Customer mobility and congestion in supermarkets

Area of Science:

• Spatial analysis
• Human mobility modeling
• Retail analytics

Background:

• Population-level mobility models are crucial for applications like urban planning and trade flow analysis.
• Existing models primarily focus on large spatial scales (intracity to intercountry).
• Understanding human movement at smaller scales, such as within retail environments, remains underexplored.

Purpose of the Study:

• To investigate the applicability of human mobility models at a small spatial scale, specifically customer flow within supermarkets.
• To adapt and apply gravity and intervening-opportunities models to empirical supermarket mobility data.
• To explore the use of mobility models for optimizing supermarket layouts and reducing customer congestion.

Main Methods:

• Utilized anonymized, ordered customer-basket data to infer empirical mobility flow within supermarkets.
• Applied variants of the gravity model and intervening-opportunities models (including the extended radiation model) to fit observed customer flow.
• Modeled supermarket zones as queues and employed a gravity model with a simulated-annealing algorithm to identify low-congestion store layouts.

Main Results:

• Doubly-constrained gravity and extended radiation models successfully estimated 65%-70% of in-supermarket customer flow.
• Optimized store layouts, derived from mobility modeling, demonstrated reduced congestion compared to original layouts.
• Popular zones in optimized layouts were frequently located along the store's perimeter.

Conclusions:

• Human mobility models are effective for analyzing and predicting customer movement within supermarkets.
• Mobility modeling offers a data-driven approach for retailers to design store layouts that minimize congestion and enhance customer experience.
• This research provides valuable insights into small-scale human mobility patterns and their practical application in retail environments.