بهینه سازی توزیع مکانی بازاریاب ها با استفاده از تحلیل های GIS

Introduction

Manufacturing and distribution companies (such as those in the food, cosmetics, hygiene products, and pharmaceutical sectors) are seeking to reduce labor costs and marketing time. Since marketers are the primary link between retailers and manufacturing and distribution companies, proper management and performance monitoring of marketers can lead to reduced distribution costs and, consequently, increased customer (retailer) satisfaction.

The distribution of marketers within a region depends on the locations of the retailers, ensuring a balance between the number of marketers and retailers. The conventional method for assigning marketers to each area involves segmenting or clustering retailers based on geographical boundaries and then assigning each area to a marketer. In the past, distribution companies manually delineated these areas on large paper maps. However, today, various technologies have been developed to facilitate the management of marketers, improving their service to customers and enhancing their efficiency.

Methodology

The first step involves preparing the main roads and constructing the network of roads for the study area. In this phase, a road network is created using the network analysis capabilities of GIS. To cluster retail points and create zones, it is necessary to extract smaller sub-areas from the road network. These sub-areas form the basis for creating larger geographic zones (clusters) for assignment to marketers.

To determine the number of retail points in the sub-areas, spatial joining is required. In other words, this analysis identifies the retail points within each sub-area by spatially overlapping the retail points with the sub-areas. Then, the centroid of each sub-area is determined, and the retail points within each sub-area are connected to this centroid. To aggregate the sub-areas and extract larger geographic zones based on the number of retail points, a regular grid of points is created. The average distance of these grid points from the centroids of the sub-areas is calculated through spatial allocation analysis, and based on this, along with the number of marketers needed, the cluster centers are determined. It is important to note that during the spatial allocation analysis, the clusters are optimized based on a specific threshold. Finally, each sub-area is assigned to a cluster, and each cluster is assigned to a marketer.



Results and discussion

Clusters represent the spatial zones designated for marketers. In other words, the outer boundary of each cluster defines the marketer's working area, and the points within it are the visit points for a marketer. Over time, due to the dynamic nature of retail points (such as relocations or changes in usage), there is no need for re-clustering. It is sufficient to update the visit routes according to the addition or removal of retail points without requiring re-clustering.

In this study, it is assumed that marketers travel on foot between retail points for visits, so the direction and type of road are not considered. Therefore, routes are calculated as the shortest distance between points within each cluster. Obviously, the less variance there is between the determined routes within clusters, the more optimal the clustering. In other words, marketers will cover equal distances within clusters to visit retail points. Equal route lengths ensure fair workload distribution among marketers and prevent spatial disparity among them. The results showed that the proposed method effectively considers this aspect.

One crucial property required for clusters is the ease with which their boundaries can be identified by a human user. Human error in this aspect could result in overlapping customer visits, which is entirely unacceptable. Therefore, defining the boundaries of clusters is vital, and they must be easily recognizable with minimal fragmentation. By comparing the compactness of the clusters, it is evident that the proposed method produces the most compact clusters with the least deviation from the mean compactness across clusters. The compactness property also facilitates shorter routes within each cluster by preventing cluster dispersion.







Conclusion

Dividing urban areas into zones or clusters such that each zone contains an equal number of retail points is a crucial issue for distribution and marketing management in distribution companies. This study proposes a method for marketing zoning in the study area. The method is based on GIS network analysis, ensuring that a series of sub-areas are arranged to achieve maximum compactness and minimal discontinuity within each cluster. The results showed that the proposed method outperforms existing methods in terms of shortest travel distance, minimal border fragmentation, minimal discontinuity, and maximum compactness. Minimizing border fragmentation reduces errors in identifying and defining the boundaries of marketing zones.

It is important to note that this algorithm was implemented only in District 6 of Tehran, and the results may vary for other areas or cities with different network structures. Therefore, it is recommended to apply this method to other regions or cities as well. The method has a relatively simple updating process, making it easy to re-cluster and reroute with the addition or removal of points within the zones. Additionally, managing marketers becomes easier due to the limited scope of their operational areas. As a future research direction, the generated zones can be used as geographical fences (geo-fences) in a mobile GIS system, where marketers receive notifications upon entering or leaving a zone.



Keywords

Location-based marketing, spatial distribution of marketers, geographic information system, network analysis



Keywords

Location-based marketing, spatial distribution of marketers, geographic information system, network analysis

Keywords

Location-based marketing, spatial distribution of marketers, geographic information system, network analysis

Keywords

Location-based marketing, spatial distribution of marketers, geographic information system, network analysis

Keywords

Location-based marketing, spatial distribution of marketers, geographic information system, network analysis

Keywords

Location-based marketing, spatial distribution of marketers, geographic information system, network analysis

Keywords

Location-based marketing, spatial distribution of marketers, geographic information system, network analysis

Keywords

Location-based marketing, spatial distribution of marketers, geographic information system, network analysis

Keywords

Location-based marketing, spatial distribution of marketers, geographic information system, network analysis