Coordinated Auto-Bidding Boosts Return-on-Spend and Total Value, ICML 2026 Study Finds

A new paper accepted at ICML 2026, 'On the Coordination of Value-Maximizing Bidders,' investigates how multiple auto-bidders can coordinate to improve advertising outcomes. The authors—Yanru Guan, Jiahao Zhang, Zhe Feng, and Tao Lin—formalize a coordination mechanism where only the highest-value bidder among a group competes in an online auction, while the others voluntarily withdraw. This approach contrasts with the standard assumption of independent bidding in auto-bidding literature.

The research shows that such coordination dominates independent bidding for a broad class of auto-bidding algorithms. Specifically, it improves both Return-on-Spend (RoS) compliance—ensuring bidders don't exceed budgets—and the total value accrued by the participating bidders or ad campaigns. The team validated their theoretical model with simulations on synthetic and real-world datasets, demonstrating practical robustness. Key scenarios include collaborative bidding managed by a third-party agent and strategic bid selection for multiple campaigns under a single advertiser.

For tech‑savvy professionals, this work has immediate implications: advertisers running multiple campaigns can coordinate bids to avoid internal competition, boosting ROI without increasing spend. Third‑party bidding platforms can adopt this mechanism to deliver better RoS for clients. The approach is practical, requiring no changes to auction infrastructure—only coordination among bidders. As auto‑bidding becomes standard, this paper offers a proven strategy to increase efficiency while maintaining budget compliance.