Track: Data Analytics

Abstract

A housing equipment company under consideration has a website, which consists of thirty blocks of pages, such as Top Page, Catalog, Products, Reform and Showroom, among others, having about 100,000 sessions per day. A session is said to make a conversion if it accesses one of the three sets of designated pages in Catalog, Reform, and Showroom, indicating that the session caller is likely to take a positive action soon for the businesses of the company. Accordingly, it is important to understand the structure of page access behaviors of sessions at the website and to identify unique characteristics of sessions making a conversion. For this purpose, this paper examines the transitions of sessions among page blocks, where sessions are categorized by referrers (Google and Yahoo with distinction of whether or not an access is through a commercially available banner advertisement), devices (PC, smart phone and tablet), weekdays vs. weekends, and landing pages, so as to capture the characteristics of conversion sessions. Based on this analysis, some suggestions will be made for improving the website so as to facilitate more sessions toward conversion.