Can Can ChatGPT Evaluate Negative Comments for Restaurant Businesses? A Study on TripAdvisor

Abstract

Purpose - This study investigates the effective use of the artificial intelligence intelligence-based application ChatGPT by restaurant businesses in handling customer complaints and recoveries. Additionally, recommendations to businesses in the context of complaint reasons and the assessment of customers' anger levels by ChatGPT are also examined. Design/methodology/approach - The research was conducted with a case study design using qualitative research method. The scope of the research encompasses the ten most reviewed businesses operating in Isparta. Considering the research objective, negative comments on these businesses are analyzed. A total of 17 negative reviews were accessed on the TripAdvisor website bet ween January 1st and September 25th, 2023. Utilizing research questions previously formulated regarding complaint reasons, recommendations to businesses, recoveries, and anger levels, ChatGPT was employed to inquire about negative reviews. Findings - ChatGPT highlighted "staff," "food," and "service" as complaint elements. Detailed recommendations for businesses were provided by ChatGPT within the context of these areas. Additionally, ChatGPT emphasized remedial strategies such as "apologizing," "speci al offers and discounts," and "regaining efforts" (discounts, special offers, or complimentary items) as recovery strategies. Discussion - In general, ChatGPT can be considered effective in responding to reviews and evaluating negative feedback. Complaint reasons, recommendations to businesses, and recovery strategies regarding negative reviews can be rapidly presented to businesses by ChatGP T. However, due to ChatGPT's evaluation of customers' anger levels solely based on the written content, ratings of 8 8-9 were frequently observed.