Market segmentation via structured click stream analysis

Abstract

Accurate market segmentation has been the basis for successful customization of products and services. To date, however, the marketing management literature has focused mainly on the exploration of segmentation variables, but lagged behind in the development of practical means for segmentation mechanisms using contemporary information technology. Motivated by this shortcoming, the current study attempts to devise an effective method that allows for systematic collection and analysis of online customers’ click stream data to facilitate market segmentation. Cohen’s CAD theory was employed in conjunction with artificial neural network models to provide the analytical foundation of this research. To test the effectiveness of the proposed method, a sizable online field experiment utilizing a disguised 7‐ELEVEN Website was conducted, and 912 useful click streams collected. The results from the subsequent data analysis supported the feasibility of the current work, but also identified the needs for further study.

Shopping Hard or Hardly Shopping: Revealing Consumer Segments Using Clickstream Data

“…consumers are using search engines to look for a brand and products (93.42%), and access the website via a mobile device (78.56%). This cluster's consumers are characterised by high engagement with e-commerce website; we observe a moderate number of past visits (3), a high number of pages viewed(25) and time spent per page (21.29'') as well as overall during the visit (9.90'). All consumers in the 'Visiting with a Purpose' cluster open the cart (100%), and 15.29% of them make a purchase, which implies goal-directed online behaviour.Last but not least, the 'Impulsive Trying' cluster is the smallest consumer group with N=370 unique visits.…”

“…Importantly, it is linked to real-time events, and thus, it can reduce the level of risk, improve profitability and efficiency of marketing actions [19] [23]. With research now advocating the advantages of clickstream data over other big data types [17] [25], 'the need to put them into scrutiny for useful applications is perfectly understandable' [23].…”

“…As demonstrated by ASOS, one potentially useful application of clickstream data in marketing research and practice is consumer segmentation [13] [25]. Consumer segmentation is defined as the division of consumers into groups of buyers who share distinct characteristics and behaviours, that might require separate products or marketing mixes [26] [27].…”

Proposing the content perception theory for the online content industry – a structural equation modeling

“…Such a perspective will not only directly echo the arguments recognizing the importance of content strategy manipulations (Dreze and Zufryden, 1997;Eisenmann, 2002), but also help to solidly clarify the effects of Web content perceptions molded by content providers upon Web users and thus provide a basis for their content policy making. In addition, although there have been many antecedent factors shown in the literature including user characteristics such as involvement, knowledge, and experience, we believe that investigating content-related variables would also contribute to the content strategy formulations of content providers for their common inadequacies regarding user's personal information (Chang, 1998;Huberman et al, 1998;Wen and Peng, 2002). In order to systematically investigate how content perceptions affect Web users, we adopt the conceptual framework proposed by Singh and Dalal (1999) to structure the relationships among Web content exposure, perceptions, attitudinal and behavioral reactions.…”

Integrating Decision Trees and Cognitive Maps for Market Segmentation in Service Sector

“…Apart from using their own database, companies are collecting data from purchasing transactions, credit card receipts, membership history and even internet usage preferences [13]. Customer information analysis becomes more and more vital for the companies.…”