Estimating attractiveness, hierarchy and catchment area extents for a national set of retail centre agglomerations
Les Dolega; Michalis Pavlis; Alex Singleton (2016). Journal of Retailing and Consumer Services, 28, 78-90. DOI: 10.1016/j.jretconser.2015.08.013
Abstract
There is a legacy of research aiming to conceptualise and empirically estimate retail store catchment areas, however, a dearth that frames such considerations within the context of retail agglomerations and their position within regional or national networks. As a result, this paper provides an extension to single store or shopping centre retail catchment estimation techniques, and presents an empirically specified and tested production constrained model for a national network of retail centres in the UK. Our model takes into account the spatial interactions between potential customers and a hierarchical network of retail centres to estimate patronage probabilities and catchment extents. The model is tested for a large metropolitan area vis-à-vis real world shopping flows recorded through a survey of shoppers. Finally, we present an open source software tool for custom model fitting, and discuss a range of theoretical and empirical challenges that such a model presents.
Extended Summary
This research develops a comprehensive model to estimate retail catchment areas for over 1,300 UK town centres, addressing the gap in understanding how shopping destinations compete within national retail networks. The study extends traditional single-store catchment analysis to examine retail agglomerations (groups of shops in town centres) and their hierarchical relationships across England and Wales. Using a modified Huff gravity model, the research incorporates multiple factors that influence shopping behaviour, including retail centre attractiveness, distance decay effects, and competition between destinations. The attractiveness measure combines quantitative attributes such as retail centre size (number of outlets), retail diversity, proportion of leisure units, presence of anchor stores, and vacancy rates. The methodology calculates shortest road distances between Lower Super Output Area centroids and retail centre boundaries, then applies different distance decay parameters based on centre hierarchy and accessibility. The model generates patronage probabilities for each area, which are used to delineate primary (50% threshold), secondary (25% threshold), and tertiary (10% threshold) catchment areas. Validation against real-world shopping survey data from Birmingham demonstrates that the model correctly predicts shopping destinations for 78% of areas after calibration adjustments. The research reveals significant spatial variations in catchment sizes, with higher-order centres like Manchester and Liverpool drawing customers from much larger geographical areas than local district centres. The study identifies challenges in densely populated urban areas where intense competition between multiple centres creates complex patronage patterns that are difficult to model accurately. Some areas, particularly around Birmingham and the West Midlands, show gaps where no centre achieves sufficient patronage probability to claim primary or secondary catchment status, highlighting the limitations of gravity-based approaches in highly competitive retail environments. The research provides an open-source R package enabling practitioners and researchers to apply the methodology to different geographical contexts or update parameters as retail landscapes evolve. This work has significant implications for retail planning, town centre management, and understanding the impact of changing consumer behaviour, including the growth of online shopping, on traditional retail hierarchies and catchment patterns.
Key Findings
- A national model successfully estimates retail catchments for 1,312 UK town centres using modified Huff gravity modelling techniques.
- The model correctly predicts shopping destinations for 78% of areas when validated against real-world Birmingham survey data.
- Higher-order retail centres like Manchester draw customers from significantly larger geographical areas than local district centres.
- Dense urban areas with multiple competing centres present modelling challenges, leaving some areas unassigned to any catchment.
- An open-source software tool enables practitioners to apply the methodology for retail planning and town centre management.
Citation
@article{dolega2016estimating,
author = {Les Dolega; Michalis Pavlis; Alex Singleton},
title = {Estimating attractiveness, hierarchy and catchment area extents for a national set of retail centre agglomerations},
journal = {Journal of Retailing and Consumer Services},
year = {2016},
volume = {28},
pages = {78-90},
doi = {10.1016/j.jretconser.2015.08.013}
}