An Automated Data-Driven Approach for Product Design Strategies to Respond to Market Disruption Following COVID-19

Abstract

Online reviews provide a source to identify customer needs. While many studies have analyzed online reviews during the pandemic, it is worth noting that many customer preference studies in this period were not conducted within a product design context. The societal challenges presented by the prolonged COVID-19 pandemic, spanning nearly three years, have significantly impacted all facets of the population in a manner unparalleled in recent decades. Therefore, this research delves into the post-COVID-19 landscape, examining shifts in consumer preferences for diverse product features through an analysis of online reviews. Our framework unfolds in five stages: First, it collects online reviews and second, delves into customer interest in product features. Third, it analyzes customer sentiments toward these features. Fourth, employing interpretable machine learning techniques, it determines the significance of each feature. Fifth, an importance-performance analysis (IPA) and Kano models are utilized to formulate and analyze product strategies. The developed method is assessed on two real-world datasets—smartphone and laptop reviews. The results reveal that after the pandemic, customer satisfaction for the screen and camera in smartphones decreased, whereas it increased for those in laptops. In addition, the importance of battery features in smartphones and laptops has increased. These insights will aid companies in promptly formulating strategies to navigate dynamic market environments.