Travel experiences documented in online reviews influence travelers’ travel intentions

Abstract

Online travel reviews (OTRs) have revolutionized the way travelers plan their trips. While previous studies have investigated the impact of OTR attributes on travel intention, little attention has been paid to the impact of travel experiences documented in OTRs. Moreover, prior research has primarily used net effect methods, overlooking the possibility that combinations of different travel experiences may jointly influence travel intention. Based on experiential marketing theory, this study investigates how sensory, affective, behavioral, and intellectual experiences jointly affect travel intention. We collected 9242 online reviews and analyzed them with negative binomial regression and fuzzy-set qualitative comparative analysis. The findings reveal that sensory, behavioral, affective, and intellectual experiences documented in OTRs form multiple combinations and jointly influence travel intention. Our study contributes to the literature by revealing that the combinations of different travel experiences can affect travel intention.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-25971-9.

Introduction

Online travel reviews (OTRs) on social media have significantly transformed the tourism industry. According to a report by American Express¹, 75% of tourists search on social media to gain inspiration for their next journey. Partly due to the intangible and experiential nature of tourism, tourists rely on OTRs for information to make travel-related decisions^2,3, such as selecting destinations and booking hotels. Since most online travel reviews are submitted by unpaid tourists, they are more credible than other forms of information, such as advertisements or commercial promotions⁴. Meanwhile, these reviews document tourists’ travel experiences, including their food experiences and behaviors while traveling^5,6, thereby providing a wealth of information for tourists’ decision-making. Due to these factors, OTRs have become a vital marketing channel for the tourism industry. Therefore, how OTRs affect the travel intentions of tourists has become an issue of considerable interest among both scholars and practitioners.

Existing studies have investigated the impacts of OTR attributes on travel intention. For example, review valence, which is an OTR attribute and provides information about the quality of the trip, has been found to influence tourists’ travel intentions⁷. The source credibility of OTRs, as an indicator of information reliability, could lower perceived risk and enhance travel intention⁸. The existing studies revealed that potential tourists rely on OTRs to obtain valuable information to reduce risk or uncertainty in their travel decision-making^3,9. However, since tourism is a hedonic experiential product¹⁰, tourists purchase for travel experience. Thus, tourists seek travel experience-related information to better understand and evaluate tourism products¹¹. Researchers found that potential tourists often search for and browse experiential content on travel-related websites to form expectations and even trigger travel impulses^12–14. OTRs represent a particularly rich source of such travel experiential content, as they are narratives of tourists’ travel experiences^15,16. There are various types of travel experiences recorded in OTRs, such as sensory experiences (e.g., descriptions of beautiful landscapes) or behavioral experiences (e.g., descriptions of hiking in the rainforest). These experiences can enable potential travelers to formulate pre-experiences¹⁷. However, previous studies did not systematically investigate how different types of travel experiences documented in OTRs affect travel intention, limiting theoretical understanding regarding the impact of OTRs on potential tourists and constraining practical insights into how OTRs generate business value for tourism stakeholders.

Moreover, previous studies have investigated the impact of OTR attributes using net effect methods, such as structural equation modeling, which overlook the combinatory effects of multiple antecedents. These methods estimate the impact of a single antecedent on the dependent variable and have limited capability to examine how the combinations of a set of independent variables affect the dependent variable¹⁸. Moreover, scholars have suggested that combinations of different types of experiences may play a crucial role in driving positive customer outcomes^19,20. However, empirical evidence on the combination effects of experiences remains scarce, which hinders our understanding of their impact on tourists.

To address these two gaps, this study aims to investigate how travel experiences documented in OTRs jointly affect travel intentions. This study adopts the experiential marketing theory (EMT) to explain the impact of travel experiences in OTRs on travel intention. EMT conceptualizes experience as the subjective and internal responses individuals have through interactions with a product or company^21,22. There are four types of experience, including behavioral, sensory, intellectual, and affective experiences²³. The EMT suggests that the greater the richness of experiences, the more positively they can influence consumer attitudes and behaviors, such as satisfaction^22,23. Previous studies have applied EMT to explore the impact of experiences on consumer attitudes^24,25 and behavioral intentions^26–28. Although EMT was originally developed in the context of consumer goods, it has been extended to various products and services, including tourism^14,29. In tourism contexts, it has contributed to understanding intentions to revisit and consumer satisfaction^24,29,30. Therefore, this study adopts EMT as the theoretical framework to investigate how experiences jointly influence tourists’ travel intentions. Drawing on EMT, we identify four types of travel experiences documented in OTRs, including behavioral, sensory, intellectual, and affective experiences. To answer the research question, we collected 9242 online travel reviews and employed negative binomial regression (NBR) and fuzzy-set qualitative comparative analysis (fsQCA) to analyze these data. The results from NBR and fsQCA complement each other and reveal that the four types of travel experiences form complex configurations that jointly affect travel intentions.

Our study makes several important contributions. First, this study contributes to online travel review research by revealing that sensory, behavioral, affective, and intellectual experiences documented in OTRs can influence travel intention. Second, by combining NBR and fsQCA, this study contributes to the literature by revealing that the travel experiences documented in OTRs form multiple combinations and jointly influence travel intention, which responds to scholars’ calls to explore the dynamic impact of experiences on individuals¹⁹. To our knowledge, this is the first study to investigate the interdependency among these four experiences. Third, this study also contributes to experiential marketing theory by shifting the research focus from examining the effects of individual experiences in isolation to adopting a configurational perspective that explores the interdependencies among different types of experiences. Practically, our study also provides important implications. First, since our study reveals the interdependency among different types of travel experiences, travel social media platforms can leverage these insights to prioritize the distribution of content that combines specific experiences—such as sensory and intellectual—to enhance users’ travel intentions. Second, tourism stakeholders can leverage the insights from this study to design promotional content that incorporates effective experience combinations to attract potential tourists. Third, travel bloggers can apply these findings to refine their content strategies, thereby enhancing long-term commercial value.

Literature review and theoretical background

Online travel review and travel intention

Online travel reviews are a type of data that documents people’s attitudes, feelings, and emotions regarding travel¹⁶ and are particularly important for potential tourists. A significant body of research has focused on investigating the impact of OTRs’ attributes on travel intentions. We conducted a comprehensive search of the Web of Science to collect and summarize relevant studies, as shown in Table 1.

Table 1.

Literature review.

Attributes of OTRs		Ref.
Attributes of review content	Information quality, review valence, review trustworthiness, content style, electronic word of mouth, perceived credibility, perceived authenticity, photograph quality, and emotional review	3,4,9,31–34,42–44
Attributes of reviewers	Source credibility	8,31
Attributes of reviewed destinations	Destination familiarity, destination image	4,33

Existing literature has found that the attributes of OTRs’ content^3,31,32, reviewers^8,31, and reviewed destinations^4,33 contributed to tourists’ travel intention. Within this stream of research, previous studies primarily focused on the impact of OTRs’ content attributes on travel intention. Specifically, researchers suggested that tourists rely on OTRs to obtain reliable information that facilitates decision-making and shapes travel intentions³. For example, source credibility of OTRs, which signals more reliable information in OTRs, has been found to lower perceived risk and enhance travel intention⁸. The perceived authenticity of OTRs has been found to foster both cognitive and affective trust toward the platform, which in turn shapes travel intention³. Moreover, other studies suggested that the content quality of OTRs could also shape travel intention. For example, empirical evidence found that the information quality of OTRs was positively associated with travel intention, since it could facilitate trust toward the platform and destinations^32,34. Bigne et al.⁴ emphasized the role of detailed OTR content in reducing uncertainty and thereby increasing visit intention. Collectively, these existing studies revealed that OTRs play an important role in providing helpful information that helps tourists reduce risk or uncertainty in travel-related decision-making and, in turn, shape travel intention.

Although the existing literature has provided valuable insights into the impact of OTRs on travel intention, little is known about the role of travel experience documented in OTRs. Given that tourism is an experiential product¹⁰, tourists seek experience-related information to better understand and evaluate it¹¹. Previous studies suggested that potential tourists often searched online for experiential information^12–14, which could affect their travel intention. Specifically, in the era of social media, individuals frequently document their unique travel experiences in online reviews^15,16 and describe these experiences using experience-relevant terms, such as sensory-related words^35,36. As tourists seek travel experiences that stimulate their senses and minds^37,38, the experiences recorded in OTRs can provide valuable insights for potential tourists¹⁶. Previous studies have found that sensory tourism advertising and sensory information on tourism websites could generate vivid pre-experiences and facilitate travel intention^17,39. In the same vein, travel experience recorded in OTRs may also affect the travel intention of potential tourists. Although existing literature has made admirable efforts to investigate the impact of OTRs, as summarized in Table 1, no prior studies have examined the impact of OTRs from the experiential perspective, which hinders our understanding of the impact of OTRs.

Moreover, all previous studies mentioned in Table 1 have employed the net effects approach, such as structural equation modelling or ANOVA, while overlooking the combinatory effects of antecedents. Specifically, the net effect approach estimates the influence of “each hypothesized independent variable associated significantly with a dependent variable after separating out the influence of other independent variables“⁴⁰. In other words, the net effects methods assess the contribution of a single variable to the dependent variable¹⁸. Therefore, previous studies on OTRs using the net effects approach have examined the impact of each OTR attribute on travel intention in isolation, whereas ignoring the potential influence of combinations of OTR attributes. A recent study revealed that review attributes, such as writing style, evaluativeness, and review sources, have complex interdependent relationships and jointly affect individuals’ perceptions⁴¹. This finding suggests that individuals are more likely to form a comprehensive understanding of OTR attributes rather than considering them in isolation. Therefore, investigating the combined effects of travel experiences documented in OTRs on travel intention could provide deeper insights.

Driven by these two research gaps, this study aims to investigate how the travel experiences recorded in OTRs jointly affect travel intention. OTRs can significantly shape potential tourists’ perceptions and travel intentions by vividly depicting these multifaceted experiences. Therefore, this study seeks to explore how the travel experiences recorded in OTRs impact the travel intentions of potential tourists through the lens of experiential marketing.

Experiential marketing theory

In contrast to traditional marketing theory, which considers consumers as only interested in functionality and benefits, experiential marketing theory (EMT) views consumers as emotional, seeking enjoyable consumption experiences²¹. EMT posits that the consumption experiences consumers desire are “directed toward the pursuit of fantasies, feelings, and fun“^21,45. These experiences can be communicated through various channels that convey emotions and complex messages, such as electronic media²¹.

Scholars further divided experiences into four types, including behavioral, sensory, intellectual, and affective experiences²³. Specifically, behavioral experience refers to individuals’ perception of the actions or physical experience related to a specific stimulus^23,46, such as the perceptions about engaging in physical activities like hiking in the mountains or windsurfing in the sea. Sensory experience is an individual’s sensory perception due to smell, touch, hearing, taste, and sight related to a specific stimulus^23,47, such as seeing the beauty of a city’s architecture. Affective experience is an individual’s perception of the emotions related to a specific stimulus²¹, such as positive feelings when visiting a city. Intellectual experience refers to the experience of thinking or problem-solving^21,23, such as learning about local history in the museum. The EMT further proposes that since experience provides value, the greater the richness of experiences, the more positively they can affect consumer attitude and behavior^21,23. The theory has been applied to investigate the impact of experiences on consumers’ attitudes and behavioral intentions^24,25,28. For example, previous studies found that sensory and behavioral experience could positively affect the consumers’ satisfaction and intention to recommend²⁹. Moreover, the EMT has been applied to various kinds of products and services, including shopping and tourism^14,48,49.

Existing studies have investigated the impact of these experiences on individuals, but little attention has been paid to the interdependency of experience. Recent research emphasizes the need for a more dynamic perspective regarding the impact of experience on individuals¹⁹. Individuals holistically evaluate product-related stimuli and form behavioral intentions rather than evaluating them in isolation⁵⁰. This is because, although individuals may perceive discrete product-related stimuli, they interpret them collectively to construct meaning and guide their behavior⁵¹. Individuals expect certain product attributes to co-occur in coherent patterns⁵². For example, when purchasing clothing, some items are typically paired together to create a harmonious ensemble, whereas mismatched combinations are often avoided by individuals due to their lack of unity-in-variety. Researchers have found that matching environmental cues to create a stronger sense of unity can enhance consumers’ pleasure and purchasing behavior more effectively than mismatched stimuli⁵³. Additionally, scholars suggest that the combination of different types of experiences could play a crucial role in driving positive customer outcomes^19,20. Taken together, these studies imply that various experiential dimensions may interact and accumulate to jointly affect individuals. However, there is little empirical evidence examining how the interdependency among experiences affects individuals. Therefore, grounded in the EMT, this study introduces negative binomial regression and fsQCA to investigate how travel experiences documented in OTRs jointly affect travel intentions.

Hypothesis

Sensory experience, which includes the visual, olfactory, auditory, gustatory, and tactile experiences related to travel and conveyed by the descriptive content in OTRs^23,47, is hypothesized to positively influence travel intention. First, the enhancement of sensory experiences is associated with positive consumer outcomes^21,54. Previous research has shown that sensory advertisements in text can serve as an important source of information to evoke sensory imagery and appeal to consumers, thus positively affecting their purchase intentions⁵⁵. Therefore, when a review describes sensory-related information such as the “fresh mountain air” or the “vivid colors of autumn leaves,” it can evoke tourists’ sensory imagery, thereby enhancing travel intention. Second, the sensory experiences described in OTRs allow tourists to mentally pre-experience the travel destination and lead to a favorable attitude about the travel^39,56. Descriptions of sensory experience in advertising could increase the amount of mental imagery in consumers, thus promoting consumers’ credibility perception and purchase intention^39,57,58. Research has found that the richness of sensory stimulation in advertisements is positively associated with perceived credibility and purchase intention¹⁷. Finally, although sensory experiences encompass visual, auditory, olfactory, gustatory, and tactile dimensions, with visual experiences playing a dominant role in shaping tourists’ perceptions, the other sensory dimensions could contribute to long-lasting and memorable travel experiences⁵⁹. Previous studies found that all these sensory experiences described in advertisements could positively influence consumers’ attitudes toward products⁶⁰. Therefore, we hypothesize that:

H1

Sensory experience in OTRs could positively affect tourists’ travel intentions.

Affective experience, which is the emotions related to a travel and conveyed through OTRs, is hypothesized to positively influence travel intention. First, the EMT suggests that the richness of affective experiences could positively affect individuals’ attitudes and behaviors^21,23. Previous research has shown that affective experiences related to a destination can positively influence tourists’ satisfaction, which in turn affects their intention to recommend the destination²⁹. Second, given that tourists seek emotionally enriching experiences, the affective experiences documented in OTRs can provide valuable insights into the emotional rewards they might anticipate from their trip, thus influencing their travel intention^61–63. For example, reading an OTR like “I was filled with joy when I saw the northern lights” allows potential tourists to understand the anticipated emotional payoff from the trip, which can influence their travel intention. Therefore, we hypothesize that:

H2

Affective experience in OTRs could positively affect tourists’ travel intentions.

Behavioral experience, which is the actions or physical experience related to travel and conveyed through OTRs^21,23, is hypothesized to positively influence travel intention. First, since individuals seek the richness of behavioral experiences, the behavioral experiences documented in OTRs can provide cues for possible self-relevant actions, thereby increasing behavioral intention to replicate the experience^23,29. For instance, reading reviews that describe behavioral experiences, such as “hiking in the rainforest” or “dancing with locals during a traditional festival,” can evoke imagined bodily participation and trigger vicarious engagement in action-oriented activities^64,65, thereby increasing travel intention. Second, by providing insights into alternative lifestyles and ways of engaging with the destination, the behavioral experiences detailed in OTRs enhance the destination’s attractiveness and may further result in changes in behavior^66,67, such as travel intention. For example, the behavioral experience conveyed through Starbucks’ slogan (“If your coffee isn’t perfect, we’ll make it over”) provides customers with an actionable expectation that could influence their perceptions and responses^68,69. Hence, behavioral experiences described in OTRs could inspire tourists to consider ways of engaging with the destination and thus positively affect their travel intention. Therefore, we hypothesize that:

H3

Behavioral experience in OTRs could positively affect tourists’ travel intentions.

Intellectual experience, which is the experience of thinking or problem-solving and conveyed through OTRs^23,70, is hypothesized to positively influence travel intention. First, tourists seek intellectual experiences because thoughtful and reflective interactions stimulate their curiosity, enhance their involvement, and encourage deeper travel-related exploration^71,72. For example, reading reviews that contain intellectual experiences, such as learning a city’s historical evolution through its architectural styles, may spark curiosity and foster a desire for meaningful engagement, thereby increasing travel intention. Existing studies have found that intellectual experience could facilitate consumer engagement and behavioral intention, such as purchase intention^65,71. Second, intellectual experiences can fulfill tourists’ intrinsic needs for thinking and learning^73,74. Researchers found that reviews highlighting intellectual experiences, such as learning about local history in the museum or planning optimal travel routes, can activate reflective thinking and increase perceived value^24,75, which is a key driver for travel intention⁷⁶. Since tourists seek intellectual experiences, intellectual experiences described in OTRs can provide tourists with insights into experiences they may encounter and thus increase their travel intention. Therefore, we hypothesize that:

H4

Intellectual experience in OTRs could positively affect tourists’ travel intentions.

Data collection and measures

Data collection

This study collected OTRs from Mafengwo (www.mafengwo.cn), which was selected to collect research samples for several reasons. First, Mafengwo is one of the largest online travel communities in China, with over 100 million registered users⁷⁷. Second, Mafengwo functions as a travel social media platform, similar to TripAdvisor, where tourists can share their travel experiences. Third, unlike online travel agency platforms (e.g., Ctrip.com) or online-to-offline service platforms (e.g., Dianping.com), which mainly focus on booking services, Mafengwo emphasizes rich user-generated content regarding travel reviews. Mafengwo hosts a significantly larger volume of user-generated travel reviews than other general travel websites⁷⁸ and has been widely adopted as a data source in OTR-related studies^79–81. Based on these considerations, Mafengwo is particularly suitable as a source of research samples to investigate the impact of OTRs in this study.

When collecting review data, the ethics are fully considered. Since Mafengwo is a public social media for anyone to post their reviews, it is not required to seek permission from the users to gather and analyze the review data^82–84. Ethical review and approval were waived for this study due to the national law “Ethical Review of Life Sciences and Medical Research Involving Human Beings,” which exempts secondary data that does not contain personally identifiable information from the full review process. We collected data from the travel tips on Mafengwo. Specifically, in June 2023, we collected data from 2015 (when Mafengwo first released travel tips) to 2022. During the data collection, we did not restrict the destinations and collected all reviews within this time period. The website of Marengwo Travel Tips is shown in Fig. 1. We finally obtained 9,739 online travel reviews, which covered 34 provinces in China. The collected data were anonymized and combined prior to access and analysis. These reviews contain the number of favorites, the identity of the bloggers, and the content of the reviews, as shown in Fig. 2. We deleted reviews with blank texts to guarantee the reliability of the review data and results, leaving 9242 reviews.

Fig. 1.

The website layout of travel tips in Mafengwo.

Fig. 2.

The layout of online travel reviews in Mafengwo.

Variables operationalization

We introduced Linguistic Inquiry and Word Count (LIWC), which is a widely used tool for textual analysis, to measure key variables in our study. LIWC has been extensively validated and applied in prior research to measure variables such as emotion, relatedness, and cognitive responses through keywords from textual data^85–87. The independent variables were measured using established LIWC dictionaries, which have been validated in previous studies. Specifically, the textual analysis using LIWC measures various variables by comparing every word in the text to the dictionary, enabling researchers to gain a deeper understanding of the underlying factors within the text⁸⁸.

Since previous studies have used the sensory dictionary to measure sensory experience in text, we followed this approach and measured the richness of sensory experience (Sensory) in OTRs with the LIWC using the sensory dictionary^36,89. Since affective experience is the richness of emotions related to a specific stimulus, we followed previous research and measured the richness of affective experience (Affective) in OTRs using the emotion dictionary in LIWC⁹⁰. The emotion dictionary contains words that represent various types of emotion, such as words that represent positive emotion, like happy and hope, as well as words that represent negative emotion, like worry and sadness. We followed the definition of behavioral experience and used the motion dictionary in LIWC to measure the richness of behavioral experience (Behavioral). The motion dictionary contains words such as walk and jog, and has been used to measure physical and behavioral experience^91,92. Since intellectual experience involves thinking processes and problem-solving²³, we used the cognitive processes dictionary in LIWC to measure the richness of intellectual experience (Intellectual). The cognitive processes dictionary captures engagement with thoughts, analyses, and problem-solving and has been used to measure the level of active information processing and thinking^93–95. Table 2 presents examples of experience measurements.

Table 2.

Examples of OTRs with experience measurement.

OTR content	Sensory experience	Affective experience	Behavioral experience	Intellectual experience
In this restaurant, you can really taste an authentic American hamburger.	0.125	0.00	0.00	0.00
In the summer, you can ride a bike and see my favorite lotus flowers.	0.143	0.071	0.071	0.00
During the trip, my child can see animals and learn lots of knowledge about animals.	0.067	0.00	0.00	0.133

We used the number of favorites (Travelint) to measure travel intentions, as saving to favorites reflects potential tourists’ deliberate interest and signals their intention to collect travel-related information for potential future action^96,97. For the control variables, we controlled for the identity of the bloggers, as it may potentially affect the travel intention. For instance, some bloggers sell travel-related products on Mafengwo, and some bloggers are verified by Mafengwo. Their identity may potentially affect the saving to favorites behaviors. A series of dummy variables has been created to control potential bias from bloggers’ identities. Four types of identity markers were identified. If the blogger is an official organization (e.g., cultural and tourism bureaus), the Official is coded as 1 (otherwise, coded as 0). If the blogger is verified by Mafengwo, the Verified is coded as 1 (otherwise, coded as 0). If the blogger sells travel-relevant products, the Business is coded as 1 (otherwise, coded as 0). Moreover, the quantity of numerical data in the content (Numerical) is controlled, as features such as “ticket prices” or “distance” could enhance the perceived credibility of the reviews. The number of sentences (Length) is controlled, as the length of the review may influence the amount of detail provided. Finally, we controlled for the destination, as it may influence travel intention. Given that the reviews covered 34 provinces, we coded it as a categorical variable (Destination), assigning a unique value from 1 to 34 to represent each province. Moreover, review valence, which refers to the extent to which a review provides a positive or negative evaluation of the review objects⁹⁸, may also influence travel intention. Following the guidance of literature⁹⁸, we coded the review valence for all the collected reviews (1 = positive, 0 = neutral, −1 = negative). The interrater reliability exceeded 0.80, and all discrepancies were resolved through discussion.

Data analysis

NBR is suitable for estimating the contribution of a single independent variable to the dependent variable; however, it has limited capability to examine the effects of combinations of multiple variables on the dependent variable. By contrast, fsQCA could identify how combinations of multiple variables affect the dependent variable¹⁸. Additionally, fsQCA can explain relationships between variables even when these relationships are statistically insignificant⁹⁹. Therefore, the combination of these two methods could complement each other to overcome their limitations and offer deeper insights into the relationships between variables. In recent years, an increasing number of studies have combined these two methods to gain a more nuanced understanding of their research topics, particularly for investigating how the independent variables jointly affect dependent variables^100–102. Following previous literature, we employed both regression and fsQCA to investigate how experience documented in OTRs jointly affects travel intention. Specifically, we applied NBR and fsQCA to analyze the same data. First, we examined the research model using NBR, and second, we used fsQCA to conduct a further analysis of the same data.

Table 3 presents the descriptive statistics for all the variables. The Pearson correlation coefficient is presented in Table 4.

Table 3.

Descriptive statistics of variables.

	Mean	SD	Min	Max
Travelint	1889.423	4325.628	5	90,344
Sensory	181.868	133.061	3	2820
Affective	104.657	75.736	2	1276
Behavioral	172.728	132.755	1	2684
Intellectual	414.643	296.277	6	6334
Verified	0.403	0.491	0	1
Business	0.290	0.454	0	1
Official	0.011	0.106	0	1
Numerical	149.493	119.580	2	2171
Length	75.083	59.141	0	1411
Valance	0.698	0.476	−1	1

Table 4.

Pearson correlation coefficient matrix of main variables.

	1	2	3	4	5
Travelint	1.000
Sensory	0.206***	1.000
Affective	0.188***	0.859***	1.000
Behavioral	0.215***	0.810***	0.849***	1.000
Intellectual	0.214***	0.862***	0.914***	0.896***	1.000

Significance levels are indicated as follows: *** p < 0.001, ** p < 0.01, * p < 0.05.

Negative binomial regression analysis

The NBR model has been chosen for this study since the dependent variable (the number of favorites) is a count variable with overdispersion. Given that all of the variance inflation factors were less than 2, multicollinearity would not be a problem in our sample. Before running the model, all variables (except dummy variables) were centered.

The results of NBR are shown in Table 5. As shown in Table 5, sensory experience has a positive and significant impact on travel intention (β = 0.283, p < 0.001), which lends support to H1. Moreover, affective experience has a negative and significant impact on travel intention (β = −0.373, p < 0.001). Therefore, H2 is not supported. The behavioral experience does not have a significant impact on the travel intention (β = 0.117, p = 0.078), which does not support H3. Intellectual experience has a positive and significant impact on travel intention (β = 0.717, p < 0.001), which lends support to H4.

Table 5.

NBR regression results.

	Model 1	IRR	Model 2	IRR
Sensory			0.283***(0.077)	1.327
Affective			−0.373***(0.076)	0.688
Behavioral			0.117(0.078)	1.124
Intellectual			0.717***(0.101)	2.049
Verified	−0.452***(0.051)	0.636	−0.478***(0.049)	0.620
Business	−0.005(0.050)	0.995	−0.023(0.052)	0.977
Official	−1.873***(0.158)	0.153	−1.799***(0.149)	0.166
Numerical	0.331***(0.066)	1.392	0.450***(0.071)	1.567
Length	0.009***(0.0005)	1.009	0.009***(0.0005)	1.009
Valance	−0.169***(0.046)	0.845	−0.132***(0.045)	0.877
Destination	0.054***(0.002)	1.055	0.054***(0.002)	1.056
Constant	6.764***(0.193)	866.292	7.998***(0.461)	2976.257
Observations	9242		9242
LRa	1505.80***		1946.57***
Pseudo-R-Square	0.0210		0.0220
AIC	151112.2		150975.3
BIC	151176.4		151,068

The LR represents the likelihood ratio based on the chi-square distribution. Significance levels are indicated as follows: *** p < 0.001, ** p < 0.01, and * p < 0.05. Robust standard errors are reported in parentheses.

We have conducted several robust tests to check the reliability of our findings. First, we replaced the dependent variable measurement with the number of shares, which could also reflect users’ travel intentions (see Model 2 in Table A1 in Appendix A). Second, the measurement of affective experience was replaced with positive emotion terms and negative emotion terms (see Model 4 in Table A1). Finally, we use the Poisson regression to test for the validity of our findings (see Model 6 in Table A1). The results show that the findings of this study are robust.

Configuration analysis

The NBR results show that sensory, affective, and behavioral experiences significantly impact travel intention, while the effect of behavioral experiences is not significant. Since fsQCA can uncover relationships between variables even when they are statistically insignificant⁹⁹, it can complement NBR by offering a deeper understanding of the insignificant relationship. Additionally, as NBR has limited capability to assess the effects of multiple variable combinations, fsQCA provides further insights into the complementary effects between variables^103,104. Therefore, we subsequently conducted fsQCA to gain a more nuanced understanding.

FsQCA calibration

The dependent and independent variables were calibrated as fuzzy sets via the direct calibration method¹⁸. We followed previous studies¹⁰⁵ and built the fuzzy sets with three cutoff points: full membership (mean + one standard deviation), full non-membership (mean - one standard deviation), and the crossover point (mean). In line with NBR regression, the independent variables are the four experiences represented in the online travel reviews, and the dependent variable is the travel intention. Following previous studies¹⁰⁶, raw consistency and PRI consistency have been set to 0.8 and 0.5, respectively.

FsQCA results

A configuration refers to a specific combination of independent variables that affect dependent variables. Table 6 graphically displays the configurations for achieving high and low travel intentions. There is no necessary condition for high (low) levels of travel intention. All the circles are core conditions. The overall solution consistency as well as coverage for high (low) levels of travel intention are 0.77 (0.80) and 0.33 (0.19), respectively, which are appropriate values for both indicators.

Table 6.

The results of fsQCA.

Configuration	High Travel Intention			Low Travel Intention
	P1	P2	P3	N1
Sensory experience	⬤	⬤		⊗
Affective experience			⊗	⬤
Behavioral experience	⊗	⬤	⬤	⬤
Intellectual experience	⬤	⊗	⬤	⊗
Raw coverage	0.23	0.21	0.23	0.24
Unique coverage	0.06	0.03	0.05	0.24
Overall solution consistency	0.77			0.80
Overall solution coverage	0.33			0.19

The ⬤ means that the condition exists, whereas ⊗ refers to the condition is absent. The blank space in the table means the condition does not matter.

The raw coverage of all four configurations (as shown in Table 6) is relatively high, indicating that all these configurations have great empirical importance. Configuration P1-3 represents the combination of independent variables that facilitate a high level of travel intention, while configuration N1 represents the combination of independent variables that lead to a low level of travel intention.

Configuration N1 represents the combination of independent variables that lead to a low level of travel intention. Configuration P1 suggests that to enhance travel intention, OTRs should include a high level of sensory and intellectual experiences, while avoiding behavioral experiences. Configuration P2 indicates that to enhance travel intention, OTRs should include a high level of sensory and behavioral experiences and exclude intellectual experiences. Finally, configuration P3 suggests that to enhance travel intention, OTRs should include a high level of intellectual and behavioral experiences, and exclude sensory experiences.

The configurations P1, P3, and N1 suggest that describing behavioral experiences in OTRs is risky. This is because behavioral experiences, when combined with sensory or intellectual experiences, result in high travel intention (i.e., P1 and P3), while their combination with affective experiences can lead to low travel intention (i.e., N1). Meanwhile, sensory and intellectual experiences are important for facilitating high travel intention, as a high level of these experiences consistently leads to high travel intention.

Discussion

This study investigates how travel experience recorded in OTRs jointly affects travel intention. By combining NBR with fsQCA, our study has several interesting findings. Specifically, the NBR results show that sensory and intellectual experiences positively affect travel intention, which is consistent with prior research indicating that sensory and intellectual experiences can positively affect consumer behavior or purchase intention^71,107,108. Moreover, the results suggest that affective experiences have a negative effect on travel intention, which diverges from previous studies reporting that affective experience is positively associated with consumer satisfaction and behavioral intention^71,109. Further, our results also showed that behavioral experience does not significantly affect travel intention, which is not in line with previous studies revealing that behavioral experience is positively associated with consumer satisfaction and behavioral intention^23,71.

The results of fsQCA could further extend the findings of NBR. Specifically, providing a high level of sensory and intellectual experiences can facilitate travel intention. This result aligns with previous qualitative interview findings, which showed that the sensory experience enhanced intellectual experience, encouraged more reflection, and ultimately persuaded customers to purchase the products¹¹⁰. Moreover, the results showed that behavioral experiences in OTRs could facilitate high travel intention when combined with sensory or intellectual experiences. Previous research has found a similar pattern, suggesting that neither sensory nor behavioral experience alone is sufficient to facilitate perceived interactivity¹¹¹. However, their combination produces a synergistic effect whereby the two dimensions mutually reinforce each other, ultimately enhancing perceived interactivity¹¹¹. Meanwhile, as for the combination of behavioral and intellectual experiences, previous studies have found that intellectual experiences, such as learning knowledge, can stimulate behavioral experiences by enabling individuals to share this knowledge with others, thereby promoting favorable attitudes toward the products¹¹⁰. Finally, the fsQCA results suggest that behavioral experiences lead to low travel intention when combined with affective experiences. This provides a possible explanation for the overall insignificant impact of behavioral experience on travel intention in the NBR results, as its effects vary depending on the combination. Specifically, when behavioral experience is combined with sensory or intellectual experiences, it produces a synergistic effect that facilitates travel intention, whereas when combined with affective experiences, it produces a de-synergistic effect that reduces travel intention. Taken together, these findings reveal that four types of travel experience are related and jointly shape travel intention.

Theoretical implications

By employing the NBR and fsQCA, this study makes several important theoretical contributions. First, this study extends research on OTRs by investigating how travel experience documented in OTRs influences travel intentions. As tourism is a hedonic experiential product¹⁰, tourists search for travel experience-related information to understand and form expectations for the tourism product¹¹. While existing studies explored the impact of OTRs’ attributes on travel intention, no study has investigated how experiential content in OTRs shapes travel intention. Therefore, grounded in experiential marketing theory, this study complements existing research by exploring the impact of travel experiences in OTRs on travel intention. Our findings reveal that sensory, affective, behavioral, and intellectual experiences could affect travel intention, though their effects vary. Therefore, this study contributes to the literature by providing four important antecedents for travel intention through the lens of experiential marketing theory.

Second, this study contributes to the literature by revealing how the interdependency among experiences affects travel intention. Previous studies have examined the impact of OTRs’ attributes using net effect methods^3,112, while ignoring the combination of OTRs’ attributes that might jointly affect the travel intention. Although scholars have suggested that combining certain experiences could drive positive customer outcomes^20,50, there is limited empirical evidence on the complex interdependencies among different experiences and their influence on individuals. By combining NBR with fsQCA, our study explores how different experiences documented in OTRs work together to influence travel intention. Our findings suggest that combinations of sensory and intellectual experiences, sensory and behavioral experiences, and intellectual and behavioral experiences can lead to higher travel intention. To the best of our knowledge, this study is the first one to empirically investigate the interdependency relationship of experience. Therefore, this study contributes to the literature and responds to scholars’ calls for exploring the dynamic impact of experiences on individuals¹⁹.

Finally, this study advances experiential marketing theory by demonstrating the interdependent nature of travel experiences. While experiential marketing theory suggests that richer experiences enhance positive consumer attitudes and behaviors^21,23, our findings reveal that this is not always the case. In fact, certain combinations of travel experiences positively influence travel intention, whereas others may have negative effects. For example, combining sensory and behavioral experiences can positively influence travel intention, while combining affective and behavioral experiences can even lead to a reduction in travel intention. This advances EMT by shifting the research focus from examining the effects of individual experiences in isolation to adopting a configurational perspective that explores the interdependencies among different types of experiences.

Managerial implications

Besides the theoretical contributions, this study also has several important managerial implications. First, this study provides valuable insights for travel social media platforms. Our findings reveal the interdependency between travel experience and thus offer nuanced guidance for content distribution and algorithmic recommendations. For example, given that combinations of sensory and intellectual experiences could facilitate high travel intention, social media platforms (e.g., TripAdvisor, Mafengwo) should prioritize and algorithmically promote content that emphasizes sensory rich, intellectually stimulating information, such as descriptions of beautiful landscapes and historical knowledge-sharing.

Second, this study provides valuable guidance for tourism stakeholders such as destination marketing organizations and travel agencies. As OTRs have become a key marketing channel for the competitive tourism market, tourism stakeholders need to understand which OTR attributes could affect tourists. Our research demonstrates that experiences documented in OTRs can significantly influence tourists’ travel intentions. Therefore, tourism stakeholders can leverage these insights by collaborating with travel bloggers to promote their products and services. Specifically, tourism stakeholders could work with bloggers to highlight several combinations of experiences in their review content. For instance, based on our findings, they could combine sensory and intellectual experiences, sensory and behavioral experiences, and intellectual and behavioral experiences, since these combinations could facilitate tourists’ travel intention. Moreover, those tourism stakeholders can become bloggers and create review content that shares travel experiences based on the findings of our research to encourage tourists’ travel intention.

Third, this research provides valuable guidance for bloggers on travel social media platforms. Many of these bloggers collaborate with tourism-related brands or travel agencies and are eager to leverage their reviews to attract the travel intentions of potential tourists. Therefore, bloggers could leverage the findings of our study to elaborate on their reviews. For example, they could create reviews that combine sensory and intellectual experiences, sensory and behavioral experiences, and intellectual and behavioral experiences to increase tourists’ travel intention. Additionally, they should be cautious when combining behavioral experiences with other experiences, as certain combinations may lead to a decrease in travel intention. By doing so, bloggers can effectively transform their reviews into commercial opportunities, fostering long-term growth in their business value.

Limitations and future research

There are several limitations of this study, which bring opportunities for future studies. First, this study focuses on the impact of travel experiences documented in OTRs on travel intention. However, different types of destinations may elicit different types of experiences. For example, intellectual experiences may be more prominent in cultural or historical destinations. Therefore, future research could explore how different types of destinations affect travel experience and travel intention to provide context-specific insights for different types of destinations. Second, this study utilized real-world data collected from travel-related social media platforms to examine how travel experiences documented in online reviews influence travel intention. However, since all variables were measured from review texts rather than individuals’ subjective perceptions, this may limit the generalizability of our findings. Therefore, future research could adopt survey or experimental designs to directly measure individuals’ subjective perceptions and analyze the data using other analytical techniques such as confirmatory factor analysis (CFA) and structural equation modeling (SEM). Such approaches would allow the collection of additional influential factors and the incorporation of important control variables, such as demographic characteristics, to further validate and extend our findings. Third, the use of real-world online review data did not allow us to directly examine the specific mechanisms through which combinations of travel experiences influence travel intention. Future studies could investigate whether, and how, the synergistic effects among sensory, affective, intellectual, and behavioral experiences in OTRs shape travel intention. Such research would provide deeper insights into how the interplay of different travel experiences translates into individual decision-making. Forth, although this study has investigated the influence of four types of experiences on travel intention, other factors may also affect tourists’ travel intention. For example, the features of photographs and the branding effect of the destination may affect the intention to visit in different contexts. Future research could explore more contingency factors to further enhance the explanatory power of the model. Fifth, the data used in this study were collected from an online travel social media platform in China, which may constrain the external validity of our findings. Future research could extend this study by gathering data from diverse cultural contexts or cross-national samples to verify our findings across different populations and cultural settings.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

H.T. and D.Y. conducted the study, and D.Y. collected the data. H.T., H.Z. and D.Y. analyzed the results, H.Z. did the project management and visualization, and all authors wrote and reviewed the manuscript.

Data availability

Data are collected from the websites of Mafengwo (https://www.mafengwo.cn/gonglve/). The dataset for this study is available upon request from the first author.

Declarations

Competing interests

The authors declare no competing interests.