Modeling Consumer Behavior in Shopping Malls: Dynamic Fatigue, Social Interactions, and Data-Driven Layout Optimization

Brick-and-mortar shopping malls play a vital role in retail ecosystems, serving as platforms where tenant layout and design choices significantly influence consumer traffic patterns. These traffic patterns are crucial for determining store-level exposure, rental negotiations, and overall mall profitability. However, predicting granular, store-level consumer traffic remains a challenge due to the complex interplay of human behavior, including physical fatigue, social dynamics, and decision-making under spatial constraints. Existing models often rely on aggregate visitation data, overlooking critical behavioral factors such as how shoppers tire over time or how groups navigate collectively. This leads to suboptimal mall designs and inefficient tenant placements. This project aims to address these gaps by developing a behaviorally grounded structural model of consumer movement that incorporates two key dynamics: (1) physical fatigue and (2) social shopping behavior. Using a unique dataset of 17.3 million second-by-second shopping trajectories collected from four large malls, we will analyze how shoppers' decisions are influenced by accumulating fatigue, rest opportunities, and group coordination costs. The dataset, captured through AI-powered video sensors, provides unprecedented spatial and temporal resolution, allowing us to observe complete shopping journeys, group compositions, and rest patterns while preserving privacy. Our research focuses on three key questions: 1. How do physical fatigue and rest affect shoppers' movement, store choices, and exit decisions? 2. Can we reliably identify social shopping groups from movement patterns, and how do their behaviors differ from solo shoppers? 3. Which tenant layouts and recovery zone placements (e.g., seating and food courts) maximize store-level exposure and overall shopper engagement? To answer these questions, we will develop a structural econometric model that explicitly accounts for fatigue accumulation, recovery, and group heterogeneity. Using this model, we will conduct counterfactual simulations to test how different mall layouts and tenant placements impact shopper behavior. These simulations will provide actionable insights for optimizing mall design, balancing trade-offs between driving store-level exposure and minimizing shopper fatigue. Our findings will help mall operators design layouts that enhance shopper experience, increase dwell time, and improve profitability. By combining advanced modeling techniques with high-resolution trajectory data, this project bridges the gap between behavioral theory and practical retail design, offering a robust framework for data-driven decision-making in the retail sector.