By Michael S. Lin, Yun Liang, Joanne X. Xue, Bing Pan, and Ashley Schroeder
International Journal of Contemporary Hospitality Management Volume 33 Issue 6

Purpose of the study

Recent tourism research has adopted social media analytics to examine tourism destination image (TDI) and gained timely insights for marketing purposes. Comparing the methodologies of social media analytics and intercept surveys would provide a more in-depth understanding of both methodologies and a more holistic understanding of TDI than each method on their own.  This study aims to investigate the unique merits and biases of social media analytics and a traditional visitor intercept survey.

Methodology

This study collected and compared data for the same tourism destination from two sources: responses from a visitor intercept survey (n=1,336) and photos and metadata from Flickr (n=11,775). Content analysis, machine learning, and text analysis were used to analyze and compare the destination image represented from both methods. In particular, Latent Dirichlet Allocation (LDA) was the main topic modeling approach.

Findings

The results indicated that the survey data and social media data shared major similarities in the identified key image phrases. Social media data revealed more diverse and more specific aspects of the destination, whereas survey data provided more insights in specific local landmarks. Survey data also included additional subjective judgment and attachment towards the destination. Together, the data suggested that social media data should serve as an additional and complementary source of information to traditional survey data.

Originality

This study fills a research gap by comparing two methodologies in obtaining TDI:  social media analytics and a traditional visitor intercept survey. Furthermore, within social media analytics, photo and metadata are compared to offer additional awareness of social media data’s underlying complexity. The results demonstrated the limitations of text-based image questions in surveys. The findings provide meaningful insights for tourism marketers by having a more holistic understanding of TDI through multiple data sources.