Tiered matching model considering quality compatibility in two-sided markets

This paper explores how platforms can effectively match heterogeneous users within tiers to increase matches and profits in two-sided matching markets. With the rapid expansion of digital platforms, user preferences and quality have become increasingly differentiated, leading to market congestion, search frictions, and inequality among users of different levels. To address these issues, we propose a centralized matching model that dynamically estimates user popularity and adjusts acceptable match sets. The model incorporates a reciprocal recommender to reduce inequality in matching and avoid congestion from users intensively sending requests to higher-tier counterparts. By analyzing user recommendation lists, the model quantifies the impact of market thickness, congestion, and matching security on participant utility. We then establish integer linear programming models to solve the stable matching optimization problem, maximizing platform revenue and social welfare. Numerical experiments validate the effectiveness of the proposed approach, demonstrating that a soft-tiered strategy using a reciprocal recommender optimizes the distribution of matching pairs across different levels, significantly increasing matches for middle and lower-tier users while enhancing platform revenue and social welfare. The model provides decision support for platforms to optimize tiered strategies and assists regulatory agencies in monitoring potential social welfare losses from excessive segmentation. The findings highlight the importance of considering user heterogeneity and market characteristics in designing effective matching mechanisms for two-sided platforms.