Social Media Experience as a Behavioral Driver in Culture-Based Music Tourism: A Comparative Study of Two Heritage Festivals in Indonesia

1. Introduction

Music tourism has rapidly emerged as one of the most dynamic and economically consequential niches of contemporary tourism. The global music tourism market, valued at USD 6.6 billion in 2023, is projected to reach USD 13.8 billion by 2032—a compound annual growth rate driven by post-pandemic event revival, the cultural turn in tourism consumption, and the digitalization of audience engagement (Custom Market Insights, 2023). Music festivals, in particular, have transcended their role as entertainment spectacles to become strategic instruments of cultural preservation, destination branding, regional identity construction, and economic multiplier generation (Mazlan et al., 2025; Onn et al., 2024; Peng, 2023; Skandalis et al., 2024).

Within this global landscape, Indonesia offers a particularly compelling empirical setting. Following the pandemic, domestic tourist movements rebounded to an estimated 1.02 billion trips in 2024 (BPS, 2025), with music festivals serving as significant travel motivators that intersect with the country's rich cultural heritage. Two festivals in Yogyakarta exemplify contrasting models of how music tourism integrates with cultural heritage: Prambanan Jazz Festival (PJF), staged within the UNESCO World Heritage Site of Prambanan Temple, and Ngayogjazz, distributed across the living kampung (urban village) spaces of Javanese community life. The former embodies tangible heritage tourism—curated, institutionally governed, and architecturally framed; the latter embodies intangible heritage tourism—organic, community-driven, and socially embedded. Together, they constitute a natural cross-contextual setting for examining how heritage type shapes experiential dynamics.

Despite growing scholarship on experience economy in tourism (Pine & Gilmore, 1999; Oh et al., 2007; Quadri-Felitti & Fiore, 2012) and an emerging stream of music tourism research (Gibson & Connell, 2005; Bolderman, 2020; Onn et al., 2024), three interrelated gaps remain. First, the classical 4E framework (entertainment, education, escapism, esthetics) was conceptualized in a pre-digital era and has not been systematically updated to account for the role of social media in extending the festival experience across time (Hudson et al., 2015; Bu et al., 2020). While scholars have begun to acknowledge social media's role in post-event engagement, it has rarely been operationalized as a structural experience dimension on equal footing with the original 4E. Second, empirical investigations of music tourism in Southeast Asian emerging-market contexts—where heritage, religion, and community structures shape experiential consumption in distinctive ways—are scarce relative to the dominance of European and North American studies in the field. Third, comparative studies examining how the type of cultural heritage (tangible vs. intangible) configures experience economy dimensions are virtually absent from the literature, despite UNESCO's longstanding distinction between these two heritage modalities.

This study addresses these gaps by formulating and empirically testing a 4E+1E model in which Extended Experience through Social Media (EESCM) is treated as a fifth, structurally co-equal experience economy dimension. Using SEM-PLS analysis on data from two contrasting heritage festivals in Yogyakarta, Indonesia, we examine how each dimension—classical and digital—influences revisit intention and word of mouth, and we systematically compare the experiential configurations across the two contexts.

1.1 Research Contributions

This study makes four contributions to the experience economy and music tourism literatures. First, theoretically, it advances the experience economy framework by empirically establishing Extended Experience through Social Media (EESCM) as a structurally co-equal—and behaviorally dominant—fifth dimension, moving beyond Pine and Gilmore's (1999) original four realms. Second, it introduces and validates a novel bifurcation pattern within the experience economy: a separation between behavioral triggers (entertainment and EESCM) and meaning-making layers (education, escapism, esthetics), which has implications for both theory and festival design. Third, methodologically and contextually, it provides one of the first comparative SEM-PLS analyses of music tourism across tangible and intangible heritage settings, demonstrating that heritage type functions as a configuration moderator—shaping the qualitative character and mechanisms of experience dimensions while leaving the structural pattern of behavioral significance largely intact. Fourth, empirically, it contributes much-needed evidence on culture-based music tourism in Indonesia, expanding the geographic and cultural scope of experience economy research.

2. Literature Review and Hypotheses Development

2.1 Music Tourism in Culture-Based Destinations

Music tourism—travel motivated wholly or partly by music-related events, performances, and experiences—has been increasingly recognized as a growing niche of contemporary tourism (Gibson & Connell, 2005; Bolderman, 2020). Its rise reflects broader shifts in tourist behavior toward experiential consumption, authenticity-seeking, identity construction, and meaning-making through travel (Richards, 2020; Skandalis et al., 2024). Within this niche, music festivals occupy a particularly important position: they aggregate large audiences, generate intense affective engagement, and provide rich settings for cultural and commercial exchange (Wood & Kinnunen, 2020).

Culture-based destinations introduce additional complexity. When music festivals are situated within sites of heritage significance, the experiential value of the event is intertwined with the symbolic, historical, and communal meaning of the place. Empirical research has shown that such contexts enhance the perceived authenticity of tourist experiences (Wang, 1999; Lee et al., 2024), strengthen place attachment (Iversen, 2022), and contribute to durable behavioral loyalty. Yet the precise mechanisms by which heritage contexts amplify, transform, or differentiate experience economy dimensions—and their downstream behavioral effects—remain underspecified. Moreover, the literature has not adequately distinguished between music tourism situated in tangible heritage settings (e.g., monumental sites, archaeological landscapes) and in intangible heritage settings (e.g., living community practices, oral traditions, ritual spaces), even though UNESCO conventions have long recognized these as analytically distinct.

2.2 Experience Economy: The 4E Framework and Its Limits

Pine and Gilmore's (1999) experience economy framework remains a foundational paradigm for understanding consumer engagement in service and tourism settings. The framework conceptualizes experience along two axes—participation (active vs. passive) and connection (absorption vs. immersion)—yielding four experiential realms: entertainment (passive absorption), education (active absorption), escapism (active immersion), and esthetics (passive immersion). Empirical applications have widely confirmed the explanatory power of this 4E model in cultural tourism (Quadri-Felitti & Fiore, 2012; Manthiou et al., 2014), festival research (Rivera et al., 2015; Aşan et al., 2020), and heritage tourism (Li et al., 2022; Culha, 2020).

However, the 4E framework was theorized in a pre-social-media era, and its capacity to account for contemporary, digitally-mediated tourist experiences is increasingly questioned. Festival experiences today are no longer confined to the on-site moment of consumption; they begin with pre-event anticipation generated through social media, intensify on-site through real-time content production and sharing, and continue long after the event through reflection, content recirculation, and community memory (Hudson et al., 2015; Ibrahim et al., 2021). This temporal stretching and digital embedding of experience has prompted scholars to call for expanded experience frameworks (Bu et al., 2020; Wood & Kinnunen, 2020), but few studies have yet formalized social media as a fifth structural dimension and tested its relative behavioral influence vis-à-vis the original four.

2.3 Extending the Framework: Extended Experience through Social Media (+1E)

In this study, we introduce Extended Experience through Social Media (EESCM) as a fifth experience economy dimension positioned alongside, not subordinate to, the original 4E. EESCM is conceptualized as the digitally-mediated continuation, reproduction, and amplification of the festival experience across the temporal arc of pre-event anticipation, on-site engagement, and post-event reflection. Drawing on Hudson et al. (2015), Bu et al. (2020), Wood and Kinnunen (2020), and Ibrahim et al. (2021), EESCM encompasses four sub-mechanisms: (a) digital engagement—active interaction with festival-related content and accounts; (b) emotional reinforcement—reliving and intensifying experiential memory through digital content; (c) social co-creation—producing and sharing content that contributes to a collective festival narrative; and (d) behavioral extension—translating digital interaction into offline action.

The theoretical case for treating EESCM as a structural experience dimension—not merely a marketing channel—rests on three observations. First, in contemporary festivals, the digital experience precedes, parallels, and outlives the on-site experience, structurally altering the temporal architecture of consumption. Second, social media does not merely transmit experience; it actively reconstructs it through user-generated content, peer interaction, and algorithmic amplification (Skandalis et al., 2024). Third, the affective and behavioral consequences of digital experience extension—particularly word of mouth and revisit intention—have been shown to exceed those of several classical 4E dimensions in some festival contexts (Tan & Lin, 2021). Together, these observations justify EESCM's elevation from supplementary variable to structural co-equal.

2.4 Behavioral Consequences: Revisit Intention and Word of Mouth

Two behavioral consequences anchor our model. Revisit intention represents the retentive tendency—the propensity to return and repeat an experience—and has been linked to satisfaction, emotional attachment, and perceived value in festival contexts (Khoo, 2022; Molina-Gómez et al., 2021). Word of mouth (WOM) represents the expressive tendency—the willingness to recommend an experience to others—and functions as a critical driver of destination reputation, new visitor acquisition, and digitally-mediated brand reach (Tan & Lin, 2021; Chou et al., 2025). Together, RI and WOM capture both the intrinsic (return) and extrinsic (advocacy) dimensions of festival loyalty, making them appropriate dependent variables for testing the behavioral consequences of experience economy dimensions.

2.5 Hypotheses Development

Entertainment, the core of festival consumption, generates emotional responses—joy, nostalgia, collective energy—that are easily articulated and shared. Consistent with Rivera et al. (2015) and Tan and Lin (2021), entertainment is expected to positively influence both behavioral outcomes.

H1: Entertainment Experience positively influences Revisit Intention.

H2: Entertainment Experience positively influences Word of Mouth.

Education experience, encompassing cultural learning and heritage interpretation, enriches the meaningfulness of festival attendance. Although its behavioral effect may be less direct than entertainment, it contributes to cognitive value capable of motivating return visits and recommendations (Li et al., 2022; Armbrecht, 2021).

H3: Education Experience positively influences Revisit Intention.

H4: Education Experience positively influences Word of Mouth.

Escapism, the experience of active immersion that breaks routine, fosters strong emotional memories and psychological recovery. Although its behavioral effects may be situational (Baldick & Jang, 2025), the construct is hypothesized to predict both outcomes.

H5: Escapism Experience positively influences Revisit Intention.

H6: Escapism Experience positively influences Word of Mouth.

Esthetics—the beauty of setting, design, and ambience—creates place-based memories and sensory richness that differentiate culture-based festivals from generic events (Iversen, 2022; Chou et al., 2025). Both behavioral outcomes are hypothesized to be influenced by esthetics.

H7: Esthetics Experience positively influences Revisit Intention.

H8: Esthetics Experience positively influences Word of Mouth.

EESCM is hypothesized to influence both behavioral outcomes through its capacity to amplify, sustain, and transmit experience across the temporal arc of the festival. Drawing on Hudson et al. (2015), Bu et al. (2020), and Ibrahim et al. (2021), EESCM is expected to be particularly powerful for word of mouth (through social amplification) and for revisit intention (through sustained emotional attachment).

H9: Extended Experience through Social Media positively influences Revisit Intention.

H10: Extended Experience through Social Media positively influences Word of Mouth.

2.6 Conceptual Framework

Figure 1 presents the conceptual model. Five experience economy dimensions—four classical (entertainment, education, escapism, esthetics) and one newly proposed (extended experience through social media)—are positioned as exogenous constructs predicting two behavioral outcomes: revisit intention and word of mouth. The model is tested separately in two culturally distinct music festival contexts, enabling cross-contextual comparison of both structural significance patterns and underlying mechanisms.

Figure 1: Conceptual Model — The 4E+1E Framework Linking Experience Economy Dimensions to Behavioral Consequences

Note: Arrows indicate hypothesized positive relationships (H1–H10). The +1E (EESCM) is presented as a structural co-equal to the classical 4E, not as a supplementary or moderator construct. The model is tested in two festival contexts (PJF and Ngayogjazz) to enable cross-contextual comparison.

3. Method

3.1 Research Design and Study Sites

This study employs a quantitative cross-sectional design using Structural Equation Modeling–Partial Least Squares (SEM-PLS). The quantitative analysis presented here forms a standalone empirical investigation, while drawing interpretive depth from a broader concurrent embedded mixed-methods project (Creswell & Plano Clark, 2018) of which it is part. SEM-PLS was chosen over covariance-based SEM (CB-SEM) for three reasons: (a) the model is predictive and exploratory in extending the 4E framework with a new dimension; (b) PLS handles smaller and unevenly-sized samples more robustly, which is critical given the contrasting respondent pools across the two festivals; and (c) PLS does not require multivariate normality assumptions, which is appropriate for Likert-scale festival survey data (Hair et al., 2019; 2022). Two research sites were purposively selected to represent contrasting heritage configurations within a single cultural-geographic region (Yogyakarta, Indonesia):

Prambanan Jazz Festival (PJF), an annual multi-day festival staged at the UNESCO World Heritage Site of Prambanan Temple. PJF represents the tangible-heritage tourism context: institutionally governed by a commercial promoter in coordination with the state heritage authority, with curated programming staged within a formally managed conservation zone. The 2025 edition recorded approximately 76,000 attendees over three days (Antara News, 2025; Injourney Destination, 2025).

Ngayogjazz, an annual one-day jazz festival distributed across the living spaces of a Javanese kampung. Ngayogjazz represents the intangible-heritage tourism context: community-driven, co-created across multiple makeshift stages embedded in everyday community life, with attendance free of charge and audience-performer boundaries deliberately blurred.

This contrast is theoretically motivated: PJF's tangible-heritage, institution-driven model and Ngayogjazz's intangible-heritage, community-driven model permit examination of whether—and how—heritage type configures experience economy dimensions and their behavioral consequences.

3.2 Sampling and Data Collection

Convenience sampling was employed at both sites, targeting visitors attending the 2025 festival editions. Data were collected through structured questionnaires administered on-site during the festivals and online via official festival community channels post-event, to maximize reach across visitor segments. After data screening (removing incomplete and inconsistent responses), 366 valid responses were retained for PJF and 98 for Ngayogjazz. Both samples meet the SEM-PLS 10-times rule minimum threshold (Hair et al., 2022), given that the model has five paths converging on each dependent variable (requiring n ≥ 50 per site). The sample size asymmetry reflects the substantially larger scale of PJF (~76,000 attendees) compared to Ngayogjazz (community-scale, intimate gathering) and is treated as a feature of contextual variation rather than a methodological limitation.

3.3 Measurement Instrument

A structured questionnaire measured seven latent constructs using a 5-point Likert scale (1 = strongly disagree; 5 = strongly agree). Indicators were adapted from validated instruments in the music tourism and experience economy literatures: entertainment (Rivera et al., 2015; Aşan et al., 2020), education (Li et al., 2022; Manthiou et al., 2014), escapism (Baldick & Jang, 2025; Molina-Gómez et al., 2021), esthetics (Iversen, 2022; Chou et al., 2025), and EESCM (Hudson et al., 2015; Ibrahim et al., 2021; Bu et al., 2020). Revisit intention and word of mouth items were adapted from Khoo (2022) and Tan and Lin (2021), respectively. Item wording was contextually adapted to the Indonesian music festival setting through expert review and a pilot test on 30 visitors prior to full deployment, leading to minor refinements for clarity.

3.4 Data Analysis

SEM-PLS analysis was conducted using SmartPLS 4. The analysis proceeded in two stages. First, the outer (measurement) model was evaluated for convergent validity (outer loadings > 0.70; AVE > 0.50), internal consistency reliability (Composite Reliability > 0.70; Cronbach's α > 0.70), and discriminant validity (Fornell-Larcker criterion and HTMT ratios < 0.85), following Hair et al. (2022). Second, the inner (structural) model was evaluated through path coefficients (β), R-square (R²) values, effect sizes (f²), and significance testing via bootstrapping with 5,000 resamples. To address potential common method bias arising from single-source self-report data, Harman's single-factor test was conducted, with the largest factor explaining less than 50% of total variance, suggesting that common method bias is unlikely to confound the findings (Podsakoff et al., 2003).

4. Results

4.1 Respondent Profile

Table 1 presents respondent characteristics for both festivals. The two samples differ substantially in profile, reflecting the contrasting nature of the festivals. PJF visitors are predominantly female (80.9%), largely first-time attendees (67.5%), and travel from across Java and beyond (59.6% non-local), reflecting an audience drawn by the prestige and novelty of the UNESCO heritage setting. Ngayogjazz visitors, by contrast, are predominantly female (61.2%), predominantly repeat visitors (41.8% attended four or more times), and heavily concentrated in the Yogyakarta region (72.4%), reflecting a loyal, community-rooted audience for whom the festival is part of an annual social tradition.

Table 1: Respondent Demographic and Visit Profile

Note: PJF = Prambanan Jazz Festival; NGJ = Ngayogjazz. Source: Primary data, 2025.

4.2 Measurement Model

All constructs in both datasets demonstrated satisfactory psychometric properties. Outer loadings for all retained items exceeded 0.70, indicating adequate item reliability. Average Variance Extracted (AVE) values were above 0.50 for all constructs, confirming convergent validity. Composite Reliability and Cronbach's Alpha exceeded 0.70 for all constructs, establishing internal consistency reliability. Discriminant validity was confirmed through both Fornell-Larcker criterion (square root of AVE for each construct exceeded its correlations with other constructs) and HTMT ratios below the 0.85 threshold (Henseler et al., 2015). These results confirm that the measurement model is valid and reliable across both festival contexts, establishing a sound basis for structural model interpretation.

4.3 Structural Model and Hypothesis Testing

Table 2 presents the structural model results for both festivals, including hypothesis decisions, significance levels, and R² values.

Table 2: Structural Model Results and Hypothesis Testing

Note: p<0.01; * p<0.001; † p<0.05 (two-tailed, bootstrapping 5,000 resamples). ENT=Entertainment Experience; EDU=Education Experience; ESC=Escapism Experience; EST=Esthetics Experience; EESCM=Extended Experience through Social Media; RI=Revisit Intention; WOM=Word of Mouth. H4 PJF (EDU→WOM): β=0.115, T=2.620, p=0.009 — statistically significant but with negligible effect size (f²=0.023); interpreted as a context-specific nuance rather than a general pattern. Source: SmartPLS 4 analysis, 2026.

The structural model reveals a striking and consistent pattern across both festival contexts. Of the ten hypotheses tested, nine show consistent results across both festivals. Entertainment (H1, H2) and EESCM (H9, H10) significantly predict both behavioral outcomes in both settings. Education, escapism, and esthetics do not predict revisit intention in either festival (H3, H5, H7 not supported); nor do they predict word of mouth in Ngayogjazz (H4, H6, H8 not supported). A context-specific nuance emerges for H4 in PJF: education exerts a statistically significant but negligible-effect influence on word of mouth (β=0.115, T=2.620, p=0.009, f²=0.023), an isolated finding that does not replicate in Ngayogjazz (β=−0.064, p=0.583) and thus does not constitute a general pattern. The dominant pattern remains a bifurcation: entertainment and EESCM function as behavioral triggers, while education, escapism, and esthetics operate as meaning-making layers without consistent direct behavioral impact. Effect size (f²) analysis reinforces this: the non-significant dimensions show very small effect sizes (f² < 0.025 across both festivals), while the supported paths range from small-to-moderate for entertainment (PJF: f²=0.017 for RI, 0.058 for WOM; NGJ: f²=0.100 for RI, 0.096 for WOM) to moderate-to-large for EESCM (PJF: f²=0.104 for RI, 0.305 for WOM; NGJ: f²=0.103 for RI, 0.251 for WOM).

EESCM emerges as the dominant predictor in both contexts, particularly for word of mouth. In PJF, the path EESCM→WOM yields the highest coefficient in the entire model (β=0.473, T=8.349, p<0.001), followed by EESCM→RI (β=0.347, T=5.434, p<0.001). In Ngayogjazz, EESCM similarly dominates: EESCM→WOM (β=0.439, T=4.215, p<0.001) and EESCM→RI (β=0.300, T=3.341, p=0.001). Entertainment, while consistently significant, shows comparatively smaller coefficients: ENT→RI at β=0.159 (T=2.247, p=0.025) in PJF and β=0.343 (T=2.892, p=0.004) in Ngayogjazz; ENT→WOM at β=0.234 (T=3.637, p<0.001) in PJF and β=0.315 (T=3.110, p=0.002) in Ngayogjazz. This is a striking empirical result: a dimension absent from Pine and Gilmore's original framework now exerts the strongest behavioral influence in contemporary culture-based music tourism, surpassing the classical entertainment effect in both heritage contexts.

The R² values indicate that the model explains a substantial proportion of variance in behavioral outcomes. For PJF: R² = 0.426 for revisit intention and R² = 0.637 for word of mouth. For Ngayogjazz: R² = 0.497 for revisit intention and R² = 0.559 for word of mouth. Notably, PJF shows a higher R² for WOM while Ngayogjazz shows a higher R² for RI—an asymmetry that aligns with the visitor profiles (first-time, spectacle-oriented audience at PJF; loyal, community-rooted audience at Ngayogjazz) and is discussed further in Section 5.

5. Discussion

5.1 The Bifurcation Finding: Behavioral Triggers vs. Meaning-Making Layers

The most theoretically consequential finding of this study is what we term the bifurcation pattern: of the five experience economy dimensions tested, only entertainment and EESCM exhibit statistically significant behavioral effects, while education, escapism, and esthetics consistently do not. Critically, this is not because the latter three dimensions are absent or unimportant in visitor experience. Companion qualitative evidence from the same research project (analyzed in the broader dissertation) confirms that visitors experience these dimensions richly and articulately. The puzzle is that experienced meaning does not translate into measurable behavior.

We propose that this divergence reflects a fundamental and previously underappreciated functional differentiation within the experience economy framework itself. Education, escapism, and esthetics operate as meaning-making layers—they enrich the cognitive, emotional, and symbolic depth of the festival experience but do so in ways that are internalized, implicit, and ambient rather than action-generating. Entertainment and EESCM, by contrast, function as behavioral triggers—they generate explicit, articulable, and digitally-shareable affective and social states that translate directly into the decision to recommend and return. This bifurcation has roots in the absorption-immersion-participation logic of Pine and Gilmore's original framework but moves beyond it by introducing a functional asymmetry: not all experience dimensions are equally action-generating, and the distinction between meaning-making and behavior-triggering may be more fundamental than the active-passive or absorption-immersion axes.

This finding aligns with and extends previous observations. Molina-Gómez et al. (2021) noted that escapism's behavioral effects are situational and personal; Iversen (2022) observed that aesthetic value is often internalized as ambient quality rather than salient decision input. Our results suggest that these are not isolated observations about specific dimensions but instances of a broader structural pattern that warrants explicit theoretical recognition.

5.2 The EESCM Dominance: A Necessary Extension to the 4E Framework

EESCM emerges as the dominant predictor of both behavioral outcomes in both festivals—a result that has significant implications for experience economy theory. While Pine and Gilmore's (1999) original framework positioned entertainment as the foundational realm and the locus of experiential value, our findings show that, in contemporary culture-based music tourism, a digitally-mediated experiential dimension exerts greater behavioral force than the classical entertainment realm itself. This is not merely an additive finding (i.e., that social media is also influential); it is a structural finding: EESCM cannot be relegated to a supplementary marketing channel or a moderator variable—it must be theorized as a structural co-equal to the 4E.

Three mechanisms underlie EESCM's dominance. First, temporal extension: by spanning pre-event anticipation, on-site engagement, and post-event reflection, EESCM accumulates and sustains experiential value over a much longer arc than the festival itself. Second, social amplification: digital content production is inherently social, embedding the festival experience within visitors' broader social networks and identities, which makes it more easily translated into word of mouth (Hudson et al., 2015; Tan & Lin, 2021). Third, memorial reconstruction: digital content allows visitors to relive, reframe, and intensify their festival memories, sustaining emotional attachment that drives revisit intention. Together, these mechanisms make EESCM particularly potent—especially for WOM, which is itself a fundamentally social-expressive behavior.

This finding necessitates a reconceptualization of the experience economy in the digital age. The 4E framework remains valid as a typology of experience modalities, but it is now empirically incomplete as a model of behavioral influence. We propose that the 4E+1E formulation, with EESCM positioned as a structural co-equal, is more appropriate for contemporary tourism research and practice.

5.3 Heritage Type as a Configuration Moderator

Although the structural significance pattern is consistent across both festivals—entertainment and EESCM dominate; education, escapism, and esthetics do not directly drive behavior—the qualitative character and underlying mechanisms of these dimensions differ systematically between the two contexts. Table 3 synthesizes this configuration.

Table 3: Cross-Festival Comparison of Experience Economy Configuration and Mechanisms

Note: PJF = Prambanan Jazz Festival (tangible heritage, institutional governance); NGJ = Ngayogjazz (intangible heritage, community governance). Source: Synthesis of quantitative findings and complementary qualitative interpretation, 2025–2026.

Three observations stand out. First, although EESCM is dominant in both contexts, its sub-mechanisms differ. In PJF, EESCM operates primarily through digital engagement and emotional reinforcement—fan content sharing and artist interaction—consistent with the technologically connected, novelty-oriented audience of a large commercial festival. In Ngayogjazz, EESCM operates primarily through behavioral extension and social co-creation—repeated attendance as social tradition, transmitted through community networks. Thus, EESCM is not monolithic; its internal sub-mechanisms are themselves context-contingent, shaped by the social capital structure of the destination community (bonding capital in Ngayogjazz; bridging capital in PJF, in Putnam's, 2000 terms).

Second, entertainment also shifts in character: it manifests as emotional engagement (nostalgia, spectacle, heritage-amplified affect) in PJF, but as social engagement (community intimacy, audience-performer fluidity) in Ngayogjazz. Although the structural effect of entertainment on behavior is consistent across both contexts, what entertainment means and how it functions is shaped by heritage type.

Third, the asymmetric R² pattern—higher WOM-R² in PJF (0.637) and higher RI-R² in Ngayogjazz (0.497)—reflects the audience-context fit. PJF's first-time, novelty-seeking audience is highly motivated to share their experience (high WOM-relevance), while Ngayogjazz's loyal, repeat-visitor audience has return decisions driven more cleanly by accumulated experience quality (high RI-relevance). This asymmetry strengthens the case that heritage type configures not only experience dimensions but also the dominant behavioral pathways through which they operate.

Taken together, these findings support our proposal that heritage type (tangible vs. intangible) functions as a configuration moderator in culture-based music tourism: it does not flip the structural significance pattern, but it systematically shapes the qualitative character and operational mechanisms of experience economy dimensions.

5.4 Theoretical Contributions

This study advances the experience economy and music tourism literatures in four specific ways. First, it extends Pine and Gilmore's (1999) 4E framework by empirically validating Extended Experience through Social Media (EESCM) as a fifth, structurally co-equal—and behaviorally dominant—experience dimension. To our knowledge, this is among the first empirical studies to formalize and test EESCM as a structural construct rather than as a moderator or supplementary marketing channel. Second, it introduces the bifurcation pattern as a theoretically novel refinement of the experience economy framework: experience dimensions cleave into behavioral triggers (entertainment, EESCM) and meaning-making layers (education, escapism, esthetics), suggesting that the framework needs a functional asymmetry layer in addition to its participation-immersion typology. Third, it advances heritage-tourism scholarship by introducing heritage type (tangible vs. intangible) as a configuration moderator that shapes the qualitative character and mechanisms of experience economy dimensions, even when the structural significance pattern is held constant. Fourth, it provides one of the first empirical investigations of music tourism in the Southeast Asian cultural context, broadening the geographic and cultural reach of experience economy research beyond its predominantly Euro-American empirical base.

5.5 Managerial Implications

For festival organizers and destination marketers, the findings suggest a clear strategic priority ordering. First, investments in digital-social ecosystem building—platform presence, artist-audience engagement, community channels, content amplification strategies, and pre-/post-event digital storytelling—yield the highest returns in behavioral outcomes. Festival marketers should not treat social media as a downstream promotion channel but as a structural component of the experiential product itself, integrated into the design of the festival across its full temporal arc. Second, investments in entertainment quality—programming, talent curation, stage production—remain essential as the second-most-influential behavioral driver. Third, while education, escapism, and esthetics do not directly drive behavior, they should not be deprioritized: they constitute the meaning-making substrate that gives entertainment and EESCM their depth and authenticity. Cutting cultural programming or heritage integration to focus solely on entertainment and digital marketing would impoverish the experiential foundation on which behavioral triggers act.

For policymakers and destination managers in heritage cities like Yogyakarta, the cross-contextual findings suggest a portfolio strategy: tangible-heritage festivals such as PJF are particularly effective for new visitor acquisition (high WOM-R²), while intangible-heritage festivals such as Ngayogjazz are effective for community embedment and loyalty cultivation (high RI-R²). The two models are complementary, not substitutable, and a destination strategy that nurtures both can serve different functions in the broader cultural tourism ecosystem. Heritage-based music tourism thus offers a model of how heritage preservation, community participation, and economic generation can co-evolve—provided that festival governance is sensitive to the distinctive logic of each heritage type.

6. Conclusion

This study contributes to the experience economy and music tourism literatures by advancing four interconnected findings. First, Extended Experience through Social Media (EESCM) operates as a structurally co-equal—and behaviorally dominant—fifth dimension of the experience economy, justifying its formal incorporation into a 4E+1E framework. Second, experience economy dimensions cleave into two functional categories: behavioral triggers (entertainment, EESCM) and meaning-making layers (education, escapism, esthetics), with implications for both theory and festival design. Third, heritage type (tangible vs. intangible) functions as a configuration moderator that shapes the qualitative character and operational mechanisms of experience dimensions, even when their structural significance pattern is consistent. Fourth, empirically, the study expands experience economy research into the under-studied context of Southeast Asian heritage tourism.

These contributions are grounded in SEM-PLS analysis of 366 visitors to Prambanan Jazz Festival and 98 visitors to Ngayogjazz—two music festivals in Yogyakarta, Indonesia, that represent contrasting tangible and intangible heritage configurations. The consistency of structural findings across these two highly different contexts strengthens confidence in the generalizability of the core results, while the contextual differences in mechanism enrich the theoretical interpretation.

Several limitations should be acknowledged. First, the cross-sectional design precludes causal inference about the temporal dynamics through which EESCM influences behavior. Future longitudinal designs could examine how digital engagement evolves across pre-/during-/post-event phases. Second, the asymmetric sample sizes (366 vs. 98) reflect the natural scale differences between the two festivals; while both meet SEM-PLS adequacy thresholds, future research with more balanced samples could enable stronger formal comparative testing through multi-group analysis. Third, the focus on two festivals in a single cultural region (Yogyakarta, Indonesia) constrains generalizability; extension to other heritage music festivals across Asia and globally would test the configuration-moderator role of heritage type more thoroughly. Fourth, the convenience sampling approach, while practical for festival contexts, limits population-level inference. Future studies could employ probabilistic sampling at festivals where attendee lists permit.

Future research could productively extend this work in several directions. First, EESCM's mediating role between meaning-making dimensions and behavioral outcomes deserves explicit investigation—does social media convert latent education/esthetics value into measurable behavior? Second, multi-group SEM analysis across more festival pairs could formally test the heritage-type-as-moderator proposition. Third, examining the four EESCM sub-mechanisms (digital engagement, emotional reinforcement, social co-creation, behavioral extension) as separate constructs could yield finer-grained theoretical insights. Finally, the bifurcation pattern itself invites cross-domain replication: does the behavioral-trigger/meaning-making-layer distinction hold in non-music, non-festival, or non-tourism experience economy contexts?

Author Contributions: Conceptualization, P.M.H; Methodology, P.M.H; Formal Analysis, P.M.H; Investigation, P.M.H; Data Curation, P.M.H; Writing – Original Draft Preparation, P.M.H; Writing – Review & Editing, E.R. & I.L.S. & J.S.L The author has read and agreed to the published version of the manuscript.

Funding: This research received no external funding.

Conflicts of Interest: The author declares no conflict of interest.

Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. The study was conducted in accordance with the ethical standards of Universitas Indonesia and the Declaration of Helsinki. Ethical approval was obtained prior to data collection.

Data Availability Statement: The anonymized dataset supporting the findings is available from the corresponding author upon reasonable request.

Acknowledgments: The author thanks the organizers of Prambanan Jazz Festival and Ngayogjazz, the visitors who participated in the survey, and colleagues at Universitas Indonesia for valuable feedback on earlier drafts.