User value identification based on an improved consumer value segmentation algorithm

Abstract

PurposeThe purpose of this study is to propose a new consumer value segmentation method for low-dimensional dense market datasets to quickly detect and cluster the most profitable customers for the enterprises.Design/methodology/approachIn this study, the comprehensive segmentation bases (CSB) with richer meanings were obtained by introducing the weighted recency-frequency-monetary (RFM) model into the common segmentation bases (SB). Further, a new market segmentation method, the CSB-MBK algorithm was proposed by integrating the CSB model and the mini-batch k-means (MBK) clustering algorithm.FindingsThe results show that our proposed CSB model can reflect consumers' contributions to a market, as well as improve the clustering performance. Moreover, the proposed CSB-MBK algorithm is demonstrably superior to the SB-MBK, CSB-KMA and CSB-Chameleon algorithms with respect to the Silhouette Coefficient (SC), the Calinski-Harabasz (CH) Index , the average running time and superior to the SB-MBK, RFM-MBK and WRFM-MBK algorithms in terms of the inter-market value and characteristic differentiation.Practical implicationsThis paper provides a tool for decision-makers and marketers to segment a market quickly, which can help them grasp consumers' activity, loyalty, purchasing power and other characteristics in a target market timely and achieve the precision marketing.Originality/valueThis study is the first to introduce the CSB-MBK algorithm for identifying valuable customers through the comprehensive consideration of the clustering quality, consumer value and segmentation speed. Moreover, the CSB-MBK algorithm can be considered for applications in other markets.

An Adaptive Parameter-Free Optimal Number of Market Segments Estimation Algorithm Based on a New Internal Validity Index

“…Compared with the separation in the BWC, we replace the minimum distance between the centroid of cluster j and the centroids of other clusters in the clustering results with the average distance, which can better reflect the structural relationship between this cluster and other clusters on the whole. As for the cluster connectivity, we use con(j, i) defined in (4) to reflect the connectivity of sample x (j) i , and use con(j) defined in (5) to reflect the connectivity of cluster j. Extremely, if the T nearest neighbors of a sample in its cluster are exactly the same as its T nearest neighbors in the entire dataset, we consider that this sample is correctly classified in the dimension of connectivity; obviously, the larger the con(j) is, the more reasonable the clustering results are.…”

“…The Chinese wine market dataset in this work is provided by the Chinese Grape Industry Technology System, with a total of 2747 effective samples, including two parts: the factors that affect consumers' decision-making and the social demographic characteristics as in [5]. The basic information is listed in Table 13 and the first part is presented through the design of Likert fivecategory attitude scale.…”

“…In the state-of-the-art studies, there are two main ways to determine the ONS: (1) the researchers directly specify an ONS or a range based on their prior knowledge [5], and (2) the combination of a validity index and a clustering algorithm is used to determine the ONS due to a fact that the clustering analysis is the current dominant market segmentation technique and most clustering algorithms also require a pre-specified number of clusters (NC), which makes the process of finding the optimal NC (ONC) coincide with the purpose of determining the ONS [6]. Compared with the former, this combination strategy is more objective and user-friendly; and its basic idea is to perform a clustering algorithm many times with different NCs, choose an appropriate validity index to evaluate the clustering results, and determine the ONC (ONS) according to the evaluation values [7,8], in which the selections of the validity index and the clustering algorithm play the very important roles in whether a reasonable ONC can be obtained.…”