Understanding the patterns behind purchasing capability: A case study of smartphone consumers

Purchasing behavior analysis plays a crucial role in pricing, recommendation, and market strategy designing. One of the fundamental questions that arises in purchasing behavior analysis is to understand, characterize and estimate purchasing capability. In this article, we investigate the patterns of purchasing capability from the perspective of network usage of smartphone consumers based on a large-scale usage detail records (UDRs). First, the purchasing capability of smartphone consumers are divided into three (high/middle/low) levels according to corresponding device retail price around observation period. Then we pairwise integrating interest and temporality information, and conduct a clustering analysis for each purchasing capability level users. Finally, the profile of each community is extracted by a visualization process. The learned patterns can not only illustrate how users behave on interest, but also describe their network usage preference on temporality, providing fine-granularity to understand the dynamics behind each purchasing capability group. Moreover, inspired by learned patterns, we find network usage distribution on temporality, spatiality and interest can serve as good indicators for estimating users purchasing capability. Our work has broad applicability in fields such as pricing, recommendation, and market strategy designing.