SACO: Sequence-Aware Constrained Optimization Framework for Coupon Distribution in E-commerce

TL;DR

SACO introduces a sequence-aware approach to constrained coupon distribution, addressing the limitations of single-round and uplift-only methods by explicitly modeling multi-round user interactions under a budget $B$. It combines trajectory-based training with a constrained-optimization Decision Transformer and a SLSQP-based dual variable optimization to produce a scalable, end-to-end policy. Empirical results across public, synthetic, and industrial datasets show SACO achieving an average revenue improvement of $3.60\%$ with robust ROI and BARate gains, while ablations confirm the importance of budget constraints and long-term return guidance. The framework offers practical applicability for real-world marketing with efficient inference and potential for extending to broader constrained optimization problems.

Abstract

Coupon distribution is a critical marketing strategy used by online platforms to boost revenue and enhance user engagement. Regrettably, existing coupon distribution strategies fall far short of effectively leveraging the complex sequential interactions between platforms and users. This critical oversight, despite the abundance of e-commerce log data, has precipitated a performance plateau. In this paper, we focus on the scene that the platforms make sequential coupon distribution decision multiple times for various users, with each user interacting with the platform repeatedly. Based on this scenario, we propose a novel marketing framework, named \textbf{S}equence-\textbf{A}ware \textbf{C}onstrained \textbf{O}ptimization (SACO) framework, to directly devise coupon distribution policy for long-term revenue boosting. SACO framework enables optimized online decision-making in a variety of real-world marketing scenarios. It achieves this by seamlessly integrating three key characteristics, general scenarios, sequential modeling with more comprehensive historical data, and efficient iterative updates within a unified framework. Furthermore, empirical results on real-world industrial dataset, alongside public and synthetic datasets demonstrate the superiority of our framework.

Figures (5)

• Figure 1: Complex and Meaningful Coupon Distribution.
• Figure 2: Single-Round Scenarios in Existing Research.
• Figure 3: Constrained-Optimization Decision Transformer
• Figure 4: Evaluation Results
• Figure 5: Ablation Study on MT Dataset.