Panic Station: Consumer Sentiment Analysis of the Evolving Panic Buying During the COVID-19 Pandemic

Introduction

Technological advances have transformed the structures of various organizations, facilitating the way firms communicate with their audiences. The impact of social networks has generated a virtual world based on communities and platforms (Sann & Lai, 2020; Tuncer, 2021), where users create, share and exchange opinions, emotions, and experiences (Liao et al., 2021). By reviewing and rating products, services, and brands, Internet users generate data that constitute a valuable source of information for companies and business decision-makers (Kauffmann et al., 2019). E-commerce has emerged as a new mechanism in the digital economy age.

In this regard, e-commerce is one of the most widely used digital tools today. Not only because the incursion into the digital ecosystem has become a global necessity, but also because of the solution, it has represented different problems such as the pandemic caused by COVID-19, where many consumers prefer online shopping as a more convenient and secure method (Hasiloglu & Kaya, 2021). As part of this transformation, a new e-commerce paradigm emerged, called social commerce or s-commerce (Qalati et al., 2021). The first hint of the implementation of s-commerce emerged in 2005 on Yahoo! when users of that platform issued opinions and ratings about products and shopping experiences, providing textual reviews (Liao et al., 2021). Nowadays, s-commerce represents an exchanging system of products or services through social networks. S-commerce is a combination of commercial and social activities, as it connects sellers and consumers through websites and social platforms like Twitter, Facebook, YouTube, and Tik-Tok (Abed, 2020).

Numerous online social networking platforms represent an opportunity to immerse oneself in the new socio-digital era due to their great popularity and ease of use in personal and business terms. Before purchasing any product or service, consumers are accustomed to seeking evaluations online, transforming and empowering the “word of mouth” (WOM) to “electronic word of mouth” (e-WOM). According to Zhou et al. (2019), approximately 83.5% of online shoppers value the opinions of other consumers. In turn, companies can leverage this information to expand their knowledge of consumer behavior and have elements to formulate marketing strategies that contribute to organizational growth and competitive advantage (Lin et al., 2020; Sann & Lai, 2020).

Due to such information being generally unstructured and its manual processing would be a complex task (Jain et al., 2021), in recent years, sentiment analysis or opinion mining constitutes an efficient technique, by extracting, classifying, and analyzing content in Big Data; sentiment analysis has become an increasingly useful tool for processing consumer attitudes and emotions in online reviews (Chaturvedi et al., 2018). In this regard, consumer sentiment analysis allows processing large amounts of data to extract and detect polarity (positivity, neutrality, and negativity) in product reviews (Do et al., 2019)which can help Internet users to make purchasing decisions and also to rationalize them, as well as companies to a better understanding of consumer behavior, identifying problems, experiences, satisfaction, trust, and position in front of their competitors. All this could provide valuable information to improve the service (Jain et al., 2021).

Consumer sentiment analysis can be done at three levels: document, sentence, and aspect (Liu & Zhang, 2012). This paper will analyze consumer polarity towards highly demanded products during the SARS-CoV-2 pandemic, and to achieve this, the authors will identify emerging trending topics related to product purchases. Each dataset (trending topic) will be analyzed using a lexicon approach to capture positivity, neutrality, and negativity levels during the pandemic.