Group14 Project Overview
Project Motivation & Problem Statement
Overall, the aim of the project is to develop a geospatial dashboard that would facilitate the ABC retail’s evaluation of their current and different potential outlet siting locations through better prediction of the catchment and number of customers it can potentially attract.
Therefore in order to start our project, our team thought best to first understand the dynamics of the different outlets for ABC retail to find the proper motivation for the project. Our team decided to understand the public demand for outlets better by looking at the geographic distribution of customers (i.e. where they come from) for each of the outlets. Our initial hypothesis was that most of the customers of a particular outlet would be residents within the immediate neighbourhoods of the outlet. In order to validate this hypothesis, a dashboard was constructed on Tableau to reflect the geographic view of the outlets location and the respective customer distribution according to subzones.
Our hypothesis proved to be valid for most of the outlets. We can use outlet 01 as an example to illustrate our hypothesis. As can be seen in Figure 1, the choropleth map reflects darker shades of blue in subzones within the vicinity of outlet 01. This is indicative that more customers of outlet 01 come from these subzones as well.
However there are some outlets which are an exception to our hypothesis. Let us take a look at outlet 12 as an example. Referring to Figure 2, we can see that the subzone comprising of the highest population of outlet 12 customers is Tampines East. The subzone with the second highest population is Serangoon Garden. When taking a look at the geographic locations of Tampines East and Serangoon Garden, they are further away from outlet 12 as opposed to many other subzones. Furthermore, Tampines East and Serangoon Garden themselves are situated far apart from each other and they are closer to their own regional outlets such as outlet 23 and outlet 21 respectively.
So why are these customers are not visiting outlets within their vicinities are travelling to other outlets instead? Why is the customer distribution for certain outlets scattered across subzones which are far apart from each other? These are some of the questions that our group started asking ourselves upon discovering the exceptions to our hypothesis. The only way to answer all these questions would be to understand the patterns of customer-flow/demand for different outlets, which can be explained through certain factors which determine the attractiveness of an outlet to customers. Variables which affect attractiveness of an outlet include distance from the customer’s location to the outlet, the scope of service of the outlet or even the commodities that are available near the location of the given outlet.
Therefore we intend to first use the Multiplicative Competitive Interaction Model (MCI Model) to identify how much of an effect the individual variables have on the attractiveness of a given outlet. Followed by that, we then use the MCI Model to apply those variables in the context of estimating the customer-flow/demand for an outlet.
Constant change management in the location and resources within the outlets have led to the dire need of a reliable and standardized method which can measure the effectiveness of past policies, as well as an accurate predictive model to conduct what-if analyses for future. Thus this research paper dives deeper into finding the perfect model fit for ABC retail which can be used to predict and measure the attractiveness of an existing outlet or potential outlet in a given vicinity.
Project Objective
Develop a geospatial dashboard that would facilitate a private retail company’s evaluation of different outlet siting locations through better prediction of the catchment and number of customers it can potentially attract.
