Refining spatial models of individual level consumer channel and store choice behaviours

Context: The Centre of Spatial Analysis and Policy at the University of Leeds has been known for its strength in applied retail modelling and work within location planning for the grocery market. The current retail spatial interaction model used allows us to see the grocery market as it is in terms of the number of retailers, their store sizes, and the types of populations residing in areas they are located. Using the model allows retailers to see how well they are performing, who their competitors are, and what they could do to improve revenue. However, the model requires refinements to include the prediction of consumer store and channel choices, as differs among populations and adding such individuality to a SIM requires significant research and justification within methodologies. This project will have access to customer purchasing and behaviour data derived from the Sainsbury's loyalty card scheme. This data will allow this project to capture, understand, and model consumer behaviours at the consumer level. This will be achieved through using the data in a custom-built agent based model and applying the knowledge and strength of spatial interaction models.
Objectives: - To expand knowledge on consumer store and chancel choice behaviours. - To build an agent-based model that utilises the data derived from Sainsbury's loyalty card scheme for research, along with the strengths of the spatial interaction model.
- To publish a project that includes new insights and understandings into the consequences of individual decision and behaviour in the retail world. Method: To carry out this project, a thorough literature review will be produced that discusses how retail has changed over the past decades, what new channels of retailing are available to consumers, and the possible impacts of this upon retail site location choices. This will include a section about the different ways consumers shop, how their shopping frequencies, missions, and budgets differ. It will highlight the importance of acknowledging individualised consumer behaviours when retailers are looking to expand, change, or relocate their stores. The next section will involve transforming the loyalty card scheme data into a format that is easy to incorporate into a model. A custom-built agent based model will be designed to include the different behaviours and channel choices of consumers. This will be achieved by drawing upon the methods used in spatial interaction models that have been successful and thoroughly developed. The model will involve individual consumer level choices and behaviours, the type of stores and channels of retail, and an element of geography. This is subject to change as the project is much more complex, however this is a simple summary of the proposed project and what it will involve.