How User-Generated Content Affects Tourist Loyalty Behaviour at Cultural Heritage Sites in Indonesia

Social media became a primary solution for fulfilling social needs and staying connected during the COVID-19 pandemic, as individuals turned to digital platforms to maintain social interactions and seek information [1]. This phenomenon accelerated the role of social media as a powerful marketing tool in the tourism industry, enabling destinations to reach global audiences through easily shareable content. Social media facilitates the dissemination of UGC, which offers travelers first-hand, authentic insights that influence perceptions and decision-making processes.

Generation Z, born between 1997 and 2012, is particularly influential in this digital transformation. Empirical studies show that Generation Z prioritizes authenticity, emotional engagement, and visually appealing content in their travel choices, relying more on peer-shared experiences over traditional marketing [2]. This generation, representing 40% of global consumers, is highly active on social media platforms, making them a key audience for UGC-based tourism promotion [3]. Their digital savviness and preference for emotional narratives have reshaped how cultural heritage tourism engages with younger travelers, emphasizing the need to understand their behavior in response to UGC. Studies have shown that Generation Z travelers actively seek visually engaging and authentic content that aligns with their values and aspirations, making UGC a crucial tool for influencing their travel decisions and fostering destination loyalty [4].

Within the tourism landscape, UGC has proven especially valuable for cultural heritage tourism. UGC—categorized into factual and emotional content—offers authentic narratives and practical information that shape destination image and tourist behavior. Emotional UGC, such as engaging visuals and stories, fosters emotional connections and enhances perceptions of a destination, while factual UGC provides useful details like hours of operation, pricing, and promotions [5].

These content types play a key role in tourist loyalty behaviour that consists of revisit intention, WOM, and satisfaction. Recent studies highlight that revisit intention is tourists' likelihood to return to a destination based on previous positive experiences and satisfaction, often influenced by memorable and engaging elements at the site [6]. WOM represents tourists' willingness to share their positive experiences, which significantly affects the destination's image and can drive new visits by influencing others' travel choices [7]. Satisfaction is the positive outcome when tourists’ experiences align with or exceed their expectations, reinforcing both revisit intentions and WOM as loyalty indicators [7]. Together, these factors form a cohesive loyalty framework, showing that higher satisfaction not only strengthens tourists' personal desire to return but also encourages them to endorse the destination to others.

The importance of understanding these behaviors is particularly critical for Indonesia, where tourism is one of the most profitable sectors due to the country’s rich cultural and geographical diversity. Indonesia is home to six UNESCO-listed cultural heritage sites, including Borobudur Temple, Prambanan Temple, and the Bali Subak System, which attract both domestic and international visitors [8]. Therefore, expanding it takes more than the development of the physical infrastucture. These sites rely on their ability to create an intriguing destination with a positive destination image with digitalized and up-to-date information [9].

In a previous study condcuted on a wildlife park in Indonesia, Faunaland Ancol, researchers found that social media has a significant and positive effect on promoting tourism [9]. Another study stated that the continuous uprising of social media significantly transforms tourism pre-travel stage, or the planning, as destination image perceptions was build in the consumers mind by their first impression online. They found that UGC positively and significantly impacts visiting intentions in Vietnam by enhancing destination image, visit intention, and tourist attitude [10]. Meanwhile, a study by Xu et al. [11] demonstrated that both factual and emotional UGC influence tourist loyalty through destination image and satisfaction, particularly at Gulangyu Island's cultural heritage sites in China. Destinations that drives on UGC plays a major role in fostering tourist loyalty that includes, but not limited to, revisit intention, satisfaction, and word of mouth (WOM).

While previous studies in other countries, have demonstrated the impact of UGC on destination image and loyalty behaviors, there is a notable geographical gap in understanding this relationship in Indonesia’s cultural heritage context.  Indonesia UNESCO-listed cultural heritage sites hold immense tourism potential, but their promotion through UGC remains underexplored in academic research. In addition to the geographical gap, previous studies have not fully explored the mediating role of destination image in explaining both types of UGC towards satisfaction, as well as the madiating role of satisfaction in connecting destination image with the three loyalty behaviors (revisit intention and WOM).

1.png

Figure 1. Conceptual model

Moreover, there is a population gap from the lack of previous research targeting Generation Z tourist, a demographic cohort that plays a pivotal role in the tourism industry and places a strong emphasis on authentic and emotionally engaging digital content. By incorporating both emotional and factual UGC, this study addresses these gaps, investigating how destination image and satisfaction mediates the relationship between UGC and tourist loyalty among Generation Z visitors to Indonesia's UNESCO cultural heritage sites. This focused approach contributes to a deeper understanding of UGC's role in cultural heritage tourism while providing insights relevant to Indonesia's tourism stakeholders.

Therefore the three main objectives for this study are (1) to determine whether factual and emotional UGC have a positive impact on destination image, (2) to explore whether destination image mediates the effect of factual and emotional UGC on revisit intention, satisfaction, and WOM, (3) to investigate whether satisfaction mediates the effect of destination image on revisit intention and WOM among Generation Z who have visited one of Indonesia's cultural heritage sites.

This research uses a quantitative and descriptive design with a cross-sectional design for data collection, meaning that data is collected at a single point in time without observing subjects over a prolonged period [16].

3.1 Research population and sample

The population of this study consists of Generation Z tourist. This demographic group was chosen due to being one of the fastest-growing travel segments, with significant purchasing power and an increasing interest in cultural and experiential tourism [17]. This demographic’s distinct reliance on social media platforms for travel-related content underscores the importance of understanding their behaviors, making them a relevant and critical population for examining UGC’s impact on cultural heritage tourism. A non-probability sampling method was applied, as not every member of the population has an equal chance of being selected as a sample [18]. The specific sampling technique is purposive sampling, where participants are chosen based on certain criteria relevant to the research objectives [16]. The sample in this study consists of Generation Z individuals aged 17-27 who have visited at least one cultural heritage site in Indonesia recognized by UNESCO.

Individuals aged 12–16 were excluded from the sample due to their limited cognitive and experiential capacity to fully comprehend and evaluate complex constructs central to this study, such as UGC, destination image, and tourist loyalty behavior. Prior research indicates that adolescents in this age range may lack the maturity or critical thinking skills needed to assess abstract concepts like perceived destination image and loyalty intentions [19].

Additionally, this age group is less likely to independently plan or engage in cultural heritage tourism compared to older individuals, making them less relevant to the study's objectives. Out of the 256 respondents who participated in the study, 200 met the study criteria and completed the questionnaire, which was distributed online via Google Forms. The sample size was determined based on the guidelines of Hair et al. [20], which recommend that the minimum sample size for Structural Equation Modeling (SEM) should be five times the number of indicators, while the maximum should be ten times the number of indicators. With 32 indicators used in this study, the minimum required sample size was 160 (32 × 5), and the maximum was 320 (32 × 10). Therefore, the sample size of 200 falls within this recommended range, ensuring adequate representation and statistical feasibility.

Although the sample size meets the thresholds recommended by Hair et al. [20], the adequacy of 200 respondents for SEM—given the complexity of the model—has been questioned in prior research. For instance, Wolf et al. [21] suggest that sample sizes in SEM should also consider model complexity, including the number of latent variables and pathways [11]. They emphasize that smaller sample sizes can reduce statistical power and increase the likelihood of estimation errors. Similarly, Muthén and Muthén [22] recommend sample sizes of 200 or more for simpler models but caution that more complex models often require larger samples to achieve robust results.

In this study, a sample size of 200 was chosen to balance practical limitations and statistical requirements. To enhance the robustness of the findings, bootstrapping and other statistical techniques were employed to account for potential limitations associated with sample size.

3.2 Research instruments and analysis

This research utilized 32 indicators to explain each variable through variable operationalization—analyzing values selected by the researcher to draw conclusions about the presence of these variables [18]. Responses were collected using a six-point ordinal scale, with categories as follows: (1) strongly disagree, (2) disagree, (3) somewhat disagree, (4) moderately agree, (5) agree, and (6) strongly agree. The use of six categories aimed to avoid neutral responses and improve reliability, as suggested by Simms et al. [23], who argued that increasing response options to six provides clearer distinctions in participant attitudes. Prior studies, such as Joshi et al. [24], have utilized six-point scales effectively, demonstrating their ability to enhance response clarity and reduce ambiguity in ordinal data collection. These scales have been shown to minimize indecision by encouraging respondents to take a definitive stance on the statements provided.

However, the choice of a six-point ordinal scale, while beneficial in some aspects, also introduces potential limitations that are not fully addressed in the study. One concern is the risk of central tendency bias, where participants may cluster their responses around the middle categories (e.g., somewhat disagree/moderately agree). Furthermore, the absence of a neutral option may compel respondents to make choices that do not fully align with their true perspectives, potentially affecting the validity of the results [25]. Although the scale design is intended to improve reliability, future research should consider these limitations and explore the use of alternative scales, such as seven-point scales, which provide a neutral midpoint while maintaining a balance between detail and clarity. Statistical analysis ensured the ordinal nature of the collected data was appropriately addressed to test the hypotheses proposed in this study.

This study utilized SmartPLS 4 for data processing and to test research instruments, employing Partial Least Squares (PLS) as an alternative to covariance-based SEM for models with latent variables. The analysis includes outer model testing for validity and reliability, and inner model analysis. Convergent and discriminant validity were assessed to ensure the model's robustness.

Hypothesis testing (path coefficient) was conducted to assess the t-statistics and p-values, which were used to determine whether the hypotheses should be accepted or rejected. The threshold for the t-statistics was set at greater than 1.96, and the threshold for the p-values was set at less than 0.05. The results of the hypothesis testing for each hypothesis are shown in Table 4.

Table 4. Path coefficient results

Source: Primary Data Proccessing on SmartPLS, 2024

5.1 Hypothesis 1

Hypothesis 1 (H1) suggested that factual UGC has a positive impact on destination image. However, the result showed an original sample value of -0.015, a t- statistics value of 0.181, and a p-value of 0.857. Since the p-value is greater than 0.05, this hypothesis is rejected, meaning factual UGC have a negative and no significant effect on destination image.

This finding aligns with research by Velayuthan and Hashim [28], which suggests that factual information does not always substantially influence the perception of a destination in some contexts. Specifically, for this study focusing on Generation Z, tourists appear to prioritize emotional narratives and firsthand experiences over factual content. Popsa [29] supports this by noting that Generation Z tends to rely more on emotionally engaging content for decision-making.

Additionally, the discrepancy in findings compared to earlier studies may be attributed to the distinct respondent profiles. Factors such as geography, with Indonesian tourists potentially valuing different content types than international tourists, and the inclusion of younger respondents aged 17-21, which were not often considered in prior research, could play significant roles. This highlights that factual UGC may lack sufficient emotional resonance to strongly influence destination image among younger audiences, underscoring the importance of tailoring content strategies to demographic preferences.

5.2 Hypothesis 2

Hypothesis 2 (H2) proposed that emotional UGC has a positive effect on destination image. The results showed an original sample value of 0.504, a t-statistics value of 5.088, and a p-value of 0.000. Since the p-value is less than 0.05, this hypothesis is accepted, indicating that emotional UGC significantly and positively impacts destination image.

The study by Cheung et al. [5] supports this conclusion by stating that the emotional components of content affectively increase a destination's appeal. Additionally, emotional interaction can boost tourist interest and enhance the destination's image, according to the principle of consumer loyalty. As a result, organizations that oversee Indonesian cultural heritage sites can use genuine emotional material to improve the emotional connection between visitors and the location.

5.3 Hypothesis 3

For Hypothesis 3 (H3), which hypothesized that destination image positively influences revisit intention, the original sample value was 0.543, the t-statistics value was 6.175, and the p-value was 0.000. With the p-value below 0.05, this hypothesis is accepted, showing that a positive destination image significantly influences the intention to revisit a location.

According to these results, it is consistent with the research of Wang et al. [30], which shows that a favorable impression of the destination can boost the inclination to return. A favorable perception of the location can also increase a person's intention to return, according to the TPB. Therefore, by emphasizing each site's distinctive qualities, managers of Indonesia's cultural heritage sites can use the destination image to draw visitors back.

5.4 Hypothesis 4

Hypothesis 4 (H4) suggested that destination image has a positive effect on WOM. The result showed an original sample value of 0.181, a t-statistics value of 1.402, and a p-value of 0.161. Since the p-value is greater than 0.05, this hypothesis is accepted, meaning destination image has a positive influence WOM, although not significantly.

This result is consistent with the study of Nguyen [31], which discovered that by fostering WOM among travelers, the destination image is essential to boosting tourism. Nonetheless, a non-significant correlation between the destination picture and WOM was discovered within the framework of this investigation. As a result, the study's sample might not accurately reflect the population.

According to research by Lam et al. [32], WOM is not always influenced by the destination picture, especially its cognitive aspects. According to this study, a favorable impression of a place is less successful in creating WOM if visitors do not feel deeply connected to the place or are not very satisfied. Because Generation Z places a higher value on emotional experiences and relationships when sharing on social media, the destination image alone might not be enough to persuade travelers to share their experiences.

The non-significant relationship between destination image and WOM suggests that a positive image alone may not suffice to inspire active advocacy. WOM is typically driven by a combination of strong satisfaction and emotional engagement. For instance, visitors who feel deeply connected to a site through personal experiences are more likely to share their stories. This highlights the need for tourism stakeholders to pair a strong destination image with initiatives that enhance visitor satisfaction and emotional bonds.

5.5 Hypothesis 5

Hypothesis 5 (H5) proposed that destination image positively influences satisfaction. The result showed an original sample value of 0.598, a t-statistics value of 1.014, and a p-value of 0.000. Since the p-value is less than 0.05, this hypothesis is accepted, indicating that destination image has a significant positive impact on satisfaction.

The study by Yunita and Tjokrosaputro [33], which claims that a favorable impression of a place results in more pleasure, is supported by this data. The Expectancy Confirmation Theory (ECT), which contends that positive views of the place will increase visitor pleasure, also explains the theory [30]. Thus, in order to increase visitor happiness, Indonesian cultural heritage site administrators can introduce experiences that enhance the destination's reputation.

5.6 Hypothesis 6

Hypothesis 6 (H6) suggested that satisfaction positively influences revisit intention. The original sample value was 0.242, the t-statistics value was 2.778, and the p-value was 0.005. Since the p-value is less than 0.05, this hypothesis is accepted, confirming that satisfaction significantly affects the intention to revisit.

The results of earlier research by Joshi et al. [24], which demonstrated that pleasure affects revisit intention, support this. The ECT, especially its last stage of repurchase intention, also explains this notion. According to ECT, happiness from expectations being confirmed reinforces the intention to revisit. Therefore, in order to increase visitor satisfaction and encourage them to return, operators of Indonesia's cultural heritage sites might raise the quality of their services.

5.7 Hypothesis 7

Hypothesis 7 (H7) proposed that satisfaction positively affects WOM. The result showed an original sample value of 0.424, a t-statistics value of 4.671, and a p-value of 0.000. Since the p- p-value is well below 0.05, this hypothesis is also accepted, indicating that satisfaction has a significant positive effect on WOM.

The study by Dini et al. [34] supports this conclusion by demonstrating that a high degree of visitor pleasure can motivate them to suggest a place. The TPB, in especially the attitude toward the behavior component, supports these findings as well, showing that positive WOM is a result of activity driven by satisfaction. Therefore, in order to promote WOM, heritage site managers must focus on visitor pleasure (Table 5).

Table 5. Specific indirect effect results

Source: Primary Data Processing on SmartPLS, 2024

5.8 Hypothesis 8

Hypothesis 8 (H8) proposed that destination image mediates the effect of factual UGC on satisfaction. The results indicated an original sample value of -0.009, a t-statistics value of 1.178, and a p-value of 0.859. Since the p-value is greater than 0.05 and the t-statistics value is below the threshold of 1.96, this hypothesis was rejected, meaning that destination image does not mediate the relationship between factual UGC and satisfaction.

This finding highlights that factual UGC, while offering accurate and objective information about a destination, fails to evoke the emotional or experiential connections necessary to enhance satisfaction. Satisfaction is often influenced by emotional engagement and interactive experiences, as supported by Dini et al. [34], who found that factual UGC had limited impact on satisfaction when mediated by destination image, especially for destinations lacking strong emotional appeal.

The ECT can also be used to understand this pathway's insignificance. ECT states that when expectations and experiences match, contentment results. Expectations based on objective facts may be created by factual UGC, but dissonance arises if these expectations do not match the destination's emotional or sensory reality. Even when mediated by destination image, factual UGC does not significantly affect satisfaction. These findings imply that, particularly for younger or experience-driven visitor segments, emotional resonance and experiential substance are more important in determining pleasure.

5.9 Hypothesis 9

Meanwhile, Hypothesis 9 (H9) proposed that destination image mediates the effect of emotional UGC on satisfaction. The analysis showed an original sample value of 0.301, a t-statistics value of 4.288, and a p-value of 0.000. As the p-value is less than 0.05 and the t-statistics value exceeds 1.96, this hypothesis is accepted, suggesting that destination image significantly mediates the relationship between emotional UGC and satisfaction.

This result aligns with Dini et al. [34], which emphasizes how emotional content enhances the positive effects of destination image. The TPB further explains this, suggesting that emotional UGC fosters stronger emotional connections between travelers and destinations, which translates into greater satisfaction. For cultural heritage site managers in Indonesia, this finding underscores the importance of leveraging emotive content in their promotional strategies to create lasting impressions and meaningful engagement with tourists.

5.10 Hypothesis 10

Hypothesis 10 (H10) suggested that satisfaction mediates the effect of destination image on revisit intention. The results showed an original sample value of 0.145, a t-statistics value of 2.684, and a p-value of 0.007. With the p-value below 0.05 and the t-statistics value exceeding 1.96, this hypothesis is accepted, indicating that satisfaction plays a significant mediating role in the relationship between destination image and revisit intention.

This finding is consistent with Nguyen Viet et al. [35], who noted that high satisfaction is a strong driver of revisit intentions. TPB provides a theoretical foundation, illustrating that when tourists perceive a destination positively and are satisfied with their experience, it reinforces their behavioral intention to return. Consequently, managers of Indonesia's cultural heritage sites should focus on enhancing the destination image to cultivate satisfaction and, in turn, encourage repeat visits, which are critical for sustainable tourism.

5.11 Hypothesis 11

While Hypothesis 11 (H11) proposed that satisfaction mediates the effect of destination image on WOM. The analysis revealed an original sample value of 0.253, a t-statistics value of 4.366, and a p- p-value of 0.000. Since the p-value is well below 0.05 and the t-statistics value exceeds 1.96, this hypothesis is also accepted, confirming that satisfaction mediates the relationship between destination image and WOM.

The findings are supported by Jayarathne [36], who highlighted that satisfied tourists are more likely to share positive experiences through WOM. Satisfaction acts as an emotional driver, transforming a favorable destination image into actionable behavior, such as recommending the destination to others. For managers of Indonesian cultural heritage sites, this underscores the strategic importance of enhancing satisfaction to amplify the reputation of their sites through positive WOM, ultimately attracting more visitors.