Predicting Browsers and Purchasers of Hotel Websites: A Weight-of-Evidence Grouping Approach

Edmond H.C. Wu, Chun Hung Roberts Law, Brianda Jiang

Research output: Journal article publicationJournal articleAcademic researchpeer-review

18 Citations (Scopus)

Abstract

A study of the online browsing and purchasing habits of some 1,400 outbound travelers in Hong Kong demonstrates the analytical power of weight-of-evidence (WOE) data mining. The WOE approach allows analysts to identify and transform the variables with the most predictive power regarding the likelihood of tourists' online preferences and decisions. The study found that just over one-third of the respondents browsed hotel-related websites, and about half of those browsers had booked a room on those sites. Browsers in Hong Kong tended to be young, well educated, and well traveled. Those who used the hotel websites for purchases were, of course, part of the browser group, and were likewise relatively well educated. However, one unexpected variable set off those who used the websites for a hotel purchase, the length of their most recent trip. One possible reason is that long-haul tourists want to be sure of their accommodations, or this may reflect hotels' free-night offers. The convenient use of model-based customer segmentation and decision rules would help hospitality practitioners effectively manage their marketing resources and activities, and enhance information-based marketing strategies to attract target customers.
Original languageEnglish
Pages (from-to)38-48
Number of pages11
JournalCornell Hospitality Quarterly
Volume54
Issue number1
DOIs
Publication statusPublished - 1 Feb 2013

Keywords

  • data mining
  • hotel website
  • information technology
  • prediction
  • travel behavior
  • weight of evidence

ASJC Scopus subject areas

  • Tourism, Leisure and Hospitality Management

Cite this