Leveraging Data-Driven Analysis To Explore Restaurant’s Market Segmentation in Indonesia

Abstract

Background: The internet and Electronic Word of Mouth (eWOM) have transformed consumer behavior in choosing dining options in Indonesia’s culturally diverse culinary landscape, yet research leveraging eWOM data to understand consumer preferences remains limited.
Purpose: This research is conducted to develop the restaurant’s market segments based on customer ratings in Indonesia using a data-driven approach.
Design/methodology/approach: The data is crawled from notable review sites in Indonesia which consist of 35.811 restaurants across Indonesia. Two clusters were generated using TripAdvisor data, encompassing users' ratings for Food, Service, Value, Atmosphere, and overall satisfaction. The research successfully segmented the Indonesian restaurant market based on customer ratings using the K-Means clustering approach.
Findings/Result: Cluster 1 valued food quality and cared about service and value. Meanwhile, Cluster 2 focused more on good service, followed by food and the restaurant’s atmosphere.
Conclusion: The research successfully segmented the Indonesian restaurant market based on customer rating, helping restaurant managers understand what customers prefer in Indonesia’s varied food scene. This can assist marketers in creating effective marketing strategies, such as advancing product development, enhancing food quality, and optimizing service offerings to better fulfill the needs and expectations of their target audience.
Originality/value (State of the art): This study can pave the way for further investigation into market segmentation in Indonesia's restaurant sector. While similar approaches have been applied in studies of other countries, the Indonesian market is unique and has distinctive features that haven't been examined in previous research. Therefore, these insights can illuminate the segmentation of the restaurant market in Indonesia.

Keywords: consumer rating, customer segmentation, digital marketing, k-means clustering, big data