TikTok marketing involves using short-form videos to promote products, brands, or services through influencer endorsements, user challenges, and storytelling (Cao & Nguyen 2025). This strategy utilises digital content creation and interactive exchanges to expand firms’ promotional efforts (Lacap et al. 2025). Unlike other social networking sites, which display static adverts, TikTok’s features enable contextual targeting through an algorithmically generated, engaging, and personalised content model (Truman et al. 2025). Typically, users can engage with marketing content via two main feeds: (1) ‘following’, which displays content from known contacts, and (2) ‘for you’, which shows content based on individual interests. Therefore, TikTok can facilitate viral marketing campaigns using trending hashtags, brand endorsements, and social media challenges (Truman et al. 2025). As Lacap et al. (2025) observe, businesses are increasingly using TikTok marketing to connect with Gen Z customers, boost brand awareness, and increase revenue, thereby shifting the social media marketing landscape.

The social media literature recognises the influence of CR (Hanaysha 2022), SI (Hazari et al. 2024; Mulcahy et al. 2025), and PE (Liu, Deng & Ayub 2025) on CAs and acceptance behaviours. However, few empirical studies examine how these variables shape Gen Z consumers’ attitudes towards TikTok marketing. Despite the significant benefits of TikTok marketing (Truman et al. 2025), existing research emphasises adoption behaviours across TikTok as a broader social media platform, neglecting its unique marketing characteristics. Furthermore, the role of DB as a barrier to TikTok marketing remains to be explored. As Cao and Nguyen (2025) suggest, further research should investigate moderating variables to enhance the understanding of consumers’ behaviours towards digital marketing content. Therefore, this study explores the social and psychological mechanisms that can predict Gen Z consumers’ attitudes towards TikTok marketing.

The theoretical foundation and research constructs for this study are based on related theories. The Social Influence Theory supports the development of SI as a construct and its associated hypotheses, while PE and CR are drawn from the Uses and Gratifications Theory (UGT). Combining these two theories provides an analytical framework for examining the mechanisms shaping Gen Z consumers’ attitudes towards TikTok marketing (see Figure 1). A detailed discussion of the relevance and applicability of the chosen theories to this study is provided below.

Building on Kelman’s (1958) early work, SI, as a theoretical concept, reflects the complex mechanisms that characterise interactions between individuals and their social environments (Angmo, Mahajan & Ranjitha 2025). In their social spheres, individuals engage with peers, family members, and other significant reference groups, which facilitates the formation of salient relationships (Hazari et al. 2024). These connections provide normative and informational cues that shape individuals’ opinions, beliefs, attitudes, perceptions, and behaviours (Sun, Chen & Xu 2025). Normative SI stems from one’s tendency to align with a significant other’s expectations (Jia et al. 2023), while informational SI occurs when one considers information obtained from others as factual (Charles & Kanani 2025). Therefore, the Social Identity Theory (SIT) posits that, through exposure to social groups, behaviour results from yielding to influence (compliance), aligning with and identifying with influencers (identification), and adjusting to conform to social norms (internalisation) (Angmo et al. 2025; Hazari et al. 2024; Kelman 1958).

Due to strong theoretical insights into how social connections influence behaviour, the SIT has been extensively applied in consumer behaviour research (Bhukya & Paul 2023). Literature on digital communication indicates that SIT is increasingly utilised to predict how consumers respond to social media influencers (Angmo et al. 2025), group-based online discussions (Sun et al. 2025), electronic word of mouth (Du et al. 2025), online brand communities (Willis & Ozuem 2025), social commerce (Charles & Kanani 2025), and TikTok brand ambassadors (Hazari et al. 2024). In this study’s context, SIT functions as a theoretical framework for understanding how Gen Z consumers’ social interactions with peers on TikTok influence their attitudes towards TikTok marketing. Mulcahy et al. (2025) contend that engagement with popular TikTok content is driven by social proof and credibility. Therefore, when Gen Z interacts with peers sharing common interests, they tend to adjust their behaviours, attitudes, and beliefs (Mulcahy et al. 2025). Consequently, SIT underpins the conceptual understanding of SI in this context.

The UGT offers a fundamental framework for understanding why individuals engage with a specific medium in active and selective ways (Menon 2024). Initially developed to analyse consumer choices related to mass media such as radio, television, and newspapers, the theory suggests that people are not passive but rather active, purposeful, and outcome-focused when selecting and using media, seeking some form of utility (Menon 2024). As a result, individuals choose which media to utilise and why based on their desire to fulfil particular unmet needs, which can lead to either positive or negative behavioural outcomes (Micallef, Hurcombe & Feng 2025). In the social media context, where users voluntarily participate and interact (Mizukoshi 2025), the UGT is employed to understand their motives for creating, sharing, and consuming content, including the needs they aim to satisfy (Menon 2024; Micallef et al. 2025; Mizukoshi 2025).

Several scholars have extended the UGT to various social media platforms, including Instagram (Menon 2024), WeChat (Kamboj & Sharma 2023), ChatGPT (Le et al. 2026), Line, and Facebook (Mizukoshi 2025). This shows that the UGT can be used to understand people’s motivations for using different digital communication platforms and the factors that influence these motivations (Le et al. 2026). More importantly, the theory can be adapted to various digital communication contexts, such as TikTok (Bhatiasevi 2024). Therefore, the current study explores how PE and CR affect Gen Z consumers’ attitudes towards TikTok marketing through the lens of the UGT.

The study adopts the hypothetico-deductive approach, grounded in the ontological perspective of objectivism and the epistemological perspective of positivism. This method is most suitable for scientific research that employs systematic, robust, and logical reasoning, aiming not only to generate knowledge but also to address key management issues (Sekaran & Bougie 2016). Accordingly, a quantitative cross-sectional survey design was utilised to collect data via a self-administered, structured online questionnaire.

The Gen Z group on TikTok in Zimbabwe served as the empirical context for this study. Often called the ‘post-millennials’ because they were born between 1995 and 2012 (Duffett & Maraule 2024), this generation exhibits notably high levels of digital literacy. A typical member is highly tech-savvy and relies on the internet and social media for product information and interactive engagement with both brands and peers (Boyle & Townsend 2024). As a result, TikTok has become a preferred source of commercial information, news, and social interaction for Gen Z consumers, surpassing traditional online search engines and short-form video formats (Lacap et al. 2025). Therefore, in accordance with ethical research guidelines, older Gen Z TikTok users (aged 18 years – 28 years) were chosen for fieldwork in this study, as they have reached the minimum age of consent in Zimbabwe to participate in a survey.

In line with previous related studies on the Gen Z cohort (see Duffett & Maraule 2024; Kara & Min 2024; Tseng, Chang & Zhu 2024), the study respondents were university students in Zimbabwe. The Zimbabwe National Statistics Agency (ZIMSTAT) reports that undergraduate enrolment at Zimbabwean universities is dominated by the 19–24 age group, which typifies the typical Gen Z cohort. Therefore, the target university in Masvingo had an overall undergraduate student population of 10 000. Using Yamane’s formula, the sample size was 370. The simple random sampling technique was employed to select the final study participants, yielding 380 responses, of which 341 were considered valid, resulting in a 90% response rate. To administer the questionnaire, a link was shared via WhatsApp in August 2025, inviting students to participate in a survey.

The scales for all latent constructs were developed from the existing literature and suitably adapted to the context of this study. The survey instrument included 25 questions addressing the main research constructs (SI, PE, CR, DB, and CA). A five-point Likert scale was used for all measurement items, with 1 = strongly disagree and 5 = strongly agree. Social influence was primarily drawn from Venkatesh (2000) and Yu et al. (2021). The measurement items for PE were adapted from Holdack, Lurie-Stoyanov and Fromme (2022) and Van der Walt, Nkoyi and Thithi (2024). Additionally, the CR scales were adapted from Jung (2017), and the DB scale was adapted from Durmuş et al. (2022). Finally, CA was assessed using items adapted from Mukherjee and Banerjee (2019). It is important to note that the scales were modified to fit the research context. Furthermore, the survey included questions collecting data on respondents’ demographic profiles. The results showed that 58% of respondents were female and 42% were male. Regarding exposure to TikTok marketing content, 52% reported frequent exposure, 30% infrequent exposure, and 18% moderate exposure.

Following the guidance provided by Collier (2020), the measurement model was evaluated by first verifying the indicator reliability (factor loadings, λ), and then conducting checks for internal consistency and construct reliability using Cronbach’s alpha (α), composite reliability, convergent validity (average variance extracted [AVE]), and discriminant validity (heterotrait–monotrait [HTMT] ratios and Fornell and Larcker). A partial least squares (PLS) algorithm was run in SmartPLS 4.0 (SmartPLS Gmbh, Rheinpromenade, Germany), and the results in Table 1 show that most factor loadings exceed the minimum threshold of 0.7 (Collier 2020), except for CR (CR1 = 0.684). However, based on Hair et al. (2021), indicator variables with low but acceptable factor loadings can be retained to support content validity if their contribution to the overall study is substantial. Hence, CR1 was considered. Owing to the coefficient alpha values, which ranged from 0.842 to 0.907, and the CR values, which ranged from 0.847 to 0.907, the observed latent constructs demonstrated internal consistency and construct reliability (Hair et al. 2021). Additionally, the AVEs exceed 0.5, ranging from 0.611 to 0.728, supporting the convergent validity of the constructs (Collier 2020).

In this study, a dual-criterion approach was employed to assess discriminant validity. According to Henseler, Ringle and Sarstedt (2015), the results in Table 2 confirm discriminant validity, as the HTMT ratios are below 0.9. This indicates that correlations among indicators measuring different constructs differ from those within indicators measuring the same construct (Collier 2020). Moreover, Table 3 shows that all requirements of the Fornell–Larcker criterion have been met. The scores shown diagonally in bold indicate that the square roots of the AVEs for a latent construct exceed the correlations among the constructs in the model (Fornell & Larcker 1981). This confirms and establishes discriminant validity.

Due to the use of self-reported items and a single cross-sectional survey, the data may be prone to measurement error, which could lead to common method bias (CMB). Therefore, Harman’s single factor analysis was conducted to detect the presence of CMB, and it was found that a single factor accounted for 37% of the total variance. Since this is below the acceptable threshold of 50%, CMB does not impact this study.

Ethical clearance to conduct this study was granted by the research committee of the Department of Management and Hospitality at Great Zimbabwe University (DMHRC/202507/001). The study adhered to all ethical guidelines set out by the university. The study participants consented to the study by responding to a filter question in the survey instrument.

The structural model was assessed through checks for collinearity, analysis of path coefficient significance, and evaluation of the model’s explanatory and predictive power. As shown in Table 4, all variance inflation factor (VIF) scores are below 5 (Hair et al. 2021), indicating that the model is not affected by collinearity. The analysis of the significance and relevance of the structural model’s relationships reveals that all path coefficients have substantial effects (Collier 2020), explaining 76.9% of the variance in Gen Z consumers’ attitudes towards TikTok marketing. The Q-square value of 0.757 indicates strong predictive relevance, indicating the model’s ability to predict the outcome variable.

The results in Table 4 demonstrate that SI is the most critical factor shaping Gen Z consumers’ attitude towards TikTok marketing (β = 0.365, t = 8.351, p = 0.00), thus supporting H1. Furthermore, the findings confirm H2 (β = 0.226, t = 3.434, p = 0.00) and H3 (β = 0.227, t = 3.717, p = 0.00), indicating that PE and CR are important predictors of Gen Z consumers’ attitude towards TikTok marketing. Regarding H4, the study examined the moderating effect of DB on the relationships between SI, PE, CR, and CA. Figure 2 shows that an adverse significant moderation effect was identified for SI (β = −0.188, t = 4.152, p = 0.00) and PE (β = −0.134, t = 2.095, p = 0.03), while CR (β = 0.233, t = 4.050, p = 0.00) showed a positive significant moderation effect. This suggests that high levels of DB reduce the influence of SI and PE on CAs towards TikTok marketing, while enhancing the impact of CR.

This study is significant in multiple ways. Social influence, PE, and CR have been identified as key predictors of CA. These findings support the SIT, which highlights the importance of SI in shaping attitudinal behaviours. Due to digital socialisation and peer interactions on social media, SI plays a crucial role in shaping Gen Z CAs. Furthermore, the study expands the SIT by analysing the moderating effect of DB, demonstrating how Gen Z consumers may become less receptive to marketing cues. It also improves understanding of UGT, indicating that PE and CR are vital factors affecting CAs towards TikTok marketing. Not only that, the study further refines UGT by introducing DB as a depletion constraint that alters the gratification-attitude relationship. Additionally, the results confirm that the predictors of CA (SI, PE, and CR) are based on validated theories. Therefore, the research extends existing scholarship on CAs and digital marketing technologies into the context of TikTok marketing. It further establishes TikTok as a source of timely, relevant, and valuable product information for young consumers making purchasing decisions. Finally, this study offers valuable insights into the behaviour of Gen Z consumers, a group that has traditionally been under-researched in marketing.

This study offers valuable practical insights for marketing practitioners in general and Zimbabwean marketing managers specifically. The findings indicate that marketing managers should focus on the factors influencing CAs, which were confirmed in this study. Social influence, PE, and CR were positively linked to CAs towards TikTok marketing. This suggests that marketers must ensure that TikTok content remains relevant to users’ needs. Furthermore, the content should be engaging and attractive to evoke positive consumer emotions. Since DB reduces consumers’ attitudes towards TikTok marketing, marketers should avoid overwhelming the platform with excessive information. Digital overload may result in negative attitudes towards TikTok marketing. Nonetheless, the study highlights TikTok’s inherent communicative potential as an emerging marketing tool with significant capacity to enhance brand visibility.

This study investigates a significant marketing communications channel and deepens our understanding of the factors influencing Gen Z consumers’ attitudes. Conducted in a Zimbabwean context, the results may have limited applicability to other regions due to differences in digital infrastructure, TikTok usage patterns, and socio-cultural factors. Since these contextual elements vary across markets, consumers’ attitudes towards TikTok marketing also differ. Therefore, behavioural responses observed in this study might not accurately mirror the perceptions of Gen Z consumers elsewhere. Given that the study utilised a small sample size, further research with a larger sample is needed to verify the conceptual model. Despite these constraints, this study provides valuable empirical evidence of the relationships among the variables under study. Future research could examine how psychosocial mechanisms shape attitudes towards mobile social media platforms among different generations. They could also examine how factors such as social media fatigue, technostress, and intrusiveness weaken the attitude-behaviour nexus across various social media platforms.