Advertising avoidance is defined as “all actions that media users employ to reduce exposure to advertising content” (Speck and Elliott, 1997). Advertising avoidance is a subordinate study to advertising effectiveness research, and most studies draw on psychological research paradigms and methods. Advertising avoidance can be classified as behavioral, cognitive, or affective (Speck and Elliott, 1997; Cho and Cheon, 2004; Kelly et al., 2018). The delimitation of this study advertising avoidance is behavioral avoidance. This study defined advertising avoidance as the behavior adopted to avoid exposure to advertisements, by interrupting the communication of an advertisement in this study (Cho and Cheon, 2004), such as by scrolling down a web page to close the advertisement, and other actions.

Research into advertising avoidance extends into its complexities, including advertising avoidance under time pressure (Duan and Wu, 2021); perceived goal impediments (Cho and Cheon, 2004; Li and Ye, 2006; Shin and Lin, 2016), perceived advertising clutter (Cho and Cheon, 2004), previous negative experiences (Cho and Cheon, 2004; Shin and Lin, 2016), perceived usefulness (Zanot, 1984; Shin and Lin, 2016), and perceived relevance (Kelly et al., 2013). Research has gone through the three stages of traditional media, web 2.0 and web 3.0, and its influencing factors have also undergone significant transformation. With the realization of user orientation in the web 3.0 era, under the influence of the technological wave of artificial intelligence and machine learning and humanistic thinking, advertising avoidance research has effectively returned to a focus on people, and more and more researchers are studying users’ behaviors and attitudes toward advertising from the perspective of their perceived benefits and risks (Chen and Duan, 2021). Knittel et al. (2016) suggests that scholars can identify which advertising-specific factors influence advertising avoidance. This could help marketing managers plan their advertising campaigns and reduce the risks associated with expensive advertising. Therefore, this study investigated the mechanism of consumers’ risk communication on advertising avoidance from the perspective of their perceptions, in order to propose insights for weakening the risk perceptions brought by targeted advertising, in order to narrow the research gap and resolve consumers’ ambivalence of both wanting to enjoy convenience and fearing to bear losses, so as to reduce consumers’ advertising avoidance.

Bauer (1960) introduced the concept of “perceived risk” from psychology into consumer research. Mangold and Faulds (2009) asserted that perceived risk is defined as the uncertainty of services or goods. The present study defines perceived risk as an uncertainty regarding the possible negative consequences of using Online Targeted Advertising (Zhang, 2019).

Studies show that perceived risk is a multidimensional construct (Kaplan et al., 1974; Stone and Grønhaug, 1993; Forsythe and Bo, 2004; Chen, 2020). Cunningham (1967) classified consumer perceived risks as financial, psychological, performance, physical, time-loss, and social; a six-dimensional model of consumer risk perception. In the study of perceived risk, scholars should reconstruct their research models based on the six-dimensional model of perceived risk according to the research context (Campbell and Goodstein, 2001). In the online environment, studies have found physical (Jacoby and Kaplan, 1972), social (Barnes, 2006; Yu et al., 2007; Faqih and Jaradat, 2015), and psychological risk (Forsythe and Bo, 2004; Shao et al., 2006) to be redundant. However, the online environment can threaten consumer privacy, making privacy risk an important dimension of perceived risk (Miyazaki and Fernandez, 2001; Chen et al., 2019; Labrecque et al., 2021). Therefore this study omitted physical, psychological and social risks and added the perceived privacy risk dimension. Combined with the online targeted advertising research context, this study is concerned with advertising avoidance rather than purchase behavior and intention, so this study combines financial and performance risks into one dimension of performance risk, as suggested by Chen (2020). More importantly, studies have found that targeted advertising influences consumers’ decision quality and effort (Xiao and Benbasat, 2018), with limited capacities to choose and buy (Aguirre et al., 2015; Newell and Marabelli, 2015), the opportunity to choose (Chen et al., 2019), and the perception that choices are manipulated (Fan et al., 2020). In particular, when consumers use the search function, the platform prioritizes advertisements from merchants that have purchased priority, and consumers’ ability to freely access information is limited by the pushing of precise advertisements (Newell and Marabelli, 2015; Chen et al., 2019). Consumers feel limited in their search results. Such restrictions can be seen as a threat to freedom in the Internet age (Liu, 2019). According to resistance theory (Chao et al., 2020) “When freedom is threatened or restricted, people are generally motivated to restore it.” This study proposes that perceived freedom risk is a new dimension of perceived risk. Therefore, combining the literature on dimension of perceived risk, and considering the characteristics of online targeted advertising, this study selected the four dimensions of perceived performance risk, perceived privacy risk, perceived time-loss risk, and perceived freedom risk to investigate the influence of perceived risk on advertising avoidance. The definition of each dimension is shown in Table 1.

In recent years perceived risk has been widely studied in consumer behavior and attitudes. The literature has focused on online purchasing intentions (Pelaez et al., 2019; Ahmed et al., 2021; Jain, 2021), buying behavior (Wai et al., 2019; Tobi et al., 2020) impulse buying (Wu et al., 2020) travel behavior (Matiza and Kruger, 2021; Sharma et al., 2021), and customer loyalty (Esmaeili et al., 2021). Perceived risk has been less studied in the advertising industry. Perceived costs, privacy concerns, and the opportunity costs of online targeted advertising can cause consumers to resist online targeted advertising (Chen et al., 2019). Chao et al. (2020) studied the relationship between perceived privacy risk, psychological resistance and attitude to advertising, and found that perceived privacy risk had no significant effect on advertising attitude; Li et al. (2016) and Kurtz et al. (2021) studied location-based advertising (LBA) and found that perceived risk has an influence; in short video situational advertising, Zhang Meizhen confirmed the positive effect of perceived risk on psychological resistance and thus on consumers’ willingness to accept through the SOR model. The literature shows that most applications of perceived risk theory focus on the effects on online shopping intention or behavior (Pelaez et al., 2019; Wai et al., 2019; Tobi et al., 2020; Wu et al., 2020; Ahmed et al., 2021; Jain, 2021; Matiza and Kruger, 2021; Sharma et al., 2021). There is less research on advertising avoidance in online targeted advertising, and there are inconsistencies in conclusions (Chen et al., 2019; Chao et al., 2020). Thus there is need to identify the impact of perceived risk on advertising avoidance in online targeted advertising.

The psychological concept of emotion refers to an individual’s reaction to a specific object in response to external stimuli. Negative emotions include anger, complaints, depression, regret, and helplessness (Jaeger et al., 2019). In this paper negative emotions are defined as negative feelings such as anxiety and anger that consumers have in response to targeted advertising (Huang, 1997). Behavior is considered as an emotion-centered coping mechanism (Akram et al., 2021). When consumers find services on a website abruptly, serendipitous information from discovering information can be elicited to their emotional affection (Umair et al., 2018). When processing information about products consumers use their emotions as a measure (Grau and Folse, 2007). Emotions play an important role in the prediction of advertising effectiveness (Edell and Burke, 1987) and can influence its effectiveness (Petty et al., 1988; Zhang et al., 2015). Negative emotions have been widely used as an important mediating variable in the study of consumer behavior and attitudes. Negative emotions mediate between cognition and consumer behaviors and attitudes (Liu and Wang, 2017; Li, 2021).

The methodology is based on a post-positivist research paradigm and used a cross-sectional quantitative questionnaire-based survey design. A third-party organization was used to recruit respondents from users who have been exposed to online targeted advertising. The questionnaire was self-administered, and completed voluntarily and anonymously. The questionnaire helped respondents understand and objectively answer whether they have been exposed to, or avoided targeted advertising, by providing specific explanations of precision advertising and advertising avoidance.

The measurement items for all variables were derived from well-established scales shown to have high reliability and validity. There were 7 latent variables and 33 measurement items (Table 2). Responses used a 5-point Likert scale, anchored by strongly disagree or strongly agree.

The questionnaire was in three parts. The first part, the preface, introduced the researcher, stated the purpose of the survey, and defined the terms used. The second part elicited respondent demographic information. The third part contained the survey items, and used images and language to construct commonly encountered scenarios of targeted advertising.

This online survey was carried out by a market research company in China with 2.6 million registered respondents. A total of 436 usable datasets were obtained. Table 3 shows the demographic characteristics of respondents included in the analysis. Of the respondents 37.61% are Bachelor degree holders, and 52.75% are postgraduate degree holders, and overall, respondents had a high education level.

This study provides insights into consumer perceptions of the risks of online targeted advertising, whether these perceived risks influence advertising avoidance; and identified the mechanism of influences of perceived risks upon advertising avoidance. The keys findings are as follows:

We first discuss the important effects of perceived risk of online targeted advertising on advertising avoidance. Perceived performance, privacy, time-loss and freedom risks positively influence online targeted advertising avoidance, in decreasing order of perceived privacy risk (β = 0.561), perceived performance risk (β = 0.203), perceived freedom risk (β = 0.165), and perceived time-loss risk (β = 0.134). Overall, perceived privacy risk is the most significant predictor of advertising avoidance. Internet companies use advanced internet and computing technologies to collect user data for increasingly accurate personalized recommendations (Kim and Kim, 2017). There is a paradox for users, who may enjoy the convenience of accurate recommendations (Shareef et al., 2017), whilst also feeling threatened by collection of user information (Chen et al., 2019; Labrecque et al., 2021), restriction of freedom of choice, etc (Aguirre et al., 2015; Newell and Marabelli, 2015). Users develop advertising avoidance when they perceive the threat of online targeted advertising. This findings are line in technology threat avoidance theory (TTAT), approach-avoidance theory, psychology reactance theory, and are consistent with literature that shows user risk perception affects their behavior and motivation (Miyazaki and Fernandez, 2001; Aguirre et al., 2015; Newell and Marabelli, 2015; Xiao and Benbasat, 2018; Chen et al., 2019; Kurtz et al., 2021; Labrecque et al., 2021). The path coefficient of the effect of user perceived privacy risk on avoidance behaviors is much larger than the other three perceived risks, indicating that perceived privacy risk has a predominant effect on online targeted advertising avoidance. This finding is consistent with those of Wang et al. (2021b) that consumer privacy concerns are the biggest impediment of online targeted advertising today (Jiang and Wang, 2020).

Secondly, mediation analysis found that negative emotions partially mediate the relationship between perceived performance, time-loss and freedom risks and online targeted advertising avoidance. In other words, in the mediation model with negative sentiment as a mediating variable, perceived risks directly influence avoidance and also cause negative sentiments, resulting in avoidance. This finding is consistent with the literature, where perceptions of risk can generate a range of negative emotions that influence consumer behavior and attitudes (Mehrabian and Russell, 1974; Petty et al., 1988; Olney et al., 1991; Li, 2021). Of particular note is that contrary to expectations, negative emotions have no mediating effect between perceived privacy risk and online targeted advertising avoidance. A plausible explanation is that close to 90% of the study respondents study had a Bachelor degree or higher. According to Coping Behavior Theory (CBT), coping styles that consumers may adopt when they experience negative emotions triggered by privacy concerns are divided into two categories, but the specific situations in which consumers will choose which coping style depends on the resources available to them (Zhang and Li, 2018). Educated people may be more familiar with the laws and regulations on personal data protection, such as the Data Security Law of the People’s Republic of China and the Personal Information Protection Law, or understand that they can deny merchants access to personal information through web browser settings. If consumers have sufficient resources (or higher controllability) to deal with negative events directly, they will actively seek solutions, thus ameliorating negative outcomes. There are also ways to mitigate negative emotions (Gross, 1998), thus when perceiving the privacy risks of online advertising, more highly educated consumers seek solutions and mitigate negative emotions. Another possible explanation is associated with the large proportion of young people in the study respondents, in that young consumer tend to have an optimistic attitude toward privacy (Zhang et al., 2020).

Thirdly, moderation analysis found that perceived COVID-19 risk moderated the relationship between negative emotions and online targeted advertising avoidance. Consumers’ perceptions of the continued health threat of COVID-19 is likely to enhance their defense motivation, and reduce their travel while increasing their demand for online information. The motivation for danger control enhances the information-seeking process (Witte, 2009). This inference is strongly supported by China’s economic statistics, that show that annual retail sales of consumer goods declined 3.9% in 2020 compared to 2019, whereas online sales grew by 10.9% over 2019 (China Government Network, 2020), and that internet advertising revenue for 2020 grew by 13.85% over 2019 (China Economic, 2021). COVID-19 associated stressors are likely to have spurred consumers to overcome negative sentiments toward online targeted advertising and shifted them from avoidance to acceptance.

The contributions of this study to theory are five-fold. Firstly, this study combined brand avoidance theory, approach-avoidance theory and SOR theory to construct a new theoretical framework for advertising avoidance. This provides a theoretical reference for future research on online targeted advertising.

Secondly, the study extends brand avoidance theory (Knittel et al., 2016) that posits that when consumers have negative emotions about advertisements, they avoid them. However, the theory does not explain why consumers develop negative emotions. By using consumers’ negative perceptions as the antecedent variable for negative emotions to explain why consumers develop negative emotions that lead to advertising avoidance, this study complements brand avoidance theory.

Thirdly, extending the SOR model. This study extends the SOR theory by adding perceived COVID-19 risk as a moderating variable to explain whether consumer perception of COVID-19 risk moderates the effect of negative emotions on targeted advertising avoidance in the post-pandemic era.

Fourthly, the applicability of perceived risk theory is expanded. The literature review found that perceived risk theory is mainly used to study consumers’ online shopping behavior. This study combines perceived risk theory with brand avoidance theory for the study of consumer advertising avoidance, which expands the application scope of perceived risk theory.

Fifthly, extending the literature by proposing an additional dimension of perceived risk, which has financial, functional, physical, psychological, social, and time-loss dimensions. Because these dimensions can vary depending on the industry or product studied (Peter and Ryan, 1976; Liu et al., 2015; Yang et al., 2020), this study focused on consumer advertising avoidance; therefore, combining the research context and psychology reactance theory, this study argues that online targeted advertising interferes with consumers’ goals and has the potential to limit their action choices, which is important for making consumers perceive freedom risk (Edwards et al., 2002). Thus the theory of perceived risk may be extended by adding perceived freedom risk as an additional dimension.

This study analyzes the mechanism of influence on online targeted advertising behavior from the perspective of consumers’ perceived risk. It has some reference value for governments, platforms and advertisers.

For the government: The results of the study show that consumers’ perception of privacy risk is the biggest influencing factor on advertising avoidance, and perceived freedom risk also significantly affects advertising avoidance. Every user gets remunerations when the relationships are bound by respect, quality and fulfillment (Umair et al., 2018). Therefore, the government needs to strengthen legislation so that consumers’ personal information and private data are actually respected and protected in a reasonable manner. At the same time, it should further strengthen the protection of consumers’ freedom rights, and promote the appropriate handling and use of personal information resources within the framework of the legal principle of “legality, legitimacy and necessity,” while ensuring that citizens’ personality and privacy rights are not infringed upon.

For advertisers: First of all, advertisers need to improve the advertisement content in terms of mitigating the performance risk and time-loss risk perceived by consumers. The content of advertisements should be as comprehensive and precise as possible to provide information about the goods like online targeted advertising with precised individual’s propensity shopping vouchers as it drives the purchase (Akram et al., 2020), and avoid exaggerating the quality and functions of the goods or services, causing consumers to think that they are not in line with expectations. It should also be noted that the content length and video duration should not be too long and should be focused so that consumers in need can have an intuitive understanding of the promoted products and services in a short time and reduce the time cost of browsing targeted advertisements. Secondly, advertisers should choose the right platform, considering the suitability of the scenario with their own products and the reliability of the platform, to further reduce consumers’ perceptions of performance and time-loss risks. Finally, during the COVID-19 pandemic, the impact of sudden public emergencies on consumers and the market is continuous. Advertisers can help themselves resume normal operations through online targeted advertising.

For platform enterprise: Platforms enterprise need to improve their advertising platforms in terms of mitigating the perceived privacy and freedom risks to consumers. It is important to collect and use consumer-generated data under the scope of legal compliance, and to find a balance between the degree of accuracy and protection of personal privacy. Reinforce consumer information usage instructions and privacy settings. Add settings related to the “right to opt out” and “right to delete,” so that consumers can see their tagged tags, and they can also freely choose to opt out or delete these tags to improve consumer recognition of the platform’s security. At the same time, the platform cannot repeatedly push homogeneous advertising content in pursuit of economic benefits, thus alleviating the perceived risk of consumer freedom and time-loss.

In conclusion, this study is an important guide for enterprises or platforms to develop advertising strategies that mitigate consumers’ risk perceptions and negative emotions and improve the effectiveness of targeted advertising. It also provides new ideas for enterprises to find a balance between the benefits and risks of precision advertising.