Cyberaddiction to TikTok during the COVID-19 pandemic Open Access

Because it was uncertain and exceptional in nature, the COVID-19 pandemic has generated a range of psychological ailments such as stress, anxiety, depression or panic (Nawaz, 2023). Emotional overload during the pandemic has led individuals to adapt themselves to the virtual world, particularly to social networks (Alfina et al., 2023). The latter allow for filling the gap of limited social interactions and lessening the sense of isolation that was created due to the pandemic (Cauberghe et al., 2021). Largely characterized by an exceptional surge in logging times during the pandemic (Zhang, 2022), these networks are now an auspicious environment for the development of addictive tendencies (Nikolinakou et al., 2024). In this regard, some studies noted that because of the omnipresence of social networks in people’s lives in times of crises, they were able to speed up a problematic use of these interfaces and the adoption of more addictive behaviors (Brailovskaia et al., 2021). Such behaviors are particularly noticeable in young populations, making them vulnerable to some addictive tendencies. Being potentially such, the social network TikTok, whose use has increased dramatically during the pandemic (Su et al., 2020), seems to be very popular for this age group, with more than one billion users (Montag et al., 2021) who can create music-filled videos or imitate the content of other videos. In this vein, Ceci (2023) pointed out that TikTok recorded more than 272 million downloads in the third quarter of 2023. However, the problems associated with this usage seem alarming, especially when the browsing habit turns into a need in itself, or even an anxiety-driven addiction (Brailovskaia et al., 2021). Accordingly, some studies have highlighted the idea that psychological disorders, often observed during a pandemic, lead to an addictive use of social networks (Shukur et al., 2023). Bearing on this proposal, it seems quite relevant to focus on the central role of cyberaddiction to TikTok in the relationship between psychological disorders (anxiety and loneliness) and attitudinal and intentional reactions, during a pandemic. It is also interesting to highlight the moderating role of fear of missing out (FOMO) in the relationship between cyberaddiction and attitude toward TikTok. Indeed, if the COVID-19 pandemic is potentially able of triggering coping mechanisms spreading over emerging social networks such as TikTok, addiction itself has not been sufficiently excavated, despite its managerial interest. In this regard, some studies clearly highlight the emotional disorders associated with this crisis and the crucial role played by these emotions in the development of affects and cognitions when using social networks (Nawaz, 2023). However, to our knowledge, very little research has focused on exploring in depth addictive behaviors to social networks and their central role in affecting users’ reactions. Even more, we know very little about the moderating role of FOMO. This latter concept has received little interest from researchers in marketing, although this emotion is associated with more attachment to social networks (Alfina et al., 2023). Accordingly, our research attempts to shed light on these behaviors while exploring the specific factors at the origin of its triggering as suggested by previous research (Zhang, 2022). This study will therefore attempt to answer the following research question: To what extent do emerging addictive psychological disorders during a pandemic affect consumers' addiction to the TikTok social network, and what impact does this addiction have on their attitude and intention to purchase the brands promoted on this network?

The aim of this study is to examine the determinants of cyberaddiction to the TikTok network during a pandemic as well as the resulting attitudinal and intentional reactions of users. To this end, two theories are used to consider TikTok users’ reactions. First, escape theory (Heatherton and Baumeister, 1991) assumes that certain self-regulatory processes, aimed to cope with the negative emotions felt by an individual, lead to extreme behaviors. These behaviors, aimed at coping with unhappy states of mind, can become addictive in some cases. Adapted to this study’s context, escape theory allows us to expect that the psychological disorders experienced during the pandemic led users to develop an addiction to TikTok. Thus, individuals seek to regulate the psychological disorders they suffer from through the adoption of excessive behavior, in this case, the excessive use of TikTok.

Second, attachment theory (Bowlby, 1958), initially developed to describe children’s biological attachment to their mother, proposes the idea of an instinctive and essential attachment to survive. By analogy, actions of TikTok users are designed in response to the attachment to this network. Such attachment is supposed to ensure their psychological comfort and well-being. Attitude toward TikTok can therefore be approached as a lever of favorable beliefs and strong emotional attachment to this network. Internet users would seek to regulate their psychological discomfort by adopting external figures as an escape.

The concept of internet addiction disorder took different names ranging from cyberaddiction, pathological, excessive or compulsive use, to a problematic use (Su et al., 2020). Internet-based addictive disorders seem to have originated with the appearance of social networks (Courbet et al., 2020). This has led authors to coin the phenomenon as social network addiction (Shukur et al., 2023). According to Bányai et al. (2017), this disorder describes an excessive concern, driven by an uncontrollable, sometimes pathological, urge to log in to social networks. The criteria behind this behavior are similar to those of addiction to substances or gambling, in particular mood disorders or conflicts (Nawaz, 2023). The COVID-19 pandemic presented itself as a setting conducive to the development of emotional disorders (Zhang, 2022). In this regard, it has been admitted that the negative emotions and psychological fluctuations experienced by individuals in these moments represent triggers for its addiction (Khrouf and Maghraoui, 2021). Specifically, the TikTok network seems to be the most addictive social media during the COVID-19 pandemic (Qin et al., 2022).

Individual vulnerabilities and psychological disorders appear to be key predictors of addictive behaviors (Cauberghe et al., 2021). However, the magnitude of the phenomenon during a pandemic incites us to explore its determinants in the context of emerging applications like TikTok.

COVID-19 anxiety is one of the most commonly experienced distresses during the pandemic (Brailovskaia et al., 2021). It represents a pathological manifestation of fear experienced in uncertain, unexpected or unknown situations (Zhang, 2022). This fear is often associated with an excessive tendency to use social networks (Courbet et al., 2020). Then, people experiencing anxiety are often phobic subjects who privilege the virtual world and seek comfort in interactions on social networks (Shukur et al., 2023). Accordingly, a direct association between anxiety and social network addiction was highlighted during the general lockdown imposed by many governments during the 2020 pandemic crisis (Brailovskaia et al., 2021). Confronted with the unique and unprecedented nature of the pandemic, users have shown an excessive attachment to the specific functionalities of TikTok which, in their eyes, allow them to relieve fear and fill an interpersonal void (Zhang, 2022). This amounts to a frantic search for an escape space from anxiety (Shi et al., 2017). Then, watching funny videos or participating in creating them on TikTok could likely alleviate this distress by creating a dopamine-filled sensation and more positive emotions (Strick, 2021). Against these proposals, the following hypothesis is then formulated:

Loneliness represents a subjective experience, created from people’s estimation of the gap between the desired level of social relationships and the one actually achieved (Peplau and Perlman, 1982). It seems that this emotion gained momentum during the pandemic because of the introduced health restrictions (Cauberghe et al., 2021). Because of the difficulty to socialize in real life during lockdowns, internet and social networks offered interesting social interaction alternatives to users while being physically distant (Cerdá-Mansilla et al., 2021a; 2021b). Indeed, the feeling of loneliness undermines the well-being of individuals, leading them to exchange with others in the virtual world in order to fill some of the experienced relational void (Shi et al., 2017). Cyberaddiction may therefore be an adaptive strategy arising from the difficulty of interacting with others, particularly during pandemic periods. In this line of thought, more recent research has suggested that loneliness was able to speed up addiction to the TikTok social network as an escape from everyday life stress and a shortcut to society (Zhang, 2022). These proposals allow us then to formulate the following hypothesis:

Attitude describes people’s beliefs and emotional attachment to a technology whose use becomes a daily routine (Ellisson et al., 2007). It seems to be an important outcome of cyberaddiction (Roth et al., 2022). Indeed, social theory considers that individuals develop beliefs that systematically gear their actions toward sources of reward (Lewinsohn and Graf, 1973). Using social networks, taking advantage of the interaction and sharing features they present, are probably major sources of reward, especially for addicted individuals (Shukur et al., 2023). In this regard, attachment theory assumes that when individuals are attached to an object with which they feel secure, they have positive attitude toward themselves and toward the object of attachment (Bowlby, 1958). Translated to the context of cyberaddiction to TikTok, users should develop positive attitude toward TikTok when they are addicted to it (Tang et al., 2021). They consider this social media as a refuge able to meet their security need, particularly in a context of pandemic, increasing thus individual vulnerabilities (Shukur et al., 2023). By analogy with the reinforcement mechanisms defined in Bandura's learning theory (Bandura, 1977), it seems that addiction as a response to an emotional deficit leads to an anticipation of affect supported by a favorable attitude to the network. These latter assumptions allow us to propose the following hypothesis:

Little research has explored the mediator role of cyberaddiction, especially in social networks. Nevertheless, Sanz-Blas et al. (2019) proved that addiction mediates the impact of Instagram overuse on emotional fatigue and instastress. According to them, cyberaddiction allows internet users to benefit from connecting with others, helping people or finding information. Cyberaddiction can thus be considered a tool appeasing psychological tensions for some internet users, especially in crisis periods. For TikTok-addicted users, the social network is this appeasing tool (Strick, 2021). Consequently, they naturally develop a positive attitude toward it. COVID-19 anxiety felt led internet users to use several strategies to reduce this state of mind. They tend to develop a cyberaddiction to social networks like TikTok which could fuel them with positive emotions and attitude (Strick, 2021; Zhang, 2022). The hedonic character of addiction makes users appreciate this source of pleasure and avoid focusing on its dark side. Cyberaddiction could therefore play a mediating role between COVID-19 anxiety and attitude toward the network. This observation allows us to formulate the following hypothesis:

For many individuals, the continuous exposure to videos on TikTok, because of cyberaddiction, is a source of positive effect and a means to recover from loneliness (Hunt et al., 2018). Thus, cyberaddiction is experienced as an escape from reality. In this regard, Shi et al. (2017) pointed out that family dysfunctioning is an important source of emotional loneliness. When internet users feel alone, they tend to look for virtual interaction and become vulnerable to cyberaddiction. Internet seems to provide feelings of autonomy and competence, increasing users’ self-esteem, especially for young users who are vulnerable. In a short time, they can become addicted to the medium and need to maintain these pleasant states, especially in pandemic periods that have the particularity of generating states of vulnerability (Hunt et al., 2018). In other words, cyberaddiction caused by loneliness could generate a feeling of pleasure and improve the internet user’s attitudinal states (Shi et al., 2017). This experience is so strong that it promotes acceptance and tolerance to the use of these networks and leads to a loss of interest in other activities (Tang et al., 2021). Bearing on these proposals, the following hypothesis is formulated:

With the rapid spread of the COVID-19 pandemic and the ever-accelerating use of social media, a truly new syndrome seems to be emerging; FOMO, also called fomophobia or FOMO. Przybylski et al. (2013) defined the concept as “a generalized anxiety that arises when individuals fail to get what they want to know in their absence, it is manifested primarily by the desire to continue to know what others are doing.” High levels of social media frequency can exacerbate the feeling of FOMO (Alfina et al., 2023). Specifically studying the TikTok social network, Alfina et al. (2023) concluded that FOMO is a factor increasing the addictive behavior of internet users. The Interaction of Person-Affect-Cognition-Execution model (I-PACE) presents FOMO as a cognitive bias intervening in situations perceived as stressful and generating problematic internet use (Brand et al., 2019). It can lead to compulsive behavior, and people may fail in resisting the urge to check their social media. Therefore, the fear of having missed something on the network is problematic for addicted persons as they start questioning their ability to control and self-regulation. Thus, the perceived lack of efficiency could decrease the favorable attitude toward TikTok as a space of addiction. TikTok is thus conceived by addicted people as a necessary malaise that they must continually control. This self-regulation could satisfy the need of personal freedom that governs attitude to end cognitive failure and psychological frustration. It is therefore expected that FOMO negatively moderates the effect of TikTok cyberaddiction on attitude toward the platform:

Brand attitude denotes a relatively enduring internal evaluation that assesses the degree to which consumers adhere to the brand (Spears and Singh, 2004). In a social network context, evaluation of content of a given network directly impacts the attitude toward the brand present on that network (Lin et al., 2019). Indeed, considering the importance of experience sharing and its effects on brand adherence, it seems that internet users tend to appreciate brands whose image is reinforced by social networks (Xu et al., 2021). It is thus quite expected that a positive attitude toward TikTok is likely to propel an attitudinal response of the same magnitude toward the brand (Lin et al., 2019). Moreover, Dwinanda et al. (2022) noticed that TikTok has a spray of features conducive to the development of greater interactivity between retailers and their followers, thereby improving attitude toward the brand. Bearing on these proposals, we propose the following hypothesis:

Purchase intention refers to “the subjective probability of making an effort to acquire a brand” (Spears and Singh, 2004). In this regard, the four-stage model of loyalty (Oliver, 1997) puts forward the following causal chain; from cognition to affect, from affect to intention and from intention to behavior. This allows for the assumption that purchase intention is explained by the internet user’s attitude and increases the likelihood of a real behavior toward a given brand. In this regard, Maghraoui et al. (2022) proved that consumers' attitude toward the brand present on social networks positively influences their purchase intention. In particular, on TikTok purchase intention seems to be strongly conditioned by the nature (favorable/unfavorable) of the internet user’s attitude toward the brand (Dwinanda et al., 2022). Hence, we can assume that liking a brand on TikTok is likely to increase the possibility of joining it through a purchase behavior in the future. These assumptions allow us to formulate the following hypothesis:

Bearing on the above proposals, we propose to test the following research model (Figure 1).

An online survey was conducted to better approach the behavior of TikTok users. The questionnaire was presented as an academic study seeking to understand the behavior of TikTok users, especially during the COVID-19 pandemic. The survey was administered to young TikTok Tunisian users, with an age range between 18 and 29. Indeed, this population belongs to Generation Z, which is born between 1995 and 2009 and represents currently the youngest and the largest population group (Wang et al., 2023). Indeed, Generation Z is digital native (Wang and Guo, 2023) and has grown simultaneously with the development of social media and is particularly familiar with online purchases (Belanche et al., 2019). It is also the population that has the highest social media usage rate (Xu et al., 2021). It has the particularity of being perpetually seeking to keep social ties, especially during the successive imposed lockdowns (Liu et al., 2021). In Tunisia, recent statistics show that most TikTok users practically belong to this age category (Novatis, 2023), which makes the target representative of the population of TikTok users. In addition, the literature suggests that cyberaddiction is more likely to affect young people, because of their vulnerability (Montag et al., 2021; Shukur et al., 2023). A link of the questionnaire was shared by the authors on social networks, particularly on TikTok as an immersive platform considered an addiction space (Montag et al., 2021). The survey took place between December 2021 and January 2022. After removing incomplete questionnaires, a convenience sample of 305 young TikTok users was obtained. Respondents have various demographic characteristics (Table 1).

The measures used in this study were validated in previous studies. Adaptations to the context of the TikTok platform were made. Then, two academic experts translated the measurement scales into French, which is the language spoken by respondents. A back-translation procedure was followed to ensure the conformity of items translated to the original ones (Frikha, 2019). The questionnaire includes items from Lee et al.'s (2020) COVID-19 anxiety scale, Russell's (1996) loneliness scale and Przybylski et al.'s (2013) FOMO scale. Similarly, Andreassen et al.'s (2014) scale to measure addiction toward TikTok, Ellisson et al.’s (2007) scale to measure attitude toward social networks, Chattopadhyay and Basu's (1990) brand attitude scale as well as Moon and Kim's (2001) scale to measure the purchase intention of the brand present on TikTok were chosen. To operationalize the studied constructs, five-points-Likert scales were used (Frikha, 2019).

To confirm the dimensionality of the used scales and to purify them, exploratory factor analyses were performed. The results showed the unidimensionality of the scales. The reliability analyses also proved that the scales had good internal consistency (Table 2). Partial least squares approach (PLS) was chosen for the data analysis (SmartPLS3 software) because it is particularly well suited for expansive and exploratory theoretical models (Xu et al., 2021). This study adapts escape theory and attachment theory (Bowlby, 1958) to the context of an emerging social media during the COVID-19 pandemic. Some concepts, like FOMO or COVID-19 anxiety, included in the research model have been little studied in the past and allowed us to adapt existing theories to the context. Furthermore, the PLS method is particularly adapted to complex models with a large number of constructs and items (Hair et al., 2014), which is the case in this study.

To avoid common variance bias that may appear when self-reported data is used, common method bias was assessed using Harman’s one-factor test (Podsakoff et al., 2003). To this end, an exploratory factor analysis was performed on all the variables of the study. The results indicate that the one extracted factor accounts for 41.33% of the variance among variables, which is less than the cutoff point of 50%. Then, common method bias is not a serious concern in this study.

Measurement model was estimated in order to check the reliability and validity of the used measurement scales. Reliability was established (Table 3) as all CR values are above 0.7 as recommended by Hair et al. (2014). Similarly, the obtained Dijkstra–Henseler's ρA are all above the threshold of 0.7 (Hair et al., 2014). Moreover, indicator reliability is satisfactory since outer loadings are above the cutoff point of 0.6 (Qin et al., 2022). Convergent validity was checked through AVEs, which are all above 0.5 (Hair et al., 2014).

Likewise, all constructs have good discriminant validity as they meet the Fornell and Larcker (1981) criterion that the square root of AVEs be higher than correlation between the latent variables (Table 4).

The HTMT matrix (Table 5) confirmed this finding and proves that discriminant validity is strong since all values are lower than the threshold of 0.85 (Qin et al., 2022).

Before proceeding to testing hypotheses, the predictive quality of the structural model was assessed using the R² and Q² coefficients (Table 6). The obtained R² coefficients are above the average of 0.33 (Xu et al., 2021). Similarly, the Stone-Geisser Q² coefficients are all higher than 0. These results therefore attest to the good predictive quality of the structural model.

To validate the hypotheses, a 5,000-iteration bootstrap was performed as recommended by Hair et al. (2014). The results (Table 7) proved first that COVID-19 anxiety promotes TikTok cyberaddiction. This finding allows us to support the first hypothesis (H1). Then, we found that loneliness felt by users also enhances TikTok cyberaddiction. Thus, the second hypothesis (H2) is supported. Moreover, it was proven that TikTok cyberaddiction favors the development of a positive attitude toward it. The third hypothesis (H3) is therefore supported.

To test hypotheses H4 and H5, the mediating effects have been estimated following the procedure of Hair et al. (2014). The results (Table 8) showed that COVID-19 anxiety has a significant effect on attitude toward TikTok. When the mediator is introduced, the effect of COVID-19 anxiety on attitude toward TikTok was still significant but decreased significantly (β changes from 0.94 to 0.29). The calculated variance accounted (VAF) indicated that mediation is partial (20% <VAF < 80%). Testing the mediation of cyberaddiction in the relationship between loneliness and attitude toward TikTok showed that loneliness has a significant effect on attitude toward TikTok. By introducing the mediator, the impact of loneliness on attitude toward TikTok became insignificant. The calculated VAF proved that this mediation is full (VAF > 80%). Thus, hypotheses H4 and H5 are supported.

Furthermore, the moderating effect of FOMO on the relationship between cyberaddiction and attitude toward TikTok was studied following the approach of Chin et al. (2003). In particular, we found that the interaction between the moderator (FOMO) and the independent variable (cyberaddiction) affected negatively the dependent variable (attitude toward TikTok). This finding means that FOMO slows down the impact of cyberaddiction on attitude toward TikTok. The sixth hypothesis (H6) is thus supported.

The results also proved that positive attitudes toward TikTok lead to positive attitudes toward the brand. The seventh hypothesis (H7) is therefore supported. Finally, findings confirm that attitude toward the brand promoted on TikTok has a positive influence on purchase intentions. Consequently, the eighth hypothesis (H8) is supported. The results of the structural model are summarized in Figure 2 below.

This study shed light on the influence of psychological disorders on TikTok cyberaddiction and its consequences on users' attitudinal and intentional reactions during the COVID-19 pandemic. Indeed, in line with previous research (Zhang, 2022), our results showed that COVID-19 anxiety as well as loneliness heightened addictive behavior to TikTok during the pandemic. Indeed, the intensive use of social media is to cope with depressive moods and promote anxiogenic thoughts of escape (Shukur et al., 2023). Moreover, restrictions and preventive measures during the pandemic accelerated the sense of isolation in many individuals. As far as TikTok is concerned, its use to escape anxiety and isolation seems to lead to problematic use and even addiction (Nikolinakou et al., 2024).

Furthermore, it was shown that TikTok cyberaddiction favors the development of a positive attitude toward it. This finding seems to be in perfect agreement with those of previous studies that indicated that users who are addicted to a given social network are the most likely to perceive it as a source of a useful and pleasant experience (Tang et al., 2021). Thus, the experience taking place when TikTok users develop addiction to the platform makes them appreciate it and generates positive attitudes toward it (Qin et al., 2022). The resulting effects of this connection activate adhesion and become the reason of connection. With regard to social networks, this cycle can be described as a dopamine loop. Accordingly, it has been proposed that addiction to TikTok leads to judge its use as pleasant and enjoyable (Roth et al., 2022). This trend may therefore be the source of a greater appreciation and more positive attitudes toward this network (Hunt et al., 2018). However, the findings indicated that FOMO slows down the impact of cyberaddiction on attitude toward TikTok. Remarkably, TikTok content was the most downloaded in 2020, the year of the pandemic (Iqbal, 2021), which could reveal a fomophobia of not being on the network. For Przybylski et al. (2013), this excessive fear questions the individual’s abilities, self-efficacy and self-esteem. To maintain their own coherence, the user seeks to develop a feeling of artificial self-efficacy. As for gourmand users, lack of situation control leads the fomofobic to dispel the appetitive effect of social networks and eliminate their aggression in order to maintain their cognitive coherence and self-esteem. Rather than a source of entertainment, TikTok is conceived by addicted people as a necessary evil that they must continually control.

Moreover, the results proved the mediating effects of cyberaddiction on the relationships between psychological disorders (anxiety and loneliness) and attitude toward TikTok. Thus, lonely or anxious users try to satisfy their social needs by interacting with other users on social networks, watching a video, playing games, etc. On TikTok, satisfaction of such needs increases cyberaddiction through tolerance and favorable attitude to the network as well as makes users lose interest in other activities that were previously important for them. The platform seems to take them to a virtual and pleasant world (Tang et al., 2021). This mediating relationship describes an info-communication phenomenon highlighting a concern with using TikTok as a dominant tool in their daily lives (Courbet et al., 2020) to escape individual vulnerabilities such as anxiety and loneliness (Cerdá-Mansilla et al., 2021a). This mediation gives a glimpse of how much this concern produces a specific emotional sphere that would be able to guide attitude of Internet users on social networks (Tang et al., 2021).

Furthermore, it seems that a positive attitude toward TikTok is more likely to develop a similar attitude toward the brand promoted on this network. This is in line with the results of Lin et al. (2019) who proposed that a transfer of effects occurs, generally, from the source of the communication channel to the brand. This transfer is an extrapolation of the attitude toward TikTok to the brand advertised on this network. Indeed, the very regular TikTok content, mostly generated by users or influencers in the form of short videos, helps brands to be known and to promote their accounts. In case the target appreciates the login activity on this network and judges it as appropriate to their needs, they are more likely to accept the brands present as well as their messages.

Finally, attitude toward the brand promoted on TikTok turns out to be a critical factor of future purchase intentions of the same brand, whether from the platform or outside. This is consistent with previous research indicating that consumers who initially developed a positive attitude toward the brand continue to desire it in the future (Lin et al., 2019). At this level, it seems that an anticipation mechanism plays a central role in action planning (Ajzen, 1991).

It is clear that today TikTok is an essential virtual space of entertainment for young people. Considered the leading platform for snack video content, TikTok is also the object of covetousness by brands that do not hesitate to exploit it in parallel with other networks such as Facebook, YouTube or Instagram. Nevertheless, TikTok represents an environment conducive to the development of cyberaddiction, especially since its users are often young, highly credulous and vulnerable (Qin et al, 2022). In particular, during the COVID-19 pandemic, internet users were more likely to develop psychological disorders that have promoted cyberaddiction (Shukur et al., 2023). In this regard, this study specifically showed that COVID-19-induced anxiety and loneliness increased cyberaddiction, which in turn affected positively attitude toward TikTok. This effect however decreases when TikTok users feel fomophobia. Moreover, TikTok cyberaddiction appears to play a mediating role in the relationship between psychological disorders (COVID-19 anxiety and loneliness) and attitude toward TikTok. It has also been proven that this attitude enhances, through a transfer of effects, attitude toward the brand promoted in the platform, stimulating purchase intentions.

From a theoretical perspective, this study has several implications (Table 9). First, it allowed us to consider TikTok an addiction medium affecting internet users’ attitude and behavioral intentions. Internet in general and social networks in particular are well known for their addictive power. Nevertheless, TikTok users seem to be even more prone to cyberaddiction, probably because of their young age, making them particularly vulnerable (Montag et al., 2021). Thus, this study has the merit of giving insights into this understudied phenomenon, despite its great importance (Qin et al., 2022). Indeed, although it is not yet recognized by the World Health Organization, TikTok cyberaddiction has several negative effects on mental health (Montag et al., 2021). Second, this research highlights two important determinants of cyberaddiction: COVID-19-induced anxiety and loneliness. Although characterizing the COVID-19 pandemic, anxiety and loneliness may appear in other crisis situations and even in the absence of a specific context. These psychological disorders seem to play a crucial role in developing an excessive attachment to TikTok that can convert into addiction. Third, this study proved the moderating role of FOMO in the relationship between TikTok cyberaddiction and attitude toward it. FOMO is considered a concept still very little known by marketing researchers (Wigman et al., 2017), yet it is particularly helpful to understand cyberaddiction (Liu et al., 2021). Fourth, the mediating role of TikTok cyberaddiction between its psychological determinants (COVID-19 anxiety and loneliness) and attitude toward the platform was revealed for the first time, to the best of our knowledge. It allows for a better understanding of how attitudinal and intentional reactions to TikTok are formed, especially in a crisis context. Fifth, two theories were applied to better approach TikTok cyberaddiction: escape theory (Heatherton and Baumeister, 1991) and attachment theory (Bowlby, 1958). Both were developed in psychology in order to understand individuals’ behaviors and were adapted in this study to the specificities of cyberaddiction.

From a managerial perspective, some recommendations could interest professionals. Then, our study’s results revealed that, during the COVID-19 pandemic, intention to purchase as well as positive attitudinal reactions stem from TikTok cyberaddiction generated by feelings of anxiety and loneliness. Nonetheless, managers should not in any case take advantage of critical periods like the COVID-19 pandemic to sell their products. Even if this study’s results are specific to the COVID-19 pandemic, they could be extended to other similar contexts, especially stressful ones. As a first step, it is recommended for firms to limit communication campaigns on TikTok during crisis periods. Thus, during health or economic crises, firms should take socially responsible actions and try to contribute to society’s well-being, especially with young vulnerable consumers like TikTok users. The large amount of information received makes them feel anxious and experience discomfort (Sanz-Blas et al., 2019). Avoiding setting-up major advertising campaigns during crises should limit the devastating effects of enhanced cyberaddiction on attitudes and the unnecessary purchases that could lead to regret or guilt (Nawaz, 2023). Second, in order to control ravaging TikTok cyberaddiction and its dark side, managers are recommended to undertake retargeting actions by sparing vulnerable TikTok users. Firms could exploit users’ data collected for advertising through machine learning techniques (Liu et al., 2021) in order to detect individuals addicted to TikTok and unstream harmful content specifically for them (Roth et al., 2022). Third, as socially responsible actors, firms should avoid putting in place means to fuel cyberaddiction or to stimulate psychological disorders like anxiety and loneliness. Advertising campaigns broadcast on TikTok should therefore be comforting and relaxing rather than stressful. Brands should therefore rely more on the fun content but also on its authenticity and reliability.

This study is not without limitations that open up some venues for future research. First, it would be interesting to see whether the obtained results are specific to the COVID-19 pandemic or not. To this end, researchers should replicate this study over a longer period of time and compare the obtained results with those of another health crisis or even in the absence of an emergency situation. Second, it is also interesting to check whether purchase intention is converted into an actual purchase, as proposed by the theory of planned behavior (Ajzen, 1991), by integrating this variable in the research model. Third, buying products in response to cyberaddiction could arise feelings of regret (Wang et al., 2023). It is therefore recommended for scholars to shed light on them by taking into account this concept as a consequence of an actual purchase. Fourth, the survey was limited to a young target audience for reasons already discussed. However, it would be interesting to take into consideration older TikTok users. Fifth, it would also be appropriate to study the effect of other determinants of cyberaddiction, such as problematic smartphone use (Nawaz, 2023) or narcissism (Wang et al., 2023). Adding other factors explaining cyberaddiction in the model would probably lead to a better comprehension of TikTok users’ behaviors.