Millennial social media users ’ intention to travel: the moderating role of social media influencer following behavior

Purpose – Social media (SoMe) influencer marketing is a popular practice. The current study examines the interplays between SoMe influencers ’ source credibility, Millennial users ’ attitudes and intention to travel. It further investigates the moderating role of SoMe influencer following behavior on the aforementioned relationships. Design/methodology/approach – Atotalof212useableresponseswerecollectedthroughanonlinesurvey. Structural equation modeling and hierarchical multiple regressions were employed for hypotheses testing. Findings – Results indicated that source credibility had a significantly positive influence on the SoMe users ’ attitudes,whichinturnwaspositivelyassociatedwiththeintentiontovisittheendorseddestination.Moreover, bothrelationshipswerestrengthenedforSoMeinfluencerfollowersthanfornonfollowers. Originality/value – The study expanded the source credibility theory to the use of SoMe influencer marketing on travel destinations among Millennial SoMe users. In addition, the research applied the self-determinationtheorytofillthegapinliteraturebyexaminingthemoderatingroleofSoMeinfluencerfollowingbehavior.


Introduction
With 3.5bn active social media (SoMe) users worldwide, spanning from major networking apps such as Twitter, Instagram, Facebook and Snapchat, SoMe is continuously increasing its power in marketing brands and influencing consumer behavior (Mohsin, 2020;Smart Insights, 2018). SoMe is changing not only the behavioral intentions and decision-making process in purchasing products but also the ways tourists gather information, decide where to travel and share their experiences (Xiang and Gretzel, 2010). The utilization of different SoMe platforms has become more prominent as consumers and tourists increasingly rely on different reviews and posts of user-generated content. Previous research suggests that destination marketing organizations (DMOs) can benefit from using SoMe platforms as a destination marketing tool (Hays et al., 2013;Lange-Faria and Elliot, 2012). With recognizing SoMe as a means of socializing with others, DMOs can streamline authentic travel information and build their brand through direct interaction with the consumer (Lange-Faria and Elliot, 2012).
Experts believe that the impact of SoMe is particularly strong on Millennials (also known as Generation Y, born between 1981 and, who are the first SoMe generation who grew up with and exposed to readily available technology (Gursoy et al., 2013;Smith and Anderson, 2018). As of early 2018, in the USA, 18-29 year olds account for 88% of users on SoMe platforms (Smith and Anderson, 2018). This age range dominates most SoMe platforms by having 91% use on YouTube, 81% use on Facebook, 68% use on Snapchat and 64% use on Instagram (Smith and Anderson, 2018).
Influencer marketing is a wildly popular practice in SoMe, which many brands incorporate in their strategy to impact the decision-making process of their consumers (Claude et al., 2018). Influencer marketing refers to the relationship between a brand and an influencer where the influencer promotes a brand's products or services on various SoMe outlets (Mathew, 2018). Sixty-one percent of consumers rely heavily on SoMe for information on products while shopping (Kramer, 2018). Researchers have found significant impact of influencer marketing and how it influences consumer purchasing habits (Claude et al., 2018;Glucksman, 2017). However, limited research exists regarding the impact of SoMe influencers on Millennials' travel behavioral intentions despite their significance in tourism. Millennials are the generation who are willing to pay more for experiences and who will consistently contribute to the tourism industry as they will have the most purchasing power in a few years (Sofronov, 2018). In addition, there are over 90.4% of the Millennials actively using SoMe in 2020 (Mohsin, 2020). With 38% of Millennials trusting digital influencers and their contents, influencer marketing is a significant factor in determining the decision-making process of this generation (Klein, 2018).
Hence, the purpose of the current study is to examine how perceived source credibility of SoMe influencers' post influences Millennial SoMe users' attitudes, which may further influence their intention to visit the destination. Furthermore, the study investigates the moderating role of SoMe users' influencer following behavior (IFB) on the aforementioned relationships. The study will contribute to the current literature by expanding the influencer marketing research into the tourism field. The investigation of the moderating role of SoMe IFB will help fill the gap in the current influencer marketing literature. The findings of the study will provide destination marketing companies with practical strategies to effectively utilize SoMe marketing tools and influencer sponsorship specifically tailored to Millennials.
2. Literature review 2.1 Influencer marketing Influencer marketing has been increasing significantly over the years as companies believe that there is 11 times more of a return on investment with influencer marketing than traditional forms of marketing (Tapinfluence, 2017). With the developing wave of interactive SoMe, influencer marketing is impacting the marketing scene, specifically for Millennials. SoMe influencer marketing uses third-party influencers to assist in promoting the brand's message, product or service and connecting to the suggested target market, through a SoMe outlet . SoMe influencers who participate in such marketing represent a new type of independent third-party endorsers who shape audience attitudes through blogs, tweets and use of other SoMe forms (Freberg et al., 2011). Influencers typically engage with their followers by regularly updating their audience with the latest information . Their contents are not only available for their followers, but for everyone in public on SoMe platforms. Users are allowed to comment and tag their friends and families on these posts and videos, which expands the spreading. A joint study by Twitter and Annalect has revealed that 49% of participated users indicated they rely on recommendations by influencers (Swant, 2016). Companies also recognize the importance of influencer marketing and desire to build a more direct relationship with the consumers in alliance with influencers (Korotina and Jargalsaikhan, 2016). A number of industries believe that engaging with influencers can help build a more authentic connection with their target customers (Cultureshop, 2015).

Social media users' intention to travel
Influencer marketing already has been shining light in the travel and tourism industry. Certain SoMe influencers create not only product content on different platforms but also valuable travel information that serve as travel motivation and decision-makers. Popular destination travel influencers on Instagram include the Bucket List Family with 1.6m followers and Jessica Stein with 2.5m followers. Both influencer pages have a diverse following and a large reach to specific target markets (Gretzel, 2018). These prominent travel influencers establish a word of mouth in the travel context, which spreads among a vast number of followers on SoMe. SoMe, which essentially releases content created by influencers, is a major source of travel information and developing factor of travel intentions (Park, 2015). Gretzel (2018) states that marketers have recognized the opportunities of amplifying messages targeted toward specific niche afforded by SoMe influencers. Destination marketing has already been utilizing SoMe as a critical tool in promoting destinations (Bokunewicz and Shulman, 2017;Hays et al., 2013;Pan et al., 2007). Travel marketers can benefit from influencers' ability to reach the targeted audience, engage on a higher level and build relationships and images that align with their brands (Gretzel, 2018).

Millennial SoMe users and perceived source credibility
Millennials, also referred to as Generation Y, were born roughly between 1980 and the mid-2000s (Pentescu, 2016). They are known to be the most tech-savvy group of people, hailed as "digital natives" (DeVaney, 2015). Millennials have accepted and familiarized themselves with technology early on. For Millennials, the use of mobile technologies is very intensive and diverse in planning all trip stages (Femenia-Serra et al., 2019). Especially, the use of SoMe is prominent on their mobile devices. In a generational analysis done for the Internet use of trip planning, employing many online intermediaries, search engines and paying special attention to user-generated content on SoMe are shown to be specific traits of the Millennials (Kim et al., 2015). Millennials spontaneously seek for destinations and plan their trips on SoMe. Taking pictures and sharing them on different SoMe platforms were the main reasons for using their mobile devices during trips (Gotardi et al., 2015).
Source credibility is defined as "believability," and it is one of the main factors in persuasion (Fogg and Tseng, 1999;Jaso, 2011). Source credibility theory states that people are more likely to be persuaded when the source presents itself as credible (Hovland et al., 1953). Source credibility theory is an established theory as it has received comprehensive examination (Hovland and Weiss, 1951;Ohanian, 1991;Umeogu, 2012). Especially, source credibility has been thoroughly studied in marketing and its relation to consumers in the online context (Ayeh, 2015;Lowry et al., 2013). Researchers found that receivers tend to have respect and readily accept the words of communicators with high level of source credibility (McCroskey et al., 1974).
Source credibility is being presented in recent studies along with the rise of SoMe influencers. The information available on SoMe and the recency of such information impact the perceptions of source credibility (Westerman et al., 2014). Many researchers argue that source credibility is a major player in establishing successful influencer marketing (Djafarova and Rushworth, 2017;Grafstr€ om et al., 2018;Xiao et al., 2018;Zietek, 2016). When influencer's credibility decreases, the potential to influence also decreases (Zietek, 2016). Djafarova and Rushworth (2017) show that bloggers, YouTube personalities and "Instafamous" profiles are powerful and influential in the purchase behavior of young female SoMe users. They are essentially categorized as influencers who often participate in influencer marketing in various fields. Findings from a study confirm that influencers are a key component in fashion marketing as they tend to have high authenticity and not motivated by monetary reasons (Zietek, 2016). Many SoMe influencers already have a large group of followers, which is presented as more likeable because they are considered more popular (De Veirman et al., 2017).

IHR 36,2
For influencers who are transparent, respect and trust build along with connections with the targeted market and may result in greater impact on influencing the consumer behavior (Kramer, 2018). Previous literature has shown a positive relationship between influencer credibility and followers' behavioral intentions (Lou and Yuan, 2018;Weismueller et al., 2020). However, studies of SoMe influencers' source credibility in the context of tourism are still scant. More research is needed to better understand how influencer source credibility may play a role on SoMe users' attitudes and intention to travel.

Attitudes toward social media influencer marketing
Eagly and Chaiken (1993) described attitude as a tendency to evaluate an entity with some degree of favor or disfavor, ordinarily expressed in cognitive, affective and behavioral responses. Attitudes are considered to shape people's minds as better the attitude a person has toward a brand, the more likely he or she is to use the product of that brand. Akar and Topcu (2011) have found that consumers' use of SoMe, the frequency of the use, their knowledge of SoMe, their following of SoMe and their fears about marketing with SoMe all affect their attitudes toward SoMe marketing. Various research findings have demonstrated that consumers' attitudes toward SoMe influencer marketing were positively associated with their purchase intentions of the endorsed product (Evans et al., 2017;Lim et al., 2017).
SoMe influencer marketing is being commonly practiced by many organizations to reach their potential consumers. The access to the audience and the engagement are main benefits of using influencer marketing. According to the source credibility theory (Hovland et al., 1953), individual users tend to be persuaded if the influencer is perceived as being credible, expert and trustworthy. Influencers are "who the consumers are looking at" and the brand's message can be amplified via word of mouth, which can potentially trigger purchasing intentions (Woods, 2016). Kolarova (2018) stresses the importance of influencer's SoMe message being presented as a brand and sponsored. Congruency between the influencer and the advertised brand is significant in the levels of perceived source credibility as it may further impact consumer behavioral intentions (Breves et al., 2019). Information presented by the influencer is perceived as credible and to highly affect users' beliefs, opinions, attitudes and behavior . Research findings indicate that the credibility of the influencers contributes to attitudes toward influencer marketing on SoMe (Xiao et al., 2018;Yoo and Jin, 2015). In the tourism context, studies suggest that the impact of perceived source credibility on online travelers' attitudes toward user-generated content is significant (Ayeh et al., 2013;Ayeh, 2015). Also, Millennials consider influencer marketing negatively when they perceive the influencer as being less credible (Grafstr€ om et al., 2018). Therefore, source credibility serves as a determining factor in forming consumer attitudes toward influencer marketing , which should also apply to Millennial SoMe users' attitudes toward tourism-related SoMe influencer marketing. Hence, the following hypothesis is posited: H1. SoMe influencers' source credibility is positively associated with Millennial SoMe users' attitude toward SoMe influencer marketing.

Intention to visit
Several studies have argued that attitude is regarded as an antecedent of behavioral intention (e.g. Chen, 2007;Vermeir and Verbeke, 2007;Wang and Ritchie, 2012). Just as when a consumer favors a certain product, his/her purchasing intention increases. The same can apply when a service or destination is marketed by a consumer's favorite influencer. Various industries have presented studies regarding influencers' impact on consumers' attitudes and behavioral intentions (Bush et al., 2004;Evans et al., 2017;Lim et al., 2017). Lim et al. (2017) Social media users' intention to travel demonstrated that consumers' attitudes toward SoMe influencer marketing were positively associated with their purchase intention of the endorsed product.
In the tourism field, researchers indicated that travelers' attitudes toward user-generated content were positively related to their intention to use the information in planning their trips (Ayeh et al., 2013;Ayeh, 2015). More specifically, influencer marketing has been found to have impact on the decision-making of Millennials when choosing a rural tourism area as a potential destination (Chatzigeorgiou, 2017). A recent study shows that Chinese Millennial tourists' attitudes toward the advertisement lead to positive attitudes toward the endorsed destination, which may further transform into higher likelihood of visit when there exists a higher degree of congruence between tourists and the SoMe influencer (Xu and Pratt, 2018). Therefore, the following hypothesis is proposed: H2. Millennial SoMe users' attitude toward SoMe influencer marketing is positively associated with their intention to visit the endorsed destination.

The moderating role of SoMe influencer following behavior
Most SoMe influencer marketing studies to date only focused on SoMe influencers' impact on their followers (e.g. Ki et al., 2020;Mart ınez-L opez et al., 2020). Few studies have examined whether SoMe influencers' posts have the same effect on followers and nonfollowers as the content is available to all SoMe users. Previous study suggested that consumers' attachment to celebrities can vary in degrees of strength (Thomson, 2006) and the degree of SoMe users' attachment to SoMe influencers may also vary depending on their following behavior. The strong attachments to SoMe influencers are associated with strong feelings of connection, affection and passion (Thomson et al., 2005). The self-determination theory can be used to understand the various degrees of strength of SoMe users' attachment to SoMe influencers. The self-determination theory can be described as an approach to human motivation and personality that highlights inherent growth tendencies and innate psychological needs as the basis (Ryan et al., 1997;Ryan and Deci, 2000). This particular theory has been embodied in previous studies as a framework for human behavior and motivation in which three types of needscompetence, relatedness, autonomyare identified as the notion for psychological needs (Engstr€ om and Elg, 2015; Gagn e and Deci, 2005; Gilal et al., 2009;Lin et al., 2009;Patrick et al., 2007). La Guardia et al. (2000) carefully examined the needs where each was significant in the development of attachment and of which the needs for relatedness served as the strongest predictor of attachment. In terms of the context in the current study, it can be suggested that individuals can develop feelings of attachment through the need for relatedness, which can be defined as the need for belonging to and connectedness with others (Engstr€ om and Elg, 2015; Palmatier et al., 2008;Ryan and Deci, 2000). Therefore, it is reasonable to assume that SoMe users who follow a specific SoMe influencer will feel connected to and tend to develop a stronger attachment to that influencer than those who do not follow. In addition, previous study indicated that higher level of attachment will result in higher level of trust and commitment (Thomson, 2006). Consumers' attachment to the influencer will have a stronger effect on their attitudes toward the endorsement and will result in a higher level of purchase intention (Ilicic and Webster, 2011). Following the same logic, SoMe IFB as an indicator of influencer attachment can lead to a higher level of perceived source credibility of the influencer and enhance Millennial SoMe users' attitudes toward the influencers' post as well as result in a higher level of intention to visit the endorsed destination. Therefore, the relationship between SoMe influencer source credibility and Millennial SoMe users' attitude and the relationship between attitude and intention to visit both will be strengthened for SoMe influencer followers than for nonfollowers. Hence, the following hypotheses are proposed: (see Figure 1).

IHR 36,2
H3. SoMe IFB moderates the relationship between SoMe influencers' source credibility and Millennial SoMe users' attitude so that the relationship is stronger for followers than for nonfollowers.
H4. SoMe IFB moderates the relationship between Millennial SoMe users' attitude and their intention to visit the endorsed destination so that the relationship is stronger for followers than for nonfollowers.

Sample and data collection
The target population of the current study is American Millennial (1981-2001) SoMe users with an active Instagram account. Instagram was chosen since it is one of the most popular SoMe platforms for influencer marketing (Evans et al., 2017). Tourism marketers are turning to Instagram to assist interaction with their target market (Pereira, 2019). In addition, Millennials are picture-driven with online communication and are more inclined to use Instagram over other SoMe platforms such as Facebook (Ting et al., 2015). A self-report online survey prepared via Qualtrics was employed for data collection. Data were collected in April 2019. Convenience sampling and snowballing sampling approaches were used where the survey link was posted on different SoMe platforms such as Facebook, Instagram and Snapchat since the target population is tech-savvy. Participants were asked to share the survey links on their own SoMe or use personal referrals to encourage their followers, friends, families to participate in the survey. To ensure that only Millennial SoMe users with Instagram accounts participate in the study, participants were required to indicate whether they were born after 1981 and whether they have an Instagram account. Participants were allowed to continue with the survey only if their answers to both questions were "yes." A brief definition of SoMe influencer and examples of SoMe influencer were then provided to the qualified participants before answering the questions. Participants were instructed to think of a specific SoMe influencer on Instagram and respond to questions about the SoMe influencer posts' source credibility and their attitudes toward this SoMe influencer's posts. Then, participants were asked to indicate their intention to visit a travel destination if this specific SoMe influencer posts about a destination on Instagram.

Measurement scales
All the scales used in the current study were adopted from prior studies and have been validated extensively. Attitude toward SoMe influencer's posts was measured using a sevenpoint semantic differential scale. It is adopted from Moon and Kim (2001) with five items such as "unpleasure/pleasure" and "unfavorable/favorable." Source credibility and intention to visit were both measured using a seven-point Likert scale, ranging from 1 (strongly disagree) to 7 (strongly agree). Source credibility was measured by three-item scale adapted from Xu and Chen (2006). Example of items includes "I think the content of this SoMe influencer's posts is consistent with facts." Intention to visit was measured with two items adapted from Social media users' intention to travel Chen et al. (2014) such as "If I get the chance to travel, I intend to visit the destination mentioned in the SoMe influencer's posts." SoMe IFB was measured by whether or not participants are following the SoMe influencer on Instagram. Demographic information such as age, gender, education and frequency of SoMe use was also collected from the participants. A detailed list of measurement items can be found in Table 2.

Data analysis
Data cleaning and descriptive data analysis were performed with SPSS 24.0. Confirmatory factor analysis (CFA) with maximum likelihood (ML) estimation was used to examine measurement reliability and validity. Common method bias was checked with CFA following Podsakoff and Organ's (1986) procedure. Moreover, structural equation modeling was carried out through SPSS Amos 23.0 to test hypotheses 1 and 2. Moderation hypotheses 3 and 4 were examined using hierarchical multiple regressions.

Results
A total of 267 responses were collected from the online survey, out of which 55 were either incomplete or invalid, resulting in 212 usable responses. Close to 65% of the respondents were female. Almost 75% of the participating Millennials are in the age group of 18-24. The majority of them (78.3%) have travel plans in the following 12 months. Nearly 84% of the participants indicated that they use Instagram at least once a day. Among the 212 respondents, the most popular Instagram activities are viewing timelines only (46.2%) or viewing both timeline and suggested pages (47.2%). The demographic information of the respondents is presented in Table 1. CFA was conducted using the ML estimation to examine the model fit. Tables 2 and 3 show the standardized factor loadings and fit statistics, indicating a good fit between the theoretical model and the data (χ 2 (30) 5 41.94, p 5 0.07; comparative fit index (CFI) 5 0.99; goodness of fit index (GFI) 5 0.96; root mean square error of approximation (RMSEA) 5 0.04; standardized root mean square residual (SRMR) 5 0.03).
Both Cronbach's alpha and composite reliability (CR) of the construct were used to measure the internal consistency. Both Cronbach's alpha and CR values ranged from 0.85 to 0.95, showing good internal consistency. Convergent validity was investigated using factor loadings and each construct's t-value to check whether it is statistically significant (Bagozzi and Yi, 1988). Table 2 indicates that all factor loadings are greater than 0.70 (Bagozzi and Yi, 1988) and were significant at 0.05. Moreover, the average variance extracted (AVE) values for all constructs (0.74, 0.79, 0.85) are above 0.50 (Fornell and Larcker, 1981), suggesting good convergent validity. Fornell and Larcker (1981) suggested that discriminant validity could be tested by comparing squared pairwise correlations between constructs and the AVE value of each construct. Results demonstrate that the square root of each construct's AVE value (between 0.86 and 0.92) is higher than its correlations with other constructs (Table 3). Hence, discriminant validity is confirmed, indicating that each construct is statistically different from the other.
A structural model was estimated using ML through SPSS Amos 22.

Social media users' intention to travel
The results demonstrated that the overall fit of the structural model was adequate, (χ 2 5 76.52, df 5 32, p < 0.05; CFI 5 0.98; GFI 5 0.93; RMSEA 5 0.08; χ 2 /df 5 2.39). The results showed that source credibility is positively related to SoMe users' attitudes toward influencer marketing (0.62, p < 0.01), suggesting that the more credible SoMe influencers' posts are, the more favorable attitude SoMe users tend to hold toward their post. Similarly, SoMe users' attitude is positively associated with their intention to visit the endorsed destination (0.33, p < 0.01). This result indicates that the more positive attitude SoMe users exhibit toward SoMe influencer's post, the higher level of intention they will have toward visiting the destination. The results lend support to both hypotheses 1 and 2. Following Baron and Kenny (1986), hierarchical multiple regression analyses were used to test the moderating role of SoMe IFB. IFB was dummy coded (1 5 following SoMe influencer, 0 5 not following SoMe influencer). Both source credibility and IFB were standardized prior to multiplication. Source credibility and IFB were entered in Step 1, and the interaction term (SC 3 IFB) in Step 2 to predict SoMe users' attitude toward influencer marketing. According to the results in Table 5, the interaction of source credibility and IFB is significant (β 5 0.28, p < 0.05), and the interaction accounts for the significant incremental variance of SoMe users' attitude toward influencer's post (ΔR 2 5 0.02, p < 0.05). Following a similar procedure, the moderating effect of IFB on the relationship between SoMe users' attitude and their intention to visit the endorsed destination is examined. The results indicated significant interaction between SoMe users' attitude and IFB (β 5 0.34, p < 0.05). The interaction accounted for the significant incremental variance of intention to visit (ΔR 2 5 0.03, p < 0.05).
To understand the nature of interaction effects, we plotted SoMe users' attitude score at combination of the mean ± 1 SD (high and low levels) for both source credibility and IFB. The interaction plot (Figure 2) indicates that the effect of source credibility on SoMe users' attitude is positive for both SoMe influencer followers and nonfollowers. The positive relationship is enhanced when SoMe users follow influencers on SoMe, supporting hypothesis 3. Similarly, we plotted SoMe users' intention to visit score at combination of the mean ± 1 SD (high and low levels) for both users' attitude and IFB. The second plot (Figure 3) shows that the effect of users' attitude on intention to visit was positive for both SoMe influencer followers and nonfollowers. The positive relationship between attitude and intention to visit was strengthened for SoMe influencer followers than those who do not follow. Therefore, hypothesis 4 is supported.

Discussion and implications
The results show that there is a significantly positive relationship between source credibility and Millennial SoMe users' attitudes toward influencers' posts, which is consistent with the source credibility theory that people are more likely to be persuaded when the source presents itself as credible and people tend to have respect and readily accept the words of communicators with high level of source credibility (Hovland et al., 1953;McCroskey et al., 1974). It is important to note that many previous studies proved that source credibility is a key element in establishing successful influencer marketing (Djafarova and Rushworth, 2017;Xiao et al., 2018;Zietek, 2016). Additionally, a research has demonstrated that high source credibility leads to more favorable attitudes (Tormala et al., 2006). Similar results were revealed in a previous study testing the effects of source credibility on attitudes and intention toward user-generated content for travel planning (Ayeh et al., 2013). The current study, on Independent variables Dependent variables and standardized regression weights (β) Step 1 Step 2 The current study further progressed into the positive association between Millennial SoMe users' attitude and intention to visit the endorsed destination, which aligns with Chatzigeorigiou's (2017) finding that influencer marketing has an impact on Millennials' decision-making when choosing a rural tourism destination. Furthermore, the result echoes a previous finding, which indicated consumers' attitude toward SoMe influencer marketing was positively related to their intention to purchase the endorsed product . In the tourism context, travelers' attitudes also have a strong influence on their intentions to purchase travel product online (Amaro and Duarte, 2015). The finding is also consistent with a recent study that Chinese Millennials' attitude toward the SoMe influencer endorsed destination may lead to higher level of intention to visit when there exists a higher degree of congruence between the person and the SoMe influencer (Xu and Pratt, 2018).
The results further indicated that positive relationship between perceived source credibility and Millennial SoMe users' attitudes was strengthened for SoMe influencer followers than nonfollowers. Similarly, the relationship between Millennial SoMe users' attitudes and intention to visit the endorsed destination was strengthened for SoMe influencer followers. This could be because Millennial SoMe users' attachment to SoMe influencers varies in degrees of strength. Based on the self-determination theory, human behavior and motivation are identified with three types of psychological needscompetence, relatedness and autonomy (Engstr€ om and Elg, 2015;Gagn e and Deci, 2005). Among them, the needs for relatedness served as the strongest predictor of attachment (La Guardia et al., 2000). As the need for relatedness refers to the need for belonging to and connectedness with others (Engstr€ om and Elg, 2015;Palmatier et al., 2008;Ryan and Deci, 2000), SoMe followers tend to develop a higher level of attachment to the SoMe influencer than nonfollowers due to the stronger sense of relatedness and could result in a higher level of trust and commitment (Thomson, 2006). Consistent with previous finding that consumers' attachment to influencers Effect of UA and IFB on SoMe users' intention to visit IHR 36,2 will transfer into positive attitudes toward the endorsement and will result in a higher level of purchase intention (Illicic and Webster, 2011), SoMe influencer following behavior can moderate the relationship between SoMe influencer source credibility and Millennial SoMe users attitudes and the relationship between Millennial SoMe users' attitude toward the post and their intention to visit the endorsed destination.

Implications
The current study expands the source credibility theory to the use of SoMe influencer marketing on travel destinations among Millennial SoMe users and adds to a growing body of evidence suggesting that source credibility is highly associated with SoMe users' attitudes toward SoMe influencer marketing. In addition, the research applied the selfdetermination theory to fill the gap in the literature by examining the moderating role of SoMe influencer following behavior. The study further expanded the current knowledge of SoMe influencer following behavior by suggesting that the positive relationship between SoMe influencer source credibility and SoMe users' attitudes and the positive relationship between attitudes and Millennial SoMe users' intention to visit the endorsed destination are strengthened for Millennial SoMe influencer followers than nonfollowers.
For application in general circumstances, destinations marketers can invite influencers whose opinions are perceived more trustworthy as destination endorsers (Xu and Pratt, 2018). Marketers should carefully choose the SoMe platforms that are most frequently used by their target travelers to have an ongoing engagement and interaction level with them. Destinations can recognize credible influencers by using those with accurate and reliable content along with transparent image. It is considered that the reputation and image of SoMe influencers are gradually built by the users' input and their opinions. For example, destination marketers and tourism operators can select travel-related SoMe influencers who have an adequate number of Millennial followers and who regularly post information that is perceived useful and reliable by the followers. In addition, it would be critical to ensure there is alignment between the destination and the selected influencer, which would have a higher impact on the target audience. For example, historic and jazzy city should be endorsed by a vibrant, artsy person with a unique personality. As for SoMe influencers, auditing the quality of follower niche, limiting sponsored contents on the timeline and disclosing sponsored marketing campaigns as necessary can help build source credibility. Engaging with followers frequently on SoMe platform such as showing their personality and tone of voice on live stories and IGTV can also help build credibility and form follower loyalty.
The results between attitude and intention imply that destinations and tourism operators should further develop marketing strategies to trigger travel intentions from the positive attitudes of Millennial SoMe users. Various marketing strategies and influencer campaigns can be practiced in order to increase brand awareness and drive sales, such as distributing unique codes and offering giveaway trips to a destination (Patel, 2019). CVBs or DMOs can work with influencers to create promo codes that offer discounts or free admission to local events for the niche market. Free getaway trips can also be arranged occasionally where the package may include a hotel stay and free breakfast meals. Also, with the impact of COVID-19, partnering with local communities, such as neighboring resorts and restaurants, will help design the authentic and immersive experiences. Influencers may practice these types of strategies to market the destination to their followers. In addition, creating not only captivating headlines when establishing mentioned promotion, but also providing genuine cause of content and honesty are suggested to impact Millennial consumers' attitudes as well as impact their travel intentions.
Destination marketers and tourism operators should also strategize to choose influencers with a high number of Millennial followers as results indicated that SoMe influencer followers Social media users' intention to travel tend to exhibit an enhanced association for the relationship between credibility and attitude as well as the relationship between attitude and intention to visit. Providing positive brand experience and engagement through the selected influencers may potentially lead to purchase intentions. Destination marketers can also share cross-post content from the influencers on their destination SoMe account. It is suggested to work with SoMe influencers who especially have a group of Millennial followers as it will enhance promoting awareness of the account and more importantly the travel destination.
The study also provides practical implications to destination marketers during the COVID-19 recovery stage. With travel restrictions and policies are still forecasted to be in place for longer to prevent the spread of COVID-19, domestic tourism is suggested to be the main engine for driving economic recovery (OECD, 2020). Once the travel restrictions are lifted, domestic tourism will be likely to recover more quickly than international travel. Hence, destination marketers should consider domestic travel for SoMe influencer marketing as stay-at-home orders have led to 44% of the travelers increased their time spent on browsing SoMe (World Travel and Tourism Council, 2020). Moreover, building customer confidence and trust online will be a key when utilizing influencer marketing since most of the communication is transferred to online platforms. In addition, it has been noted that travelers' awareness of good hygiene practices is crucial (World Travel and Tourism Council, 2020). Making sanitizing information known in a clear way via reliable sources would be critical from a traveler's point of view. Destination marketers can collaborate with credible SoMe travel influencers to reassure travelers of disinfecting practices at targeted places to ensure high quality of safety as well as service since travel influencers will most likely be the first to travel and their perspectives will bring people an idea of what the new travel portrays and help with rebuilding travel confidence in travelers.

Limitations and future studies
The majority of the participants of the study were in the age group of 18-24, which might have influenced the results. Future study can include SoMe users from other generations and investigate the moderating role of generation. In addition, the study utilized a self-report survey. Social desirability bias might be an issue. Moreover, data were collected before the COVID-19 pandemic, the findings may not be applied to SoMe users' travel intentions during COVID-19. Future research may explore factors that can drive SoMe users' travel intentions during the pandemic and study what actually determines perceived source credibility from the users' perspective. Furthermore, future study can examine the proposed relationships in the context of hotel industry and investigate the moderating role of different types of guest (business vs leisure).