an effective method for enhancing allocation efficiency and revenue. However, existing mechanisms for joint advertising fail to realize the optimality, as they tend to focus on individual advertisers and overlook bundle structures. This paper identifies an optimal mechanism for joint advertising in a single-slot setting. For multi-slot joint advertising, we propose extbf{BundleNet}, a novel bundle-based neural network approach specifically designed for joint advertising. Our extensive experiments demonstrate that the mechanisms generated by extbf{BundleNet} approximate the theoretical analysis results in the single-slot setting and achieve state-of-the-art performance in the multi-slot setting. This significantly increases platform revenue while ensuring approximate dominant strategy incentive compatibility and individual rationality.

Problem

• Design optimal auction for joint advertising bundles
• Improve revenue and efficiency in multi-slot advertising
• Ensure incentive compatibility and individual rationality

Innovation

• Optimal mechanism for single-slot joint advertising
• BundleNet: bundle-based neural network for multi-slot
• Ensures revenue, incentive compatibility, and individual rationality