Exploring dynamic effects on classifying service quality attributes under the impacts of COVID-19 with evidence from online reviews

Purpose The purpose of this paper is exploring the effects of segment dynamic and temporal dynamic triggered by the COVID-19 pandemic on classifying service quality attributes, thereby formulating improvement strategies to satisfy customers and respond to threats. Design/methodology/approach Given the dynamics of the attractive quality theory, this paper designs a framework with four phases by embedding techniques of text mining and deep learning based on evidence from online reviews. Findings This paper figures out dynamics of service quality attributes for distinct segments and their dynamic proportion along with different stages of the pandemic. Another finding demonstrates segment dynamic and temporal dynamic effects of sentiments toward service quality attributes on customer satisfaction under the impacts of pandemic. Classification results and improvement strategies are derived for varying segments at different pandemic situations. Practical implications This paper reveals dynamic effects on classifying service quality attributes, which contributes to assisting hospitality practitioners from different segments in improving service quality when facing with the challenges of crisis and potential risks. Originality/value Given hospitality industry is time- and segment-sensitive, the authors achieve the quantification of dynamics of attractive quality theory and extend it into hospitality marketing and crisis management from the perspective of dynamics with evidence from online reviews.