Sage Journals: Discover world-class research

Abstract

The COVID-19 pandemic has significantly impacted global society and economies, prompting a shift in online consumer purchasing behavior (OCPB). This study aims to explore the evolving typologies of OCPB in the context of post-pandemic retail transformations. Utilizing data from a leading Chinese e-commerce platform, this study employed data mining techniques and the Bidirectional Encoder Representations from Transformers (BERT) model to analyze OCPB typologies. The BERT model, known for its advanced multi-task capabilities, demonstrated high reliability in classifying OCPB into three distinct types, as evidenced by robust performance metrics including accuracy, precision, recall, and F1-score. However, the BERT model alone did not fully capture the nuanced concept of OCPB typology. To address this, this study integrated grounded theory to further delineate OCPB typologies, identifying key dimensions such as social and economic adaptation, product attributes, and user experience. Our findings offer a comprehensive understanding of OCPB typologies and provide valuable insights for retailers navigating the post-COVID-19 landscape. This research not only contributes to the academic discourse on consumer behavior but also offers practical guidance for enhancing retail strategies in a rapidly changing environment.

Plain language summary

Online consumer purchasing behavior typology

The COVID-19 pandemic has changed the way people shop online. This study looks at how online shopping habits have evolved since the pandemic started, focusing on different types of shopping behaviors. We used information from a big online shopping site in China and special computer programs to study these behaviors. Our research found three main types of online shopping behaviors. However, to get a deeper understanding, we also looked at why people shop the way they do, considering factors like how they adapt socially and economically, what they think about the products, and their shopping experience. Our findings help us better understand online shopping habits and offer advice to stores on how to improve their online selling strategies in the new normal after the pandemic. This study is useful for both researchers interested in consumer behavior and for businesses looking to adapt to changes in how people shop.

Keywords

post-COVID-19 consumer behavior typology data mining data analysis

Introduction

The COVID-19 pandemic caused severe damage with global social and economic consequences (Donthu & Gustafsson, 2020). At the social level, the World Health Organization reported more than 6.5 million COVID-19-related deaths (World Health Organization, 2022). At the economic level, Statista reported that the global Gross Domestic Product was 3.59%, the global unemployment rate was 6.18%, and the global Purchasing Managers’ Index of the industrial sector was 53.6 for 2022 (Statista, 2022). These enormous social and economic facts cause a profound and lasting influence on global development and inspire many studies to combine study objects and the post- COVID-19 pandemic, especially consumer behavior.

The Chinese government implemented a lockdown and quarantine to limit the spread of COVID-19 from 23 January 2020 to 8 January 2022 (Luo & Tsang, 2020). However, an increasing number of consumers are actively participating in e-commerce activities to purchase food and living supplies in the emergency context of the COVID-19 pandemic. Consumers must protect their health by reducing or even stopping offline shopping while shifting to purchasing from social media or online applications when they are quarantining or maintaining social distancing. While the studies of post-COVID-19 within the social and economic reality are significant, an analysis of online consumer purchasing behavior (OCPB) is required.

Social media marketing, product knowledge, and crisis awareness have become the new pattern for Chinese consumers in the post-pandemic era (Sun et al., 2022). In addition, consumer behavior has changed from traditional to technological (Sheth, 2020). For instance, to become better informed, consumers pay more attention to favorable and critical online reviews when new products are launched (Chen & Farn, 2020; J. Hussain et al., 2022; Park, 2019a; Vana & Lambrecht, 2021). Most retailers and stores closed during the COVID-19 pandemic quarantine. Therefore, consumers had to purchase food and supplies from e-commerce platforms or social media instead of offline retailers, leading to a substantial change in their purchasing behavior.

Identifying the fundamental and most important OCPB typology is critical for finding solutions after the COVID-19 pandemic. This study presents a novel research perspective for analyzing the distinctions between current and conventional OCPB. For instance, purchasing channel disparities (online vs. offline) and shifts in perceptions toward product characteristics resulting from the post-COVID-19 pandemic, along with rising unemployment rates prompting consumers to reassess or decrease their consumption, as well as the impact of quarantine on the timeliness of product delivery, have necessitated a focus on streamlined operations rather than redundant functions due to challenges in online communication. By capitalizing on these notable gaps in research, this study aims to address the research question of categorizing, elucidating, and defining OCPB. They are revealed as consumer purchasing behavior patterns evolve in the dynamic social and political environment as the purchasing attitude changes in the real world.

The significance of this study is utilizing the Unified Theory of Acceptance and Use of Technology (UTAUT), social distance, capital and identity theories, and distance of information state transition theory to describe typology differences between normal consumer purchasing behavior and OCPB. In addition, it uses big data analytics, such as data mining and analysis, to classify the OCPB typology from online reviews on e-commerce platforms. The results will be more likely to reflect reality, be less biased, better describe the real situation, and avoid policy interference to a maximum extent compared to those of empirical studies, because data are directly acquired and selected according to the most representative characteristics of online consumers. Our research addresses the research gap by employing big data analytics and grounded theory to analyze the typology of OCPB after the COVID-19 pandemic. The key contributions of this study are to investigate how consumers balance their income and consumption concepts, what product attributes are essential to consumers, and how reactive and affective factors influence OCPB.

The remainder of this paper is organized as follows. Section of theoretical basis and literature review presents an extensive theoretical foundation and literature review. Section of method presents a detailed description of the research method. Section of grounded theory of OCPB typology provides the grounded theory to generalize the typology of OCPB. Section of implications discusses the theoretical and managerial implications. Section of conclusion and limitations offer conclusions, a description of limitations, and suggestions for future study.

Theoretical Basis and Literature Review

Theoretical Basis

UTAUT was developed in previous consumer behavior studies (Venkatesh et al., 2022). It proposes that consumer behavior is affected by six factors: performance expectancy, effort expectancy, social influence, facilitating conditions, behavioral intentions, and user behavior. UTAUT contributes to the analysis of various consumer interests. These interest areas are e-books (Lawson-Body et al., 2018), e-government (Al Mansoori et al., 2018), mobile banking (Zhou et al., 2010), e-health (Kalavani et al., 2018), online games (Ramírez-Correa et al., 2019), the factors of online purchase intentions (Sheikh et al., 2017), the value of e-commerce and acceptance by consumers (Abed, 2018).

Social distance theory refers to individuals’ degree of similarity in opinions about observed objects, scenarios, or behaviors, which is how individuals build their mental representations of an event (Trope et al., 2007). Social distance theory proposes that a shorter distance will manifest when relating to familiar people, while a longer one will display when interacting with strangers (Linke, 2012). Similarly, the relationship between consumers and brands or companies could be close or distant (Chernev et al., 2011; Escalas and Bettman, 2005). For instance, consumers are closer to product brands regularly used in daily life or frequently included in their social activities compared with those that appear less frequently in everyday activities (Fournier, 1998). Social distance theory explains why consumers are more influenced by closer brands (Choi & Winterich, 2013). Consumers have favorable attitudes toward closer brands compared to those that are unrelated to them or their social group (Mantovani et al., 2017).

Social capital theory refers to the analysis of individuals and connections in the overall network dimensions between them. As the frequency, depth, and breadth of their information exchanges increase, more social interactions occur among them (Larson, 1992; Ring & Van de Ven, 1994). Chiu et al. (2006) found that social capital factors, namely social interaction ties, trust, reciprocity, identification, and shared vision and language, are related to individuals’ knowledge sharing in virtual communities. Huang (2016) proposed that social capital theory was explained by bridging and bonding. L. Yuan, Deng, and Zhong (2021) found that social bonding and bridging positively influence passive users’ seeking and sharing opinions.

Big Data Analytics of OCPB

The definition of big data are the five V’s: volume, velocity, variety, veracity, and value (Erevelles et al., 2016; Lycett, 2013; Safi, 2022). Volume is the massive amount of data present in big data generated from a broad range of transactions. Velocity refers to the high speed of data generation. Variety is a key distinction between big and traditional data. Big data is unstructured and behavioral (e.g., comments and user-generated content), while traditional data is structured (e.g., scanner or sensor data, records, files, and databases). Veracity refers to data quality and accuracy. Compared to volume, velocity, and variety, veracity is a critical dimension to evaluate. Finally, establishing value involves identifying and analyzing useful data, while deleting that which is irrelevant.

According to a consumption data analysis based on 724,685 consumers from a Chinese international shopping center, Du and Lin (2022) conducted a social network analysis. They created a brand network map to test associations through visual and cohesion subgroup analysis. Establishing the importance of big data within retail organizations addressed two interrelated issues: identifying the awareness, availability, and development of big data environments and exploring opportunities and challenges associated with modern big data-based decision support systems within retail organizations (Aversa et al., 2021). The unified theory of acceptance and use of technology within the context of big data techniques was applied to explain the factors of resistance to use, perceived risk, and opportunity cost in service companies (Cabrera-Sánchez & Villarejo-Ramos, 2020).

Furthermore, big data technologies were also utilized to explain the effects of user experience elements on customer revisitations through sentiment analysis based on over 100,000 online review comments. The results of a structural equation model (SEM) indicated that customers’ satisfaction mediated the relationship between their emotional perceptions and customer revisitation. Meanwhile, Park (2019b) also combined big data approaches and SEM to analyze the relationship between customer satisfaction with online airline services and in-flight customer experiences. The primary difference between general and OCPB-specific big data is that the latter mainly comprises data drawn from online reviews and comments. Specifically, the increase in online review volume could lead to changes in sales.

Relationship Between the Typology of OCPB and Online Reviews

Online reviews are generated when consumers regularly and actively share feedback and comments about products or services on different shopping websites or platforms. Online reviews have been reported to have a direct relationship with the OCPB typology. Specifically, X. Xu et al. (2021) analyzed the empirical data on Airbnb in eight cities in the US. They found that consumer purchase behavior had a concave relationship with room information. The platform’s recommendations and providers’ verification information positively impacted consumers’ purchase behavior. Nilashi et al. (2021) utilized big data online reviews of vegetarian-friendly restaurants to analyze the relationship between customer satisfaction levels and service quality using Latent Dirichlet Allocation methods for text mining, Self-Organizing Map for data clustering, classification, and regression trees for preference prediction. The results indicated customers’ preferences could be predicted by online reviews for their segmentation. Meanwhile, S. Hussain et al. (2018) built an SEM to examine the relationship between online review credibility, perceived risk, argument quality, information usefulness, and information adoption purchasing behavior. Specifically, online review credibility positively impacted OCPB regarding argument quality and perceived risk.

Moreover, online reviews have transcended the limitations of traditional ones. Their broad diffusion scope strongly influences the OCPB typology, which comprises online consumers’ after-sale feedback, the main information source for understanding their needs and requirements. Thus, companies have gradually come to realize that online reviews play a significant role in consumers’ preferences for services or products. This realization allowed companies to conduct more accurate market analyses and design products that meet consumers’ needs.

In addition, consumers also rely on online reviews to reduce potential purchasing risk. S. Hussain et al. (2017) showed that online reviews positively affected perceived risk through source credibility. Nieto-García et al. (2017) found that the online review’s valence and volume, and internal reference prices, directly influenced consumers’ willingness to pay. In addition, when Lin and Xu (2017) utilized social distance theory to analyze the cross-cultural perception of trust, they found that three key factors (valence, reviewer ethnicity, and social distance) had a significant relationship with customers’ perceived reviewer trustworthiness. However, only valence had a significant relationship with brand attitude and purchase intention. Based on a systematic analysis, a social communication framework was used to understand prior online review studies. The results showed that the context of an online review (which comprised information quality, valence, volume, and veracity) was the main factor influencing OCPB (Cheung & Thadani, 2012).

Meanwhile, the effect of online reviews was mediated by the mechanism of social psychological distance. Based on credibility, Q. Xu (2014) concluded that online consumers’ affective perception and cognitive perception were highly connected with goods and services’ reputation and profile picture. When Jin and Phua (2014) explored social capital theory and source credibility to illustrate online reviews via Twitter, they showed that valence and the number of Twitter followers were closely associated with product involvement, buying intention, and online review spreading intention. Moreover, Blal and Sturman (2014) and Melián-González et al. (2013) explained that customers’ hotel choices, hotels’ income, and hotels’ strategies were greatly influenced by online reviews, while the valence and volume of online reviews had a different influence on hotels of different classes and the positive and negative representations of hotels’ property. Online reviews were generated when online consumers regularly and actively shared reviews and comments about products or services on different online shopping websites or online platforms. Given its accessibility to a wider audience through online platforms, online reviews’ influence on OCPB was far greater, compared with that of traditional reviews, which is usually limited in scope. The relationship between the online reviews and OCPB has been clearly shown in various studies.

Related Studies of OCPB Typology

Chinese consumers’ behavior patterns were identified by way of comparison, such as income level, expenditure structure and level, purchase method, study method; dining habits were not deeply influenced, while food price and quality were highly considered (Venkatesh et al., 2022; X. Yuan, Li, et al., 2021). Three OCPB patterns were introduced from the perspective of the distance of information state transition theory; three representative Chinese e-commerce enterprises—Tianmao Mall, Jingdong Mall, and Suning Easy-to-buy were chosen as study objects. Specifically, the three patterns were shown as online shopping for specific goods, online shopping for certain types of goods, and online leisure shopping. The results revealed that OCPB patterns mainly focused on the entire shopping process, quality of commodities, price, service, and satisfaction (Li et al., 2020). Moreover, the impact of a luxury brand’s social media marketing activities on customer engagement was analyzed by big data analytics. A total of 3.78 million tweets from the top 15 luxury brands were crawled, and customer engagement was enhanced by the typologies of entertainment, interaction, and trendiness dimensions of a luxury brand’s social media marketing efforts (X. Liu et al., 2021).

Moreover, Brandtner et al. (2021) utilized more than 533,000 consumer satisfaction ratings and more than 153,000 textual comments to provide the basis for a comprehensive and sound discussion of the impact of post-COVID-19 on consumer satisfaction and perceptions. The machine learning technique of big data analytics was utilized to predict consumer behavior on social media, which was clustered into four classes (Chaudhary et al., 2021). It was found that impulse fashion buying behavior online is always motivated by boredom, and can be easily attracted by the price, easy access, and free delivery (Sundström et al., 2019). Dangelico et al. (2021) also stated that consumers’ green purchase behavior had three typologies, which were green purchase satisfaction, green purchase frequency, and willingness to pay a premium price.

Furthermore, significant differences exist in the effect of consumers’ perception of their gendered behavior offline in contrast with online utilitarian shopping motivation and purchase intentions (Davis et al., 2017). Factors such as pre-sales offers, product variety, post-sales policies, and product significance influenced consumers’ decisions to make purchases both online and offline (Sunil, 2015). Meanwhile, the main differences in online and offline consumer behavior were values, lifestyles, and future intent. For instance, the effects of values on behavior were found to be more significant for OCPB, vacations and lifestyles had a stronger relationship, and the relationship between OCPB and future intent was much stronger (Díaz et al., 2017). C. W. Liu et al. (2017) stated that the consumers’ browsing behaviors on online shopping observation duration and browsing frequency of products with high brand awareness were respectively higher than those of products with low brand awareness when under time pressure.

The study revealed that environmentally conscious consumers believe purchasing eco-friendly smartphones fulfills their moral obligations, improves their perceived consumer efficacy, and aligns with their emotional values. Furthermore, the study showed that an increased awareness of environmental issues correlates with greater skepticism toward environmental claims, a perceived decline in product quality, and a perceived personal inconvenience (Raj et al., 2023). A significant relationship between the fear of missing out and problematic smartphone use, as well as consumer’s phubbing behavior, where smartphone addiction acts as a crucial intermediary, exerting a partial mediating influence (Paul et al., 2024). Based on a representative sample of Spanish consumers, the findings demonstrate that the diversity of shopping methods hinges on the variations in e-commerce usage, smartphone utilization, both offline and online interaction, and the interchangeability of online devices (Redondo & Charron, 2023). The study aims to identify masstige brands in Pakistan’s smartphone industry by employing the masstige mean index and to examine the impact of self-esteem on consumer buying behavior, with a particular emphasis on masstige, self-gifting behavior, and brand personality (Khan et al., 2024). The results indicate that consumers are willing to sacrifice a significant portion of the discount in exchange for a remanufactured smartphone that includes a warranty and appropriate labeling (Kerber et al., 2023). The study examines both the direct and mediated influences of consumers’ perceptions of their purchasing budgets on their purchase intentions. This examination is conducted through the lenses of perceived quality, perceived price, and perceived value in an international context. The objective is to understand the extent to which the perception of purchase budget affects the forecasting of consumer purchase intentions for smartphones available on worldwide online retail platforms (N’da et al., 2023). The study aims to investigate the relationship between traditional and sustainability-oriented customer requirements and the intention to purchase sustainable smartphones in China. Furthermore, it examines the mediating effect of perceived sustainable value and the moderating role of price sensitivity (Horani & Dong, 2023).

Method

This study employs a big data analytical approach to examine the typology of OCPB among Chinese consumers in the post-COVID-19 era. The methodology is designed to mitigate the limitations associated with traditional empirical data collection, which may be biased or incomplete, particularly in the context of online consumer feedback. Utilizing big data analysis could also be timely and precise based on the massive number of online reviews. Moreover, an online review can be effectively collected and analyzed using big data analysis software, such as Python packages. By leveraging real-world data from online reviews, this research aims to provide an objective and scientifically grounded analysis.

Data Mining and Processing

The primary data source for this study consists of approximately one million online reviews from major Chinese e-commerce platforms, Jingdong and Taobao, collected between February and May 2023. The data collection process utilized Python scripts to scrape the reviews, with a focus on smartphones from leading brands such as Apple, Samsung, and Huawei. Smartphones were chosen as the focal object due to their indispensable role in contemporary life, especially in the context of health and communication functionalities that have become critical during the COVID-19 pandemic.

To ensure the integrity of the dataset, reviews that were identified as fabricated or redundant were excluded. The preprocessing stage did not require additional noise filtering due to the application of the Bidirectional Encoder Representations from Transformers (BERT) model, which is adept at handling raw data.

BERT Model Application

BERT is designed to pre-train deep bidirectional representations from the unlabeled text by jointly conditioning the left and right context in all layers (Devlin et al., 2018). The BERT model, a cutting-edge multi-task natural language processing (NLP) framework, was employed to analyze the collected reviews. BERT’s architecture allows for the deep bidirectional training of representations by conditioning on both left and right context, making it particularly suited for understanding the nuances of consumer sentiment. The model was trained using two self-supervising tasks: the Masked Language Model (MLM) and the Next Sentence Prediction (NSP). This training facilitated the extraction and analysis of consumer sentiment and behavior from the unstructured text of online reviews. Moreover, NSP aims to judge whether sentence B is the actual next sentence that follows sentence A. If so, that case is labeled as “IsNext,” otherwise; “NotNext.” Two consecutive sentences randomly extracted from a parallel corpus generated the training tokens. Fifty percent of the extracted sentences were retained, that is, the ones that were consistent with the “IsNext” relationship, and the other 50% of the second sentence was randomly extracted, which was the “NotNext” relationship. After training the BERT, it can be applied to various natural language processing tasks. Therefore, the conditional probability can be shown as follows:

P = soft max (C W^{T})

(1)

where C is the classification symbol in BERT output, and $W$ is the learnable weight matrix.

Analysis and Performance Metrics

The analysis identified three distinct typologies of OCPB, represented visually in Figure 1. These typologies were derived from the BERT model’s ability to categorize reviews based on sentiment and content, with examples provided for each category.

Figure 1.

Typology results by BERT.

The model shows that typology of OCPB has three types represented by red, green, and blue. The data for the fourth typology (yellow) were not significant during the BERT process, so they were deleted based on this analysis. The examples of OCPB typology types are presented as follows. Type 1: “photography effect is so good,”“sound quality of the telephone is good,” and “battery lasts up to a day.”Type 2: “logistics are very fast,”“I made the order yesterday, and it was delivered today,” and “attitude of customer service was enthusiastic; she introduced the function online when I was quarantined.”Type 3: “operating system is so user-friendly,”“fingerprint unlocking does not always work,” and “functions can fully satisfy my needs.”

Second, the performance of the BERT model was evaluated using standard metrics: accuracy, precision, recall, and the F1-score. The accuracy is the ratio of the correct result pushed by the retrieval system to the total results. Precision and recall are pair indicators for retrieval system measurement. The f1-score is the harmonic mean of the precision and the recall rates:

Accuracy = \frac{n_{correct}}{n_{total}}

(2)

where $n_{correct}$ is the number of samples that are accurately classified in the test set, and $n_{total}$ represents the total number of samples in the test set.

In Table 1, TP means that predicted and actual samples are positive, while FP means that predicted samples are positive but actual ones are negative. Meanwhile, TN means that predicted and actual samples are negative, while FN means that predicted samples are negative but actual ones are positive. Table 1 shows precision and recall calculated for each category. Precision refers to all samples that the system determines to be positive, while the proportion of true positive samples. Recall refers to all samples that are positive, while the system determines the proportion of true positive samples.

Table 1.

Confusion Matrix.

	Actual positive	Actual negative
Predicted positive	TP	FP
Predicted negative	FN	TN

Precision = \frac{TP}{TP + FP}

(3)

Recall = \frac{TP}{TP + FN}

(4)

F 1 Score = \frac{2 \times precison \times recall}{precison + recall}

(5)

These metrics confirmed the model’s effectiveness in classifying consumer reviews into the identified OCPB typologies, with all performance indicators showing robust results (Figure 2). Specifically, Classification 1 and 3’s performance metrics are all above 0.9. Classification 2 performance metrics are all over 0.82, while the F1-score is above 0.9. Accuracy and recall surpass 0.9, while precision and F1-score surpass 0.88. In summary, this study’s methodology leverages advanced big data analytics and NLP techniques to uncover insights into consumer behavior in a post-pandemic context. By focusing on real-world online reviews and employing the BERT model, the research offers a timely and precise analysis of OCPB typologies among Chinese consumers.

Figure 2.

Performance metrics of BERT.

Grounded Theory of OCPB Typology

“Grounded theory, although clearly a qualitative method, endeavored to integrate the strengths inherent in quantitative methods with qualitative approaches” (Walker & Myrick, 2006). Shi et al. (2021) integrated big data analysis of online ordering platforms and grounded theory to explore factors influencing the offline customer experience. Mattoni (2020) examined local activists’ experiences of big data and grounded theory using digital media to hinder corrupt behaviors. OCPB phenomena, particularly in the context of the post-COVID-19 emergency among Chinese consumers, are new and evolving. However, because the results of BERT are annotated as independent sentences, they cannot directly summarize the typology of OCPB. Thus, this study explores and analyzes the OCPM typology by combining big data analysis and grounded theory.

Open coding, axial coding, and selective coding are the main processes to compare the main categories and the corresponding open coding categories. As the number of original data is massive, software Nvivo is utilized to manage the data and report results.

First, open coding is to encode and label the original data word by word to generate initial concepts and discover conceptual categories. The original data can maximally reduce the bias of the researcher. As the number of initial concepts is very high and there is a certain degree of crossover, while the category is the reclassification and recombination, the obtained initial concepts will be classified in the next step. The initial concepts that are less than 3 times of repetition will be deleted. Meanwhile, irrelevant and meaningless data will also be deleted. Table 2 shows the open coding of the OCPB typology.

Table 2.

Open Coding Results.

Category	Source material	Initial concept
Expressage	“Because of the COVID-19 pandemic, many companies refuse to deliver products, so I could only buy a smartphone from Jingdong, as Jingdong has its own delivery system.”	Refuse to deliver
Logistics	“My smartphone broke during COVID-19 pandemic; however, the express company is temporarily out of service and the stores are closed. Thanks to Jingdong’s logistics system, I received my smartphone the next day. What an amazing experience!”	Normal logistics
Financial issue	“Because of the 24-month zero-interest payment plan, I only had to pay around two hundred yuan per month. This is a really good deal.”	Loan
Promotion	“The retail offered a special discount on International Women’s Day, so I decided to buy a brand-new smartphone as a gift to myself.”	Special discount
Additional attribute	“This smartphone looks so good. I believe this is the best gift for my girlfriend.”	Gift
After-sales service	“… the customer service representative’s attitude is very enthusiastic. She described every detail of the smartphone. I really hope it stays that way …”	Customer service
Camera	“It has so many cameras! Regardless of whether the objective is near or far, I can always capture wonderful moments as clear photos! Its macro photography and super zoom are quite powerful. Its artificial intelligence can always automatically recognize the shooting scene, and the automatic object recognition and text translation feature also show the powerful capability of its cameras.”	Smart cameras
Screen	“The screen effect is not as good as I expected.”	Screen effect
Sound	“The sound of the single speaker is really weak, and it is a little difficult to hear when the environment is noisy.”	Sound effect
Battery	“It has a long standby time, lasting 24 hours on a single charge, I play for several hours and this only consumes less than 10% consumption.”	Standby time
Communication	“The voice is clear when I am making a telephone call.”	Clear voice
Chip	“My new smartphone was running faster than my old one, its processing speed is very fast, as is its download speed.”	Operating speed
Additional function	“The Bluetooth headset is really cool, and I can listen to music and answer the phone easily.”	Bluetooth
Storage	“Its internal storage is 256GB, which allows me to download many apps.”	High capacity
Price	“This smartphone is very convenient to use and cheap. I have to say that the smartphone is cheaper than similar products and the cost-performance ratio is nearly perfect.”	Cost-performance ratio
Quality	“This is my first time purchasing this smartphone, and I hope the reliable quality could use more than two years.”	Reliable quality
Positive emotion	“I was very happy when I received the phone.”	Happiness
Negative emotion	“The effect of the screen showed yellow when I first got it, I was so angry because I thought it was defective.”	Anger
Psychological reaction	“Its high pixel count and image resolution offer a clear display when I play a game or watch a video during the COVID-19 pandemic, which makes a super nice feeling.”	Nice feeling
Physiological reaction	“It feels good once you get used to it, but the touch of the screen takes a while to operate.”	Hard operation
Belief	“HUAWEI is the best among Chinese smartphones, I will support HUAWEI forever!”	China-made product
Preference	“I am an Apple fan, I bought the full range of Apple products.”	Apple fan
Usability	“My last smartphone had a serious lag, whereas the new smartphone gave me a surprise and spectacular smooth experience!”	Smooth experience

Second, axial coding is processed based on open coding results, which analyzes and extracts explicit and implicit logical relations between each category. The main categories can be summarized from corresponding sub-categories. The axial coding results produced three main categories social and economic adaptation, product attributes, and user experience (UX). Table 3 shows the axial coding results.

Table 3.

The Main Categories Formed by the Axial Coding.

	Main categories	Corresponding categories
1	Social and economic adaptation	Expressage
		Logistics
		Financial issue
		Promotion
		Additional attribute
		After-sales service
2	Product attributes	Camera
		Screen
		Sound
		Battery
		Communication
		Additional function
		Storage
		Price
		Quality
3	User experience	Positive emotion
		Negative emotion
		Psychological reaction
		Physiological reaction
		Belief
		Preference
		Usability

Third, selective coding mainly summarizes and extracts the main categories, which can obtain the core, overarching category. Then, a new theoretical framework can be built using the connection through the core, main, and other categories. The core category is regarded as an OCPB typology based on the research object and open and axial coding. Figure 3 shows the selective coding results.

Figure 3.

The results of selective coding.

Social and economic adaptation refers to coping with complex and difficult social and economic environments, such as pandemics, war, or natural disasters (Kirk & Rifkin, 2020; Zwanka & Buff, 2021) by making individual decisions. Specifically, expressage and logistics are essential conditions for consumers who purchase online during the lockdown or quarantine of COVID-19. Taobao occupied a large part of the e-commerce market before 2019, while Jingdong shared equal or even surpassed its market share after 2020 because Taobao used third-party expressage and logistics, while Jingdong had its own. Thus, when most courier services and logistics companies were temporarily suspended, consumers canceled Taobao orders and repurchased from Jingdong, which could deliver products consistently. Furthermore, during the lockdown and quarantine period, most small entrepreneurs and freelancers lacked sources of income. Based on sharply decreased income and consumption downgrading, they would not choose credit installment payments to face the financial crisis. More than ever, they were pressured to get the maximum discount or purchase cheaper products. Consumers also selected gifts for their spouses, relatives, or partners, showing additional attributes. Moreover, the online customer service’s role was to introduce the product’s function, and solve its repair or replacement issues or returns and complaints. If the consumer requirements were not addressed appropriately, they could leave unfavorable comments and adversely impact electronic word of mouth.

The product attribute concept is divided into tangible and intangible. Concrete, physical, and objective are the main features of tangible attributes, such as size, color, smell, product design, weight, and others. In contrast, abstract, beneficial, and subjective are intangible attributes’ main features, such as price, quality, and aesthetics. Camera, screen, sound, battery, communication, additional function, and storage are tangible product attributes; price and quality are intangible product attributes.

UX is subjective, situated, complex, and dynamic. It relates to technology more than instrumental needs to some extent. Hassenzahl and Tractinsky (2006) stated that “UX is a consequence of a user’s internal state (predispositions, expectations, needs, motivation, mood, and other features), the characteristics of the designed system (e.g., complexity, purpose, usability, and functionality.) and the context (or environment) where the interaction occurs (e.g., organizational/social setting, the meaningfulness of the activity, voluntariness of use).” Furthermore, UX is defined as: “person’s perceptions and responses resulting from the use and anticipated use of a product, system or service” based on ISO 9241-11 (ISO 9241-11: 2018, 2018). UX has a close relationship with usability, attributes, and subjective, objective, and aggregated measures (Law & Van Schaik, 2010). Factors that positively influence UX are user satisfaction, context-of-use, quality, enjoyment, and ease of use. Specifically, favorable and critical emotions reflect consumers’ internal state of mood when they receive a product. Both psychological and physiological reactions define consumers’ perceptions and responses after they a product. Consumer beliefs shows the UX’s subjective characteristics. For instance, some consumers believe that using domestically-produced China products shows patriotism. Moreover, usability is the designed system’s UX characteristic.

Finally, a theoretical saturation test was conducted to verify the model’s integrity. The results demonstrate that the model’s categories are sufficiently developed. No new categories remain to be formed.

Implications

Theoretical Implications

The study’s findings have several implications. First, consumers swiftly change their online behavior patterns in different contexts, namely post-COVID-19, mobile payments, e-commerce, e-healthcare, and e-learning systems (Mou & Benyoucef, 2021). Social distance theory explains why consumers change their purchasing attitude according to real-world conditions. When consumers adapt to the lockdown and social distancing due to the pandemic, they are more likely to adopt more convenient, technological behavior in work, leisure, and education. Consumers’ longer-term adaptive responses would show the potentially transformative changes in consumption. Consumers tend to combine post-COVID-19 and new adaptive behaviors in the current situation, such as using electronic banking, e-learning, e-commerce, and social media. Consumer behavior typology will be less reactive and more resilient according to UTAUT and these results, as consumers choose social and economic adaptation to confront uncertainty. Therefore, this study confirms social and economic adaptation as an OCPB typology and extends UTAUT. UTAUT and social distance theory provide empirical support for the theoretical framework of OCPB and serve as the foundational basis for classifying different types of OCPB.

Second, consumers are driven by brand identification, which relates to the brand identity’s attractiveness (Elbedweihy et al., 2016). Price-performance ratio and product attribute quality are of higher relevance to consumers than sustainability (Rausch et al., 2021). Most consumers pay more attention to durability and lower price instead of luxury because product attributes become a living guarantee. Consumers may reduce panic and impulsive purchasing behavior du product attributes. Product quality and brand are a living guarantee and a stabilizer for consumers after the COVID-19.

Third, reactive and affective factors are important stimuli in facilitating OCPB. UX includes factors of positive and negative emotions, psychological and physiological reactions, beliefs, preferences, and usability. UX has a significant relationship with consumer satisfaction. For instance, an OCPB typology could initially form through positive emotions and psychological and physiological reaction, belief, preference, and usability, while it may have a passive form. Therefore, as the foundation of social capital theory, this study provides a theoretical UX viewpoint according to the definition of social capital theory to enrich consumer behavior research.

Practical Implications

This study provides practical suggestions for both consumers and retailers. First, it explains that social and economic adaptation are important solutions to manage the uncertainty of post-COVID-19. The findings suggest that consumers should choose the normal operation of expressage or logistics when the area is in lockdown instead of panic purchasing or stockpiling, as some essential products can be stockpiled while others cannot. Retailers should consider cooperation with express companies. This strategy may reduce profit but is better than stopping delivery or building special expressage or logistics channels such as Jingdong. Because of quarantine, unemployment and bankruptcy rates are constantly increasing. Consumers can consider installment payments when the costs cannot be covered. Retailers can offer lower installment interest or increase the frequency of promotional campaigns. Some consumers choose products as gifts to express their feeling or celebrate a special occasion. Thus, retailers should consider additional attributes for customization. Moreover, after-sales service quality is a consumer concern and the key to success for retailers during such difficult periods. Customer service attitudes and skills should be well-trained and strictly monitored to reduce consumer complaints and increase satisfaction.

Second, this study finds that product attributes are the fundamental OCPB typology. These findings assume that consumers may consider price and quality more than other attributes due to consumption downgrade. Thus, retailers can reduce the flagship products ratio while increasing the low-price product ratios to expand market share. Moreover, counterfeit products should be eliminated, and low-quality products improved. Even if consumers choose lower price products during a difficult period, counterfeit products and low-quality products should not dominate the market. Furthermore, basic functions are needs that motivate the consumer to choose the product. Retailers should focus on the improvement of basic functions.

Third, UX elements can explain the consumers’ brand preference and brand loyalty and create a more prominent brand image for retailers. For instance, belief, positive psychological and physiological reactions and emotions, and usability may be strong motivations for consumers to choose the product. Retailers should consider improved product design to enhance consumers’ positive psychological and physiological reactions and prevent usability problems. However, they should also try to enhance brand belief through advertising to secure a consumer base and promote favorable consumer shopping experiences to improve brand image.

Conclusion and Limitations

In conclusion, this research utilized big data techniques to collect online reviews from the leading Chinese e-commerce platforms, aiming to analyze the typology of OCPB. Employing data mining methodologies, we leveraged the pre-trained BERT model, renowned for its deep bidirectional representations. This approach, incorporating Multi-Task Learning models such as MLM and NSP, facilitated a nuanced classification of OCPB typologies. Our findings, derived from the BERT model’s analysis, delineate OCPB into three distinct categories.

The robustness of our results was confirmed through rigorous evaluation using performance metrics including accuracy, precision, recall, and F1-score, all of which underscored the reliability of our findings. Additionally, the application of grounded theory enabled the systematic categorization of the data, leading to the identification of main categories and subcategories within the OCPB typology. This process involved initial concept analysis, followed by open and axial coding.

The resultant typology of OCPB was categorized into three main areas: social and economic adaptation, product attributes, and User Experience (UX). Social and economic adaptation encompasses factors such as delivery efficiency, logistics, financial considerations, promotions, product extras, and after-sales service. Product attributes were identified to include elements like camera quality, screen resolution, sound quality, battery life, connectivity, processor performance, additional functionalities, storage capacity, pricing, and overall product quality. Lastly, the UX category was found to encapsulate both positive and negative emotional responses, psychological and physiological reactions, consumer beliefs, and product usability.

This study not only advances our understanding of OCPB typologies but also offers valuable insights for e-commerce platforms and marketers aiming to enhance consumer satisfaction and engagement. By identifying and understanding the nuanced factors that influence online consumer behavior, businesses can tailor their strategies to better meet the needs and preferences of their customers.

This study contributed to consumer behavior theory and practices. However, it has some limitations. First, this study chose smartphones as a data mining object because of their varied application contexts, such as electronic banking, e-learning, e-commerce, and social media. However, food and other goods were also important after COVID-19. Thus, future research should focus on these product types. Second, this study exclusively investigated Chinese consumers’ situation. However, most countries have different post-COVID-19 policies. Future research should focus on consumer behavior in other cultural settings, such as a consumer comparison of Eastern and Western cultures during the lockdown. Third, this study combines big data analysis and grounded theory to investigate and explain the OCPB typology. However, corresponding categories may have been overlooked. For instance, this study finds that corresponding categories of UX are positive and negative emotions, psychological and physiological reactions, beliefs, preferences, and usability. Thus, future research should extend the scale of elements or OCPB factors. Finally, the data obtained from the e-commerce platforms is retained for a period of 3 months. Consequently, solely data from the most recent 3-month timeframe can be accessed, while any data predating the post-COVID-19 period is considered outdated and cannot be procured through web scraping methods.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work is supported by the following funding projects: Philosophy and Social Science Research Project of Jiangsu Higher Education Institutions (Grand No. 2017SJBFDY153); Jiangsu Provincial Social Science Fund Project (Grand No. 21YSB011); Special Project on College Student Ideological and Political Education of Jiangsu University (Grand No. JDXGA201103), Doctor Fund of Zhengzhou University of Light Industry (Grant No. 2020BSJJ022); Postgraduate Research & Practice Innovation Program of Jiangsu Province(Grant No. KYCX18_2211).

ORCID iD

Zongkang Yang

Data Availability Statement

The data used to support the findings of this study are available from the corresponding author upon request.

References

Abed

(2018). An empirical examination of Instagram as an s-commerce channel. Journal of Advances in Management Research, 15(2), 146–160.

Al Mansoori

K. A.

Sarabdeen

Tchantchane

A. L.

(2018). Investigating Emirati citizens’ adoption of e-government services in Abu Dhabi using modified UTAUT model. Information Technology & People, 31(2), 455–481.

Aversa

Hernandez

Doherty

(2021). Incorporating big data within retail organizations: a case study approach. Journal of Retailing and Consumer Services, 60, Article 102447.

Blal

Sturman

M. C.

(2014). The differential effects of the quality and quantity of online reviews on hotel room sales. Cornell Hospitality Quarterly, 55(4), 365–375.

Brandtner

Darbanian

Falatouri

Udokwu

(2021). Impact of COVID-19 on the customer end of retail supply chains: A big data analysis of consumer satisfaction. Sustainability, 13(3), Article 1464.

Cabrera-Sánchez

J. P.

Villarejo-Ramos

Á. F.

(2020). Acceptance and use of big data techniques in services companies. Journal of Retailing and Consumer Services, 52, Article 101888.

Chaudhary

Alam

Al-Rakhami

M. S.

Gumaei

(2021). Machine learning-based mathematical modelling for prediction of social media consumer behavior using big data analytics. Journal of Big Data, 8(1), 1–20.

Chen

M. J.

Farn

C. K.

(2020). Examining the influence of emotional expressions in online consumer reviews on perceived helpfulness. Information Processing & Management, 57(6), Article 102266.

Chernev

Hamilton

Gal

(2011). Competing for consumer identity: Limits to self-expression and the perils of lifestyle branding. Journal of Marketing, 75(3), 66–82.

10.

Cheung

C. M.

Thadani

D. R.

(2012). The impact of electronic word-of-mouth communication: A literature analysis and integrative model. Decision Support Systems, 54(1), 461–470.

11.

Chiu

C. M.

Hsu

M. H.

Wang

E. T.

(2006). Understanding knowledge sharing in virtual communities: An integration of social capital and social cognitive theories. Decision Support Systems, 42(3), 1872–1888.

12.

Choi

W. J.

Winterich

K. P.

(2013). Can brands move in from the outside? How moral identity enhances out-group brand attitudes. Journal of Marketing, 77(2), 96–111.

13.

Dangelico

R. M.

Nonino

Pompei

(2021). Which are the determinants of green purchase behaviour? A study of Italian consumers. Business Strategy and the Environment, 30(5), 2600–2620.

14.

Davis

Smith

S. D.

Lang

B. U.

(2017). A comparison of online and offline gender and goal directed shopping online. Journal of Retailing and Consumer Services, 38, 118–125.

15.

Devlin

Chang

M. W.

Lee

Toutanova

(2018). Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805.

16.

Díaz

Gómez

Molina

(2017). A comparison of online and offline consumer behaviour: An empirical study on a cinema shopping context. Journal of Retailing and Consumer Services, 38, 44–50.

17.

Donthu

Gustafsson

(2020). Effects of COVID-19 on business and research. Journal of Business Research, 117, 284–289.

18.

Lin

(2022). Brand connection and entry in the shopping mall ecological chain: Evidence from consumer behavior big data analysis based on two-sided markets. Journal of Cleaner Production, 364, Article 132663.

19.

Elbedweihy

A. M.

Jayawardhena

Elsharnouby

M. H.

Elsharnouby

T. H.

(2016). Customer relationship building: The role of brand attractiveness and consumer–brand identification Journal of Business Research, 69(8), 2901–2910.

20.

Erevelles

Fukawa

Swayne

(2016). Big Data consumer analytics and the transformation of marketing. Journal of Business Research, 69(2), 897–904.

21.

Escalas

J. E.

Bettman

J. R.

(2005). Self-construal, reference groups, and brand meaning. Journal of Consumer Research, 32(3), 378–389.

22.

Fournier

(1998). Consumers and their brands: Developing relationship theory in consumer research. Journal of Consumer Research, 24(4), 343–373.

23.

Hassenzahl

Tractinsky

(2006). User experience - A research agenda. Behaviour & Information Technology, 25(2), 91–97.

24.

Horani

L. F.

Dong

(2023). Understanding sustainable purchase intention of smartphone users Interface: Evidence from China. Frontiers in Psychology, 14, Article 1122801.

25.

Huang

L. T.

(2016). Flow and social capital theory in online impulse buying. Journal of Business Research, 69(6), 2277–2283.

26.

Hussain

Azhar

Ahmad

H. F.

Afzal

Raza

Lee

(2022). User experience quantification model from online user reviews. Applied Sciences, 12(13), Article 6700.

27.

Hussain

Ahmed

Jafar

R. M. S.

Rabnawaz

Jianzhou

(2017). EWOM source credibility, perceived risk and food product customer’s information adoption. Computers in Human Behavior, 66, 96–102.

28.

Hussain

Guangju

Jafar

R. M. S.

Ilyas

Mustafa

Jianzhou

(2018). Consumers’ online information adoption behavior: Motives and antecedents of electronic word of mouth communications. Computers in Human Behavior, 80, 22–32.

29.

ISO 9241-11: 2018. (2018). Ergonomics of human-system interaction—Part 11: Usability: Definitions and concepts. International Standard Organization, Geneva, Switzerland.

30.

Jin

S. A. A.

Phua

(2014). Following celebrities’ tweets about brands: The impact of twitter-based electronic word-of-mouth on consumers’ source credibility perception, buying intention, and social identification with celebrities. Journal of advertising, 43(2), 181–195.

31.

Kalavani

Kazerani

Shekofteh

(2018). Acceptance of evidence based medicine (EBM) databases by Iranian medical residents using unified theory of acceptance and use of technology (UTAUT). Health Policy and Technology, 7(3), 287–292.

32.

Kerber

J. C.

Fettermann

D. C.

Bouzon

(2023). Effects of green and circular products’ features on consumers’ choice of remanufactured smartphones. Business Strategy and the Environment, 3(33), 2331–2344.

33.

Khan

M. M.

Ishaq

M. I.

Iqbal

Raza

(2024). Impact of self-esteem and self-gifting on masstige purchase intentions. International Journal of Consumer Studies, 48(2), Article e13012.

34.

Kirk

C. P.

Rifkin

L. S.

(2020). I’ll trade you diamonds for toilet paper: Consumer reacting, coping and adapting behaviors in the COVID-19 pandemic. Journal of Business Research, 117, 124–131.

35.

Larson

(1992). Network dyads in entrepreneurial settings: A study of the governance of exchange relationships. Administrative Science Quarterly, 37(1), 76–104.

36.

Law

E. L. C.

Van Schaik

(2010). Modelling user experience–An agenda for research and practice. Interacting with Computers, 22(5), 313–322.

37.

Lawson-Body

Willoughby

Lawson-Body

Tamandja

E. M.

(2018). Students’ acceptance of E-books: An application of UTAUT. Journal of Computer Information Systems, 60(3), 256–267.

38.

Zhao

(2020). Measuring ease of use of mobile applications in e-commerce retailing from the perspective of consumer online shopping behaviour patterns. Journal of Retailing and Consumer Services, 55, Article 102093.

39.

Lin

C. A.

(2017). Effectiveness of online consumer reviews: The influence of source trustworthiness, valence, reviewer ethnicity and social distance. Internet Research, 27(2), 362–380.

40.

Linke

L. H.

(2012). Social closeness and decision making: Moral, attributive and emotional reactions to third party transgressions. Current Psychology, 31, 291–312.

41.

Liu

C. W.

Hsieh

A. Y.

S. K.

Hwang

(2017). What consumers see when time is running out: Consumers’ browsing behaviors on online shopping websites when under time pressure. Computers in Human Behavior, 70, 391–397.

42.

Liu

Shin

Burns

A. C.

(2021). Examining the impact of luxury brand’s social media marketing on customer engagement: Using big data analytics and natural language processing. Journal of Business Research, 125, 815–826.

43.

Luo

Tsang

K. P.

(2020). China and world output impact of the Hubei lockdown during the coronavirus outbreak. Contemporary Economic Policy, 38(4), 583–592.

44.

Lycett

(2013). ‘Datafication’: Making sense of (big) data in a complex world. European Journal of Information Systems, 22(4), 381–386.

45.

Mantovani

de Andrade

L. M.

Negrão

(2017). How motivations for CSR and consumer-brand social distance influence consumers to adopt pro-social behavior. Journal of Retailing and Consumer Services, 36, 156–163.

46.

Mattoni

(2020). The grounded theory method to study data-enabled activism against corruption: Between global communicative infrastructures and local activists’ experiences of big data. European Journal of Communication, 35(3), 265–277.

47.

Melián-González

Bulchand-Gidumal

González López-Valcárcel

(2013). Online customer reviews of hotels: As participation increases, better evaluation is obtained. Cornell Hospitality Quarterly, 54(3), 274–283.

48.

Mou

Benyoucef

(2021). Consumer behavior in social commerce: Results from a meta-analysis. Technological Forecasting and Social Change, 167, Article 120734.

49.

N’da

Ji-Fan Ren

Wang

(2023). Perception of the purchase budget (BGT) and purchase intention in smartphone selling industry: A cross-country analysis. PLOS ONE, 18(7), Article e0279575.

50.

Nieto-García

Muñoz-Gallego

P. A.

González-Benito

. (2017). Tourists’ willingness to pay for an accommodation: The effect of eWOM and internal reference price. International Journal of Hospitality Management, 62, 67–77.

51.

Nilashi

Ahmadi

Arji

Alsalem

K. O.

Samad

Ghabban

Alzahrani

A. O.

Ahani

Alarood

A. A.

(2021). Big social data and customer decision making in vegetarian restaurants: A combined machine learning method. Journal of Retailing and Consumer Services, 62, Article 102630.

52.

Park

(2019a). Motivations for customer revisit behavior in online review comments: Analyzing the role of user experience using big data approaches. Journal of Retailing and Consumer Services, 51, 14–18.

53.

Park

(2019b). The role of satisfaction on customer reuse to airline services: An application of Big Data approaches. Journal of Retailing and Consumer Services, 47, 370–374.

54.

Paul

Manchanda

Arora

Aggarwal

(2024). “I can’t look at you while talking!”–Fear of missing out and smartphone addiction as predictors of consumer’s phubbing behavior. Journal of Research in Interactive Marketing, 18(4), 666–687.

55.

Raj

Singh

Lascu

D. N.

(2023). Green smartphone purchase intentions: A conceptual framework and empirical investigation of Indian consumers. Journal of Cleaner Production, 403, Article 136658.

56.

Ramírez-Correa

Rondán-Cataluña

F. J.

Arenas-Gaitán

Martín-Velicia

(2019). Analysing the acceptation of online games in mobile devices: An application of UTAUT2. Journal of Retailing and Consumer Services, 50, 85–93.

57.

Rausch

T. M.

Baier

Wening

(2021). Does sustainability really matter to consumers? Assessing the importance of online shop and apparel product attributes. Journal of Retailing and Consumer Services, 63, Article 102681.

58.

Redondo

Charron

J. P.

(2023). Variety of shopping modes: Theoretical framework, pivotal factors, and managerial implications. Journal of Business Economics and Management, 24(5), 857–876.

59.

Ring

P. S.

Van de Ven

A. H.

(1994). Developmental processes of cooperative interorganizational relationships. Academy of Management Review, 19(1), 90–118.

60.

Safi

(2022). What consumers think about product self-assembly: Insights from big data. Journal of Business Research, 153, 341–354.

61.

Sheikh

Islam

Rana

Hameed

Saeed

(2017). Acceptance of social commerce framework in Saudi Arabia. Telematics and Informatics, 34(8), 1693–1708.

62.

Sheth

(2020). Impact of Covid-19 on consumer behavior: Will the old habits return or die? Journal of Business Research, 117, 280–283.

63.

Shi

Pei

(2021). Influencing factors of catering O2O customer experience: An approach integrating big data analytics with grounded theory. Tehnički vjesnik, 28(3), 862–872.

64.

Statista. (2022). Global economic indicators - Statistics & Facts. https://www.statista.com/topics/5442/global-economic-indicators/#topicHeader__wrapper

65.

Sun

Leng

Xiong

(2022). Research on the influencing factors of consumers’ green purchase behavior in the post-pandemic era. Journal of Retailing and Consumer Services, 69, Article 103118.

66.

Sundström

Hjelm-Lidholm

Radon

(2019). Clicking the boredom away–Exploring impulse fashion buying behavior online. Journal of Retailing and Consumer Services, 47, 150–156.

67.

Sunil. (2015). Trends and practices of consumers buying online and offline: An analysis of factors influencing consumer’s buying. International Journal of Commerce and Management, 25(4), 442–455.

68.

Trope

Liberman

Wakslak

(2007). Construal levels and psychological distance: Effects on representation, prediction, evaluation, and behavior. Journal of Consumer Psychology, 17(2), 83–95.

69.

Vana

Lambrecht

(2021). The effect of individual online reviews on purchase likelihood. Marketing Science, 40(4), 708–730.

70.

Venkatesh

Speier-Pero

Schuetz

(2022) Why do people shop online? A comprehensive framework of consumers’ online shopping intentions and behaviors. Information Technology & People, 35(5), 1590–1620.

71.

Walker

Myrick

(2006). Grounded theory: An exploration of process and procedure. Qualitative Health Research, 16(4), 547–559.

72.

World Health Organization. (2022). Who coronavirus disease (COVID-19) dashboard. https://covid19.who.int/

73.

(2014). Should I trust him? The effects of reviewer profile characteristics on eWOM credibility. Computers in Human Behavior, 33(2), 136–144.

74.

Zeng

(2021). The impact of information disclosure on consumer purchase behavior on sharing economy platform Airbnb. International Journal of Production Economics, 231, Article 107846.

75.

Yuan

Deng

Zhong

(2021). Encouraging passive members of online brand communities to generate eWOM based on TAM and social capital theory. IEEE Access, 9, 12840–12851.

76.

Yuan

Zhao

(2021). The changing patterns of consumers’ behavior in China: A comparison during and after the COVID-19 pandemic. International Journal of Environmental Research and Public Health, 18(5), Article 2447.

77.

Zhou

Wang

(2010). Integrating TTF and UTAUT to explain mobile banking user adoption. Computers in Human Behavior, 26(4), 760–767.

78.

Zwanka

R. J.

Buff

(2021). COVID-19 generation: A conceptual framework of the consumer behavioral shifts to be caused by the COVID-19 pandemic. Journal of International Consumer Marketing, 33(1), 58–67.

Analysis of Online Consumer Purchasing Behavior Typology After the COVID-19 Pandemic

Abstract

Plain language summary

Keywords

Introduction

Theoretical Basis and Literature Review

Theoretical Basis

Big Data Analytics of OCPB

Relationship Between the Typology of OCPB and Online Reviews

Related Studies of OCPB Typology

Method

Data Mining and Processing

BERT Model Application

Analysis and Performance Metrics

Grounded Theory of OCPB Typology

Implications

Theoretical Implications

Practical Implications

Conclusion and Limitations

Footnotes

Declaration of Conflicting Interests

Funding

ORCID iD

Data Availability Statement

References