Good practice for product management decision-making: Using Amazon's dataset and ChatGPT

This study examines how Chat Generative Pre-trained Transformer (ChatGPT) can be effectively utilised as a good practice to analyse Amazon reviews within the creative and cultural industry category of Design, particularly Arts, Crafts & Sewing. By focusing on specific products and reviews extracted from a dataset containing over 800,000 products and nearly nine million reviews, ChatGPT identified common themes and customer issues such as software problems and product reliability. When we compared ChatGPT's results with those from manual data reviews, we found that ChatGPT was adept at identifying main topics but sometimes missed detailed insights or altered the original reviews when providing examples. This indicates that while ChatGPT can quickly highlight important areas for product managers, it should be used in conjunction with human analysis to achieve a comprehensive understanding. Nonetheless, the results demonstrate that when considering group creativity, ChatGPT can be a valuable member.