Introduction

Social media has become ubiquitous in contemporary life, particularly in travel and tourism. Indeed, short videos on social media often influence consumers’ travel decisions. Individuals with extensive travel experience and an aptitude for sharing information may become opinion leaders online (Wu et al. 2017). Potential tourists, especially consumers with low self-esteem (Liu et al. 2018), tend to search for content shared by opinion leaders when making travel decisions. Meanwhile, consumers are becoming less sensitive to traditional marketing, consequently leading many marketers to turn to online influencers to promote their brands and products on social media (Leung et al. 2022a). As online influencers, internet celebrities in tourism play the role of opinion leaders in their fan community. Strong celebrity–fan relationships exist in “online influencer marketing” (Teng and Chen 2020), with consumers becoming active members of fan communities and participating in content marketing (Geng et al. 2020).

Destination marketing is becoming increasingly reliant on mobile short-form videos to stimulate viewers’ sense of immersion, social presence, and entertainment. Consequently, these videos are innovating the destination marketing model, which is becoming increasingly consumer-oriented (Wang 2020). In recent years, some travel destinations in China have become widely known in a relatively short amount of time by virtue of internet celebrities’ short travel videos. The popularity of “Zibo Barbecue” (a phenomenon in which people are flocking to Zibo for tourism and indulging in local barbecue) across China is a typical example of short video marketing. Internet celebrities’ short videos about a destination appear more effective and attractive than traditional advertising or promotional videos, which has inspired other DMOs to follow suit. However, some influencers’ short videos about a destination had a lukewarm response online, raising doubts about the effectiveness of short travel videos. Hence, it is crucial to understand how influencers’ short videos affect audiences.

As a new type of endorsement between no endorsement and celebrity endorsement, internet celebrities possess some characteristics that warrant the concerns of researchers and managers. Unlike traditional celebrities, internet celebrities have limited public credibility (Leung et al. 2022b); simultaneously, they are usually deeply engaged in a particular field and have mastered a certain degree of professional knowledge. Moreover, internet celebrities tend to automatically apply beauty filters to smooth out pores and remove blemishes in their videos to enhance the beauty effect, which is done to gain attention and increase influence (Xie et al. 2023). These characteristics—trustworthiness, expertise, and attractiveness—are key to explaining how internet celebrities affect their audiences. In addition, the production and dissemination of short travel videos on social media platforms can be monetized to generate revenue, contingent upon their ability to garner online popularity through metrics such as likes, followers, reposts, comments, etc. Internet celebrities carefully manage the content they create and match themselves to the destination in short videos to reduce incongruity (Lee and Eastin 2020). Sometimes, the destinations showcased in short videos often serve solely as video material or filming locations. It is not uncommon for destinations described in popular short videos to lack tourist presence. The above situations complicated the influencers’ short video marketing.

Prior research related to tourism has mainly focused on the effectiveness of celebrity endorsements (Zhang and Xu 2023; Halder et al. 2021; Choi and Rifon 2012) or official promotional videos (Cao et al. 2021; Gan et al. 2023; Guerrero-Rodríguez et al. 2020). As a third-party endorsement, the impact of an internet celebrity’s short travel video on travel decisions and destination marketing is not well understood due to its recent development and rapid growth during the we-media era. Thus far, the literature on endorser effectiveness has mainly focused on the endorser’s characteristics (e.g., source credibility) or a match between the endorser and the product (e.g., match-up hypothesis). Source credibility is an important informational and heuristic cue that is highly relevant in online environments (Ozanne et al. 2019), Simultaneously, appropriate matches between endorsers and products can result in a more favorable attitude toward the endorsements (van der Veen 2008). Therefore, in the present study, source credibility and match-up theory were combined to construct a conceptual framework to explore the underlying effect mechanisms linking internet celebrities’ characteristics, audiences’ attitudes toward short videos, audiences’ attitudes toward destinations, and visit intentions. In addition, which characteristics of internet celebrities, such as trustworthiness, attractiveness, expertise, and celebrity–destination congruence, can significantly affect the visit intentions of social media audiences were also examined. The results aim to contribute to the knowledge on internet celebrity endorsements in tourism, extend the application of source credibility and match-up theory in the field of online influencer marketing, and provide valuable implementation suggestions for DMOs to promote short video marketing.

Literature review and research hypothesis

Internet celebrities and short videos about travel destinations

Traditionally, celebrity endorsers usually refer to movie stars, singers, models, athletes, politicians, and businessmen (Hsu and McDonald 2002). In more recent times, with the increasing number of social media platforms (e.g., YouTube, Facebook, Twitter, and TikTok), the definition of celebrity has been expanding, propelling the emergence of “internet celebrities” known in prior studies as social media influencers (SMIs) or online influencers (OIs), who shape audience attitudes through blogs, tweets, and the use of other social media (Freberg et al. 2011). Unlike traditional celebrities who have succeeded in some credentialed fields (e.g., sports, acting, music), internet celebrities are not certified by any formal institutions (Leung et al. 2022b; McQuarrie et al. 2013). They gain followers by sharing content in their own styles and weaving brand endorsements into their personal stories and posts (Lou and Yuan 2019). Short travel videos filled with personal narratives (Dong et al. 2023) and destination information can serve as reference content, playing a significant role in shaping audiences’ cognition, evaluation, and aspiration to the destination. That is, a positive attitude towards a destination can impact audiences’ willingness to visit or recommend it (Yin et al. 2023).

The emergence of internet celebrities devoted to travel and to making short travel videos has attracted an increasing amount of attention, drawing young people to “tick off” an internet-famous destination. An online renowned destination refers to a place that can evoke symbolic interaction and emotional resonance among netizens, thus gaining wide network attention. Internet celebrities’ short travel videos dissolve the hierarchy of tourist destinations by highlighting and promoting visits to obscure or unknown locations (Sun 2022). Short travel videos can provide users with abundant information (Du et al. 2022) and positively affect electronic word-of-mouth on destination image (Yin et al. 2023). The existing literature has highlighted the effects of visual content, visual perspective (Gan et al. 2023), and audio-visual elements in promotional videos on consumers’ information processing and meaning construction. However, the influencing factor in social short video marketing is another important dimension that lacks discussion and exploration. It is thus crucial to understand the causal mechanisms connecting short travel videos and visits to travel destinations from the perspective of internet celebrity endorsements.

Source credibility

The concept of source credibility (the credibility of information and its source) was initially proposed by Hovland et al. (1953) and has been found to exert a significant influence on the persuasion process. Under the same information conditions, when the credibility of the source is high, an individual is easily persuaded (Chung et al. 2015; Filieri 2015). The application of source credibility theory in prior studies has demonstrated that celebrity endorsements have a positive effect on consumers’ brand awareness, brand trust (Hung 2014), brand preference (Albert et al. 2017), and purchase intention (Argyris et al. 2021; Mainolfi et al. 2022; Pradhan et al. 2016; Saldanha et al. 2018) when it comes to endorsed products, services, or destinations. Concerning the tourism industry, the intangibility of tourism-related products and the risks accompanying purchasing decisions make the issue of credibility even more important. In this case, source credibility can affect tourists’ perceptions of satisfaction with a destination (Veasna et al. 2013). On social media, credibility is essential for generating positive attitudes toward the endorser (Belanche et al. 2021), as users are more inclined to accept information from highly credible sources (Ho and Gebsombut 2019; Hussain et al. 2017; Sokolova and Kefi 2020).

Source credibility is composed of trustworthiness, expertise, and attractiveness (Tzoumaka et al. 2016). Trustworthiness can be viewed as consumers’ confidence in the information source (the endorser), that is, whether the information source is objective and truthful when provided (Erdogan 1999). In tourism, celebrity trustworthiness positively affects tourists’ attitudes toward travel advertising and their intention to visit a location (Johns et al. 2015; Kim et al. 2018). Expertise is derived from the endorser’s knowledge and experience in the relevant field (Erdogan 1999), which can be defined as the endorser being skilled, experienced, and possessing a certain degree of knowledge about the endorsed product. Attractiveness refers to how attractive, elegant, or cute the endorser appears to the audience (Wei and Lu, 2013). There is a popular saying, “Beauty is power,” which means that being physically attractive is a form of power, so whatever one says and does is seen as being right. Unconsciously, people are eager to identify with attractive people, so they are more willing to accept information from attractive endorsers (Kim et al. 2021) and allow the “beautiful is good” heuristic to guide their decisions (Ozanne et al. 2019). Theoretically, attractiveness includes physical attractiveness and moral characteristics, such as personality, lifestyle, and intellectual skills (Erdogan 1999). However, in the practice of celebrity endorsement, advertisers mainly focus on the physical attractiveness of the endorser (Tantiseneepong et al. 2012) due to its ease of observation.

In the field of tourism, user-generated content, such as photos, travel notes, and videos, can be used as information sources for tourists to make travel decisions. Studies have found that, as information sources, short videos are superior to photos and travel notes because of their storytelling and experiential qualities (Hautz et al. 2014). Sharing short videos and projecting attractive internet celebrity images can help new media audiences generate a positive flow experience, thereby affecting their travel attitudes (Zhang et al. 2022). In celebrity endorsements, attitudes toward products or brands mainly refer to positive or favorable dimensions. In tourism, attitudes toward a destination refer to audiences’ cognition, evaluation (Knoll and Matthes 2017), or aspiration (Xu and Pratt 2018) to the destination. The intangible nature of tourism products and the economic and psychological risks associated with visit decisions elevate the importance of credibility. When the information source is highly reliable, the social media audience is more likely to rely on the information when traveling. Accordingly, the following hypothesis is proposed:

H1: An internet celebrity’s trustworthiness has a significant positive influence on the audience.

H1a: An internet celebrity’s trustworthiness affects the audience’s attitude toward the short video.

H1b: An internet celebrity’s trustworthiness affects the audience’s attitude toward the destination.

H1c: An internet celebrity’s trustworthiness affects the audience’s intention to visit the destination.

H2: An internet celebrity’s expertise has a significant positive influence on the audience.

H2a: An internet celebrity’s expertise affects the audience’s attitude toward the short video.

H2b: An internet celebrity’s expertise affects the audience’s attitude toward the destination

H2c: An internet celebrity’s expertise affects the audience’s intention to visit the destination.

H3: An internet celebrity’s attractiveness has a significant positive influence on the audience.

H3a: An internet celebrity’s attractiveness affects the audience’s attitude toward the short video.

H3b: An internet celebrity’s attractiveness affects the audience’s attitude toward the destination.

H3c: An internet celebrity’s attractiveness affects the audience’s intention to visit the destination.

Match-up hypothesis

The influence of internet celebrities’ short travel videos on consumers’ visit intentions can be tested using source credibility. Meanwhile, the endorser–product match-up hypothesis provides another dimension (Kamins and Gupta 1994; Zhang and Xu 2023). Initially, the literature mainly focused on specific matches, namely, consistency between spokesperson’s characteristics and attributes of the product. Belongingness, similarity, and so forth, that can be used to describe the match-up hypothesis (Misra and Beatty 1990). Subsequent research focused more on overall matching; that is, the endorser being congruent with the product/brand and consumer perception of the congruence between the celebrity and product being endorsed at an overall level is more important (Till et al. 2008). Choi and Rifon (2012) proposed that endorser–product congruence positively affects consumer attitudes and behavior toward advertising and brands. Research on the match-up hypothesis has suggested that endorsements are more effective when the endorser’s image is congruent with the endorsed product (Till and Busler 2000; Till et al. 2008). A good match between the endorser and product leads to more positive consumer perceptions of advertising and product than a poor match (McCormick 2016; Yang et al. 2022), Moreover, native celebrity endorsers are more appropriate spokespeople for a destination compared with non-native (van der Veen and Song 2014).

Travel destinations are more complex and multidimensional than products and brands, and the destination image has been proven to be an important factor affecting potential tourists (Tasci and Gartner 2007). Content posted on social media by internet celebrities can transform the destination’s original image into an induced image. Therefore, congruence between the overall image of internet celebrities and destination image is crucial. In the hypothesis model proposed by Choi and Rifon (2012), celebrity–product congruence influences consumers’ advertising attitudes, brand attitudes, and purchase intentions. The prior tourism literature has mainly focused on the congruence between celebrities and tourists relying on the self-consistency theory (Chan et al. 2018; Li et al. 2022; Pradhan et al. 2016) without emphasizing the congruence between celebrities and destinations (Xu and Pratt 2018). A good match between internet celebrities’ overall image and the destination image may trigger a stronger willingness to visit. Therefore, the following hypothesis is proposed:

H4: Internet celebrity–destination congruence has a significant positive influence on the audience.

H4a: Internet celebrity–destination congruence affects the audience’s attitude toward the short video.

H4b: Internet celebrity–destination congruence affects the audience’s attitude toward the destination.

H4c: Internet celebrity–destination congruence affects the audience’s intention to visit the destination.

Dual mediations and destination advertising response (DAR) model

Attitude toward the advertisement, attitude toward the brand, and purchase intention have long been examined as advertising effectiveness variables, and their interrelationships have been theorized and corroborated in prior research (Heath and Gaeth 2019; Hwang et al. 2011; Kalwani and Silk 1982; MacKenzie and Lutz 1989). The dual mediation hypothesis proposes that consumers’ advertising attitudes will affect consumers’ brand attitudes and purchase intentions (MacKenzie and Lutz, 1989; MacKenzie et al. 1986). In the tourism literature, Park and Nicolau (2013) proposed the destination advertising response (DAR) model, in which the attitude toward the destinations replaced the attitude toward the brand. The major stages of the DAR model are attitude toward advertisements, attitude toward the destination, purchase trip facets, and total trip expenditure (Stienmetz et al. 2015). In addition, Xu and Pratt (2018) found that attitudes toward advertisements affect attitudes toward destinations and travel intentions in the context of social media influencers as endorsers. Compared with advertisements, short travel videos are more experiential and interactive, usually telling a travel story, which is more likely to arouse the audience’s positive attitude toward the destination and stimulate the willingness to travel. Thus, based on the dual mediation hypothesis and the DAR model, the following hypothesis is proposed:

H5a: Audiences’ attitude toward the short video affects their attitudes toward the destination.

H5b: Audiences’ attitude toward the destination affects their visit intention.

The conceptual model constructed in the present study is shown in Fig. 1.

Fig. 1
figure 1

Proposed model and hypotheses.

Full size image

Materials and methods

Case selection

TikTok (Douyin) is China’s most popular short video social platform, the number of its users reached 842 million, and the majority of users are under 35 years old. The present study used FangQi Kiki, a short travel video producer on TikTok, as an example. Fang Qi Kiki used to be the location host of “Beautiful China Tour,” a channel on Discovery Journey. She currently has 20 million followers and 456 posts on TikTok and is a typical producer of short travel videos. This study employed her representative short video “Big Fish and Begonia” as a case study. The video has 1.97 million likes and 89,000 comments online and was filmed in the Tulou buildings located in Yongding and Nanjing in Fujian Province, China. The short video tells a traveling story against the background of the film Big Fish, which was shot in Fujian Tulou.

Questionnaire design

The questionnaire consisted of four parts: The first was demographic characteristics, including gender, age, education, frequently used social platforms, frequency of watching short videos per day, and familiarity with FangQi Kiki. The second part measured the audience’s attitude toward the short video, attitude toward the destination, and visit intention. The third part measured the influence of the internet celebrity’s trustworthiness, expertise, and attractiveness. The fourth part measured the influence of internet celebrity–destination congruence. In the second, third, and fourth parts, a 7-point Likert scale was adopted, with 1–7 representing “strongly disagree,” “disagree,” “somewhat disagree,” “uncertain,” “somewhat agree,” “agree,” and “strongly agree,” respectively (Tables 1 and 2).

Table 1 Profiles of the respondents (N = 649).
Full size table
Table 2 Reliability and validity of the measurement model.
Full size table

Sampling survey and data collection

At the beginning of the study, the questionnaire was predistributed using the snowball technique. After collecting some samples, the structure and content of the questionnaire were optimized. The questionnaire was officially distributed by a combination of random sampling and stratified sampling through the Chinese online survey platform SoJump. A short video of FangQi Kiki was embedded in the questionnaire. The participants were asked to fill out a questionnaire after watching the short video (1 min and 06 s). The official distribution period for the questionnaire was from March 26 to April 8 and October 28 to November 5, 2023, 783 questionnaires were received, 134 invalid questionnaires (including those with missing values, the same score for all items, and samples who living in Fujian province) were excluded. Ultimately, 649 valid questionnaires were obtained, with an effective rate of 82.89%. According to IP addresses, the distance between the sample location and Fujian Tulou ranged from 360 km (Ganzhou, Jiangxi Province) to 4600 km (Altay, Xinjiang Autonomous Regions). The samples came from 116 cities in 31 provinces (including municipalities and autonomous regions) in China.

Among the 649 valid samples, men accounted for 48.4%, women for 51.6%, young people aged 18–26 for 53.6%, 27–35 for 25.9%, and 36–45 for 11.1% of the sample. The post-95s (those born after 1995, known as Generation Z) are the main group conforming to the consumption characteristics of internet celebrities’ short videos. In the educational structure, those with an education below the college level accounted for 18.5%, junior college accounted for 21.4%, undergraduates accounted for 52.2%, and postgraduate and above for 7.9% of the sample. The most frequently used social platforms were WeChat and TikTok, accounting for 93.8% and 77.3% of the sample, respectively. Meanwhile, 33.3% occasionally watched short videos every day, and 52.5% frequently watched short videos. Of the total sample, 46.2% were unfamiliar with FangQi Kiki, 37.0% knew something about her, and 12.6% often watched her short videos. Samples were diversified and representative of the major audiences of social short videos (Table 1).

Reliability and validity

First, a reliability test was conducted on latent variables: the internet celebrity’s trustworthiness, expertise, attractiveness, internet celebrity–destination congruence, attitude toward the short video, and attitude toward the destination. The results showed that Cronbach’s alpha was between 0.865 and 0.958. The average variance extracted (AVE) ranged from 0.719 to 0.857 ( > 0.5), and the combined reliability (CR) ranged from 0.911 to 0.968 (> 0.7), indicating that the reliability of the questionnaire was good. Second, the KMO value measurement and the Bartlett sphericity test were carried out. The results showed that KMO ranged from 0.748 to 0.904, and Bartlett sphericity test P-values were all less than 0.001, indicating that all variables in the scale were significantly correlated (Table 2).

Finally, the discriminant validity was analyzed (Table 3). The square root of the AVE of each variable (the number in bold diagonal) was greater than the correlation coefficient between any two variables, indicating that the discriminant validity of each latent variable was good. In the present study, the square root of the AVE of each latent variable was higher than the correlation coefficient of each latent variable, indicating that the discriminant validity met the analytical requirements.

Table 3 Convergent and discriminant validity of the measurement model.
Full size table

Results

Structural model results

A path analysis of the structural equation model (SEM) was utilized to test the hypothesis model with the help of Amos 26.0. In this case, \({\chi}^{2}\)/df = 2.348, CN = 319 > 200, RMSEA = 0.046 < 0.05, RMR = 0.044 < 0.060, GFI = 0.933 > 0.9, AGFI = 0.909 > 0.9, CFI = 0.980 > 0.9, NFI = 0.966 > 0.9, and TLI = 0.975 > 0.9 all met the adaptation standard, indicating that the model hypothesis was appropriate.

The first group of hypotheses examined the influence of internet celebrities’ trustworthiness regarding short videos (H1a), destinations (H1b), and visit intentions (H1c). The results showed that trustworthiness significantly affected the audience’s attitude toward the short video (β = 0.233, p = 0.002 < 0.05), the audience’s attitude toward the destination (β = 0.139, p = 0.021 < 0.05), and the audience’s visit intention (β = 0.167, p = 0.023 < 0.05). Therefore, H1a, H1b, and H1c were all supported; that is, internet celebrities’ trustworthiness significantly influenced the audience (Table 4).

Table 4 Structural model.
Full size table

The second group of hypotheses examined the influence of internet celebrities’ expertise on short video (H2a), destination (H2b), and visit intentions (H2c). The results showed that, although the internet celebrities’ expertise was positively correlated with the audience’s attitude toward the short video, there was no significant effect (β = 0.144, p = 0.158 > 0.05). The internet celebrities’ expertise was negatively correlated with the audience’s attitude toward the destination, and there was no significant effect (β = −0.012, p = 0.884 > 0.05). The internet celebrities’ expertise significantly affected the audience’s visit intention (β = 0.316, p = 0.001 < 0.05). Therefore, H2c was supported, and neither H2a nor H2b was supported (Table 4).

The third group of hypotheses examined the influence of internet celebrities’ attractiveness on short videos (H3a), destinations (H3b), and visit intentions (H3c). The results showed that the internet celebrities’ attractiveness had no significant effect on the audience’s attitude toward the short video (β = 0.071, p = 0.415 > 0.05), the audience’s attitude toward the destination (β = 0.047, p = 0.478 > 0.05) and audience’s visit intention (β = −0.076, p = 0.349 > 0.05). Therefore, hypotheses H3a, H3b, and H3c were not supported (Table 4).

The fourth group of hypotheses examined the influence of celebrity–destination congruence on short videos (H4a), destinations (H4b), and visit intention (H4c). The results showed that, although the celebrity–destination congruence significantly affected the audience’s attitude toward the short video (β = 0.141, p = 0.031 < 0.05), it had no significant effect on destination attitude (β = −0.010, p = 0.843 > 0.05) and visit intention (β = 0.009, p = 0.883 > 0.05). Therefore, H4a was supported, and neither H4b nor H4c was supported (Table 4).

The fifth group of hypotheses examined the influence of the audience’s attitude toward the short video on their destination attitude (H5a) and visit intention (H5b). The results showed that the audience’s attitude toward the short video significantly affected their attitude toward the destination (β = 0.683, p = 0.000 < 0.05) likewise, the audience’s attitude toward the destination significantly affected their visit intention (β = 0.216, p = 0.000 < 0.05). Thus, both H5a and H5b were supported (Table 4).

Seven of the 14 proposed hypotheses, namely H1a, H1b, H1c, H2c, H4a, H5a, and H5b, were statistically significant (Fig. 2).

Fig. 2: Estimated model.
figure 2

***p ＜ 0.001, **p ＜ 0.01, *p ＜ 0.05; NS Not Supported.

Full size image

Based on the significant path of the direct effect, four groups of mediation effects tests were conducted using the user-defined “estimands” function. The bias-corrected bootstrap method was used, with a 95% confidence interval, and repeated sampling was conducted over 2,000 times (Hayes 2009). The following results were obtained: ICT → ATV → ATD → VI (0.010, 0.104), ICT → ATD → VI (0.001, 0.091), and ATV → ATD → VI (0.079, 0.237). These three paths did not contain 0 in the 95% confidence interval, and the mediation effect was significant. The path of C/D-C → ATV → ATD → VI (0.000–0.069) contained 0 in the 95% confidence region, indicating that the mediation effect was not significant (Table 5).

Table 5 Results of mediation effects.
Full size table

Multiple-group analysis

The stability or variation of the estimated model across different groups can be verified through multiple-group analysis, which is applicable to test control variables (grouping variables). Some of the literature has shown that consumers have little intent to purchase a product endorsed by an unfamiliar celebrity (McCormick 2016). Considering that the familiarity level of internet celebrities (including four groups: “unfamiliar with her,” “know something about her,” “often watch her short videos,” and “I am her fan”) would affect the audience’s visit intention, a one-way ANOVA was conducted using internet celebrities’ familiarity as the control variable. The results showed that different degrees of familiarity had a significant impact on the average value of visit intention (F = 15.067, P = 0.000), with the “often watch her short videos” group having the highest visit intention.

Furthermore, familiarity was used as a control variable for the multiple-group analysis to verify whether there were cross-group differences in the research results. In this case, X^2/df=2.358, CN = 297 > 200, RMSEA = 0.046 < 0.05, IFI = 0.916 > 0.9, TLI = 0.901 > 0.9, CFI = 0.914 > 0.9, and ΔCFI < 0.01, meaning they all met the adaptation standard and, therefore, the multiple-group analysis model was appropriate. The results of the multiple-group path analysis showed that there were cross-group differences in the path coefficients but no cross-group differences in significance (Table 6), indicating the results are stable across different groups. In addition, the path coefficient of ICE → VI decreased successively across groups, indicating that the effect of an internet celebrity’s expertise on visit intention decreased with increasing familiarity (Fig. 3).

Table 6 Results of multiple-group analysis.
Full size table
Fig. 3: Path coefficients of different familiarity levels.
figure 3

**p < 0.01, *p < 0.05; ICT Internet Celebrity’s Trustworthiness, ICE Internet Celebrity’s Expertise, VI Visit intention.

Full size image

Conclusion and discussion

First, the internet celebrity’s trustworthiness was found to be vitally important because it significantly affected audiences’ attitudes toward short videos, attitudes toward destinations, and visit intentions. On a social networking platform where photos and videos may have been digitally enhanced, and considering that travel often involves a high level of uncertainty, trust is particularly relevant to travel decision-making, and thus consumers place more importance on factors such as authenticity, objectivity, and sincerity (Yin et al. 2023). Celebrity credibility is important for attracting potential tourists and maintaining the sustainable development of endorsement destinations (Liu et al. 2013). The ultimate purpose of travel is to immerse oneself in the authenticity of the destination; hence, to obtain real experience, internet celebrities should be required to declare whether the camera’s beauty and filter functions are used, enabling the audience to accurately assess the information presented in the videos.

Second, internet celebrities’ expertise only significantly affected visit intention. Internet celebrities often play the role of opinion leaders in their fan communities, relying on their professional knowledge and travel experience, which can provide more valuable travel information to potential tourists. Some studies have shown that skilled internet celebrities can significantly affect consumers’ purchase intentions, reduce the cost of information searches, and even lead to impulsive purchasing behavior (Meng et al. 2020). In addition, the impact of internet celebrities’ expertise to visit intention decreases with increasing familiarity, thereby facilitating the DMOs in inviting suitable internet celebrities for cooperation.

Third, internet celebrities’ attractiveness (i.e., physical attractiveness) was found to have no significant effect on audiences’ attitudes toward the short video, attitudes toward the destination, and visit intention, indicating that appearance was not an effective factor affecting the audience, which is inconsistent with the traditional concept of “beauty is power.” In online environments, the level of physical attractiveness is not the primary attention-drawing feature as internet celebrities can enhance their appearance using the camera’s beauty function, which enables them to smoothen their skin, enlarge their eyes, and slim down their face in videos. Social media audiences are more interested in traveling stories told in short videos than in filtered looks and scenery. Prior research has shown that the more attractive the information source is, the more likely users will be to have a positive attitude, but it has no significant effect on purchase and use intention (Baker and Churchill, 1977). Similarly, Gilal et al. (2020) used advertising models with different degrees of attractiveness to endorse products, reaching a similar conclusion as Baker and Churchill (1977). Therefore, it would be risky for DMOs to choose a more physically attractive endorser than a normal one, which increases the possibility that the audience would remember the internet celebrity but ignore the destination. Internet celebrities’ attractiveness may affect the popularity of short travel videos online, but it has no significant effect on promoting destination marketing.

Additionally, celebrity–destination congruence only affects the audience’s attitude toward short videos, but the direct and indirect effect between celebrity–destination and visit intention was not significant, which is inconsistent with the research of Xu and Pratt (2018). This reveals the difference between advertisements and short videos in destination marketing to a certain extent. The main purpose of advertising is to project the destination image and conduct marketing, which usually reflects the interests and demands of the DMOs. While social short videos dominated by internet celebrities have multiple value pursuits, a well-made short video can be sought after by netizens, monetizing its popularity online. Internet celebrities making travel short videos need to align their personal image with the destination image, such as makeup, clothing, words, habits, and so forth, which is beneficial for short videos to be liked and shared online. This artificial consistency helps short travel videos gain online attention, however, when making travel decisions, audiences still prioritize the destination itself.

Finally, internet celebrities’ trustworthiness and expertise are factors significantly affecting visit intention, while internet celebrities’ attractiveness and internet celebrity–destination congruence are not significant in this regard (Fig. 2). The perception of social media audiences toward internet celebrities mainly arises from their short videos, comparatively, the perception of consumers toward traditional celebrities (stars) can be more broadly derived from their works, public events, and news reports, which highlights the importance of trustworthiness and expertise in online influencers’ marketing. In terms of the mediating effect, internet celebrities’ characteristics can be explained by four manifest variables (ICT, ICT, ICA, and C/D-C) in the structural model. Still, the dual mediation only played a role between ICT and visit intention. In comparison, the other three variables were not significant, indicating that trustworthiness still plays a major role in the chain path of “internet celebrities’ characteristics - short video attitude - destination attitude - visit intention.” If attitude toward the short video is taken as the starting point, the path of “short video attitude - destination attitude - visit intention” is always significant, which is consistent with the conclusion of dual mediation in advertising research. In other words, without considering internet celebrities’ characteristics, as long as the short video can be liked and supported, it can also make the audience have a positive attitude toward the destination and promote their intention to travel. The impact of an internet celebrity’s short video on a destination differs from that of traditional celebrity endorsements. For the former, the content and narrative of the video have a stronger impact, while for the latter, the celebrity’s charisma and appearance are more important.

Theoretical implications

The present study contributed to the knowledge concerning internet celebrity endorsements in tourism and offered a distinction between celebrity endorsements and online influencer marketing. “Appearance Fallacy” exists in internet celebrity endorsement, high physical appearance often leads audiences to remember internet celebrities rather than the destination. Furthermore, the authentic part of short videos plays a greater role than the constructed part, this concept addresses a strong explanation for issues related to internet celebrity endorsements. In online influencer marketing, information (short videos) is superior to the information source (internet celebrities), while in traditional celebrity endorsements, the information source (celebrities or stars) is superior to the information (advertisement). This explains why trustworthiness and expertise are more effective than attractiveness in online influencer marketing.

The present study has presented value to elicit the causal mechanisms in theory building, filling the gap in travel short videos of internet celebrities. Prior research has mainly focused on social media advertising and official promotional videos in tourism. As a mediation between tourists and destinations, short videos created by internet celebrities serve not only as promotions and recommendations but also fulfill the function of providing an experiential encounter. Internet celebrities’ short videos are comprehensive and interactive content that integrates the characteristics of Internet celebrities, destination scenery, and travel stories. These videos not only foster high audience engagement but also facilitate the establishment of a strong internet celebrity–fan relationship, consequently, they have emerged as an effective marketing strategy. The present study revealed the mechanism of how short travel videos affect tourism decision-making, thereby facilitating further in-depth research in this domain.

The current research extended the application of source credibility theory in the field of short video marketing. Some scholars have examined the effectiveness of celebrity endorsements and noncelebrity endorsements on advertising; in the present study, internet celebrity was defined as a third-party endorsement. In this kind of endorsement, internet celebrities’ trustworthiness and expertise are highlighted, while the impact of attractiveness was found to be insignificant, making trustworthiness and attractiveness seemingly incompatible factors for an effective marketing strategy. This further indicates that the role of the three dimensions of source credibility varies across different contexts; academic studies will provide mixed support for source credibility theory in the future.

Managerial implications

Trustworthiness and expertise are more effective than physical appearance when DMOs invite internet celebrities to project short video marketing, thus, the video content should be credible, authentic, original, and informative (Joshi et al. 2023). Normal-looking internet celebrities can attract people’s favorable impressions more than attractive ones and may be perceived as more honest and trustworthy (Bower and Landreth 2001). For instance, a normal-looking and unpretentious endorser with rich tourism knowledge can effectively promote consumers’ positive perceptions and visit intentions. Furthermore, the relationship between a native-born celebrity and a destination cannot be duplicated (Kwon and Vogt 2010), which means that a locally-born internet celebrity with a local accent can increase credibility and emotional connection between the destination and tourists.

As third-party endorsements, internet celebrities’ short videos essentially belong to the content jointly produced by internet celebrities and the DMOs. Internet celebrities intend the videos to be popular and monetized online, while the DMOs expect more video audiences to visit the destination. In this sense, effective travel short videos should reflect both internet celebrity attributes and destination characteristics. Compared with traditional advertising, short travel videos offer a more immersive experience; simultaneously, the essence of travel lies in experiencing an unusual environment. The “localism” of video presentation holds particular importance as it enables potential tourists to gain unique pre-experiences before traveling. Therefore, consistency between video style and destination characteristics is crucial in implementing effective online influencer marketing.

A dedicated video team should be formed to produce original short videos. Unlike the official promotional videos generated by DMOs, internet celebrities’ short videos are more informal, life-like, and emotional. Interesting and touching elements are the keys to success. Short travel video marketing should delve into the local culture, achieve a profound fusion of culture and tourism, and present the destination’s authenticity through a narrative perspective. Aesthetic fatigue is inevitable when encountering a large body of short videos on social media platforms, authentic, unadorned videos are more likely to resonate with audiences. Therefore, authenticity pursuit is superior to aesthetic pursuit for short video audiences, making portraying a naive image the primary marketing strategy for promoting a travel destination.

Limitations and future research

The current study took FangQi Kiki as a case study because she is a typical and representative internet celebrity in the field of short travel videos and has a large body of works and followers on TikTok. Using her as a case was conducive to conducting in-depth and intensive research with informational and heuristic case factors. Future research could conduct a multicase comparative study employing a mixed-methods approach, incorporating qualitative data such as social media content, web text, and interviews.

Social media audiences have become accustomed to the beauty enhancements in short videos, resulting in a desensitization towards physical attractiveness. Future research can use the video’s beauty or filter degree as a moderating variable to explore the impact of video modification on the audience’s travel intention and actual perceptions through experimental methods. DMOs need to consider using original videos to promote a destination’s naive image. This does not mean that internet celebrities cannot beautify videos, which depends on whether actual visitors perceive themselves to be misleading or not.

Short video narration can promote brand attitude and reduce individual psychological resistance to a destination (Cao et al., 2021). In addition, short videos not only influence travel decisions, but sharing travel stories in short videos may also reshape the tourist experience at a destination (Wang and Park 2012). Therefore, sharing a story or real-life experiences in a short video is more effective than simply showing the landscape. Future research can use personal narratives as a variable to explore their impact on audiences’ video experience and travel experiences. The topic of experience co-creation between online influencers and destinations deserves an in-depth study.