ChatGPT as a Sentiment Analysis Tool: An Assessment Based on Negative Reviews of High-End UK Restaurants
摘要
Customer reviews may play a significant role in influencing decision making; thus, their constant monitoring and handling can be a challenge for hospitality marketers. Given the great potential Artificial Intelligence (AI) has to impact hospitality operations, this study aims to assess the utility of Generative AI tools (GenAI) in analyzing customer reviews and effectively reflecting customer sentiments within a very special context, i.e. high-end restaurants. By focusing on reviews for 30 Michelin-starred UK restaurants in London and TripAdvisor’s two lowest ratings categories, ChatGPT is employed as a sentiment analysis tool. Preliminary findings show that ChatGPT can highlight prevailing sentiments and also identify critical aspects and sources of (dis)satisfaction, thus helping hospitality managers to effectively monitor customer reviews. In terms of practical implications, while restaurant managers still need to handle every customer review in a personalized manner, they can use GenAI for continuous monitoring of customer expectations, criteria, satisfaction levels and sentiments. Managerial decisions and tasks to ensure satisfaction for their highly demanding clientele may then become easier. Moreover, given the produced insights, identification and prioritization of key areas of differentiation from competition become easier. Respective knowledge can be used as input for pre-trained GenAI tools and chatbots used in customer service operations.