User segmentation of visual search service users based on their behaviour and personality for development strategy with two-stage clustering

With the rapid growth in visual search services (VSSs) over the past few years, it is essential to comprehensively understand their specific behaviour and various user segments of VSSs to facilitate visual search system development and innovation. The purpose of this paper is to leverage the current knowledge on visual search service users’ behaviour by adopting the Big Five personality traits as foundation theory. To segment VSS users, a two-stage clustering method is proposed in this study; this method consists of Ward's method in the first stage and continues with the K-means clustering operation. The initial survey was collected from 426 experienced users to identify the users’ personal information and how users engage with VSS. As a result, VSS provider companies can improve profitability and customer loyalty by identifying distinct categories of VSS users and including unique features. The analysis revealed that the users were divided into four groups: ‘Low-use introverted users’, ‘High-use distinctive users’, ‘Easy-going moderate users’, and ‘High-use role model users’. Ultimately, we aid visual search performance through feature development strategies for each group of VSS users. This paper identifies VSS users’ personalities as a different group that could be beneficial for VSS development purposes.