Company websites and mobile apps versus social media: which service experience creates more customer value for online travel agencies?

Abstract

This paper examines the effects of online travel agency (OTA) service experience (via OTA company websites and mobile apps, and social media) on customer perceived (utilitarian and epistemic) value. The data of 404 usable responses from tourists with experience using OTAs revealed that service experience through company websites and mobile apps was only related to utilitarian value. However, service experience through social media influenced both utilitarian and epistemic value. Compared to epistemic value, utilitarian value had a larger effect on customers’ intention to book via OTAs. Political instability negatively moderated the relationship between utilitarian value and behavioral intention.

Introduction

Online travel agencies (OTAs), such as Expedia, TripAdvisor, Ctrip, and Traveloka, are aggregators that offer travel-related services to prospective travelers through websites or mobile applications (hereafter, “platforms”) (Talwar et al. 2020). The hospitality industry has become one of the fastest growing sectors in the travel industry. In 2019, the worldwide OTA market was valued at $744.7 billion, with 7.9% annual growth since 2015. However, tourism has been one of the sectors most affected by the COVID-19 pandemic: The global market for OTAs dropped suddenly from $744.7 billion in 2019 to $595.8 billion in 2020, a 20% decline (GlobeNewswire 2020). In many countries, travel restrictions severely disrupted the tourism industry, which constitutes a significant proportion of the economy (Reuters 2020). The pandemic has challenged OTAs to rethink their business models and plans (UNWTO 2020). Accordingly, it is essential to understand consumers’ current needs and reexamine their experience using OTA platforms.

Behavioral intention is a significant aspect of the study of consumer behavior. Researchers have used the context of OTAs to examine factors related to booking online, such as attitudes, price, quality, and perceived value (Talwar et al. 2020). Of these, perceived value is the most interesting element. Many have investigated the antecedent roles of customer value, such as functional and emotional value, in the tourism industry (Zhou and Yu 2022), including OTAs (Shi et al. 2022). In addition, we argue that because of its global popularity, social media may simultaneously influence the customer service experience through intention to book online (Kim et al. 2019). Hence, we differentiate between OTA platforms and social media when evaluating service experience. We explore the influence of service experience via OTA platforms and social media on two types of customer value: utilitarian value and epistemic value. Utilitarian value focuses on functional and monetary factors (Prebensen and Rosengren 2016) and has been widely studied. Epistemic value concentrates on novelty and how well a service arouses a customer’s curiosity (Chi et al. 2021); however, it is not adequately explored in the context of OTAs. Examining the impact of different types of customer value on the intention to book online via OTAs will provide information for OTA marketers to consider in strategic planning to create customer value.

In addition, social influence determines the impact of attitudes on behavioral intention rather than directly influencing behavior (Mouakket and Al-Hawari 2012). Previous studies have considered social influence an antecedent of behavior (Alzubaidi et al. 2021). We complement the literature by examining the moderating effects of social influence; that is, we argue that social influence may moderate the relationship among utilitarian value, epistemic value, and behavioral intention. Moreover, the macro-environment, such as political instability, is an exogenous factor that negatively affects the tourism industry. Prior studies have reviewed the negative impacts of political instability and terrorism on tourism destination image and tourist demand (Ingram et al. 2013). Accordingly, we view political instability as a key factor moderating the relationship among utilitarian value, epistemic value, and customers’ intention to book through OTAs.

We use the stimulus–organism–response (S–O–R) framework to investigate the effects of consumer perceived value on the intention to book online via OTAs as well as to examine the effects of service experience on perceived utilitarian value and epistemic value. We put forward four research questions: (1) Do OTA platforms and social media influence utilitarian value and epistemic value? In particular, which dimension of service experience (OTA platforms vs. social media) is more influential? (2) Are utilitarian value and epistemic value related to online behavioral intention? (3) Does the interaction between social influence and customer value impact behavioral intention? (4) Does the interaction between political instability and customer value affect behavioral intention? The results will provide valuable guidelines for OTAs that wish to pursue profitable design and improve performance.

This remainder of this paper is organized as follows. Section 2 presents the literature review and hypotheses development. Section 3 describes the methodology. Section 4 presents the data analysis and results. Finally, Sect. 5 discusses the contributions to theory, implications for management, limitations, and future research directions.

Literature review and hypotheses development

S–O–R theory

According to S–O–R theory, proposed by Mehrabian and Russell (1974), environmental stimuli (S) influence the cognitive evaluations of organisms (O), which eventually leads to a response (R). For example, attributes of websites (S) stimulate an individual’s internal state (Zhang et al. 2014). The organism (O) forms cognitive and emotional evaluations (Jiang et al. 2010). The response (R) reflects the outcome of the stimulus and organism in the form of a decision made regarding a given behavior (Jacoby 2002). S–O–R theory has been widely used in research on shopping behavior, such as technological and fashion-related factors of smartwatches (Cho et al. 2019), and online shopping, such as impulse buying in mobile commerce (Yang et al. 2021).

Service experience as environmental stimulus (S)

Service experience refers to the technical and psychological environment related to subjective personal reactions and feelings experienced by consumers (Roy et al. 2019) involving interactions between customers and companies and the resulting sentiments that illustrate customers’ involvement from an emotional and physical perspective (Kabadayi et al. 2019). We divide service experience into two dimensions: OTA platforms and social media. The first dimension, OTA platforms, involves the technological elements of a customer’s direct interaction with any form of technology during an encounter with a company (Stein and Ramaseshan 2016), such as searching, browsing, selecting, evaluating information, or transacting with the firm online. In a study of collaborative accommodation platforms, a positive experience with an information portal resulted in higher satisfaction and loyalty (Cristobal-Fransi et al. 2019). Specifically, elements of OTA platforms in the focal study include ratings and reviews, information on accommodations, visual appeal, and service bundling.

The second dimension, social media, refers to networks that contain a large amount of information about products and firms and that help extend interactive communication between customers and firms and between customers, including a wide range of online word of mouth (WOM) (Yawised et al. 2017). Marketers have embraced social media as a marketing medium, given the enormous potential audience. Every facet of modern life is becoming connected to social media, with 45% of the global population being active users of social media (Lambton-Howard et al. 2021). Social media directly affects travel movement, customer experience, selection, and behavior. Customers tend to purchase products or services when they are engaged on social media (Appel et al. 2020). Accordingly, advertising through social networks has become more popular as marketers have realized that traditional communication channels are less effective at stimulating sales (Voorveld 2019).

Customer value as organism (O)

Customer value is an individual’s evaluation of products or services, and it is processed when the individual interacts with environmental stimuli (Chang and Chen 2008). Previous studies have identified multiple dimensions of customer value. For example, Sheth et al. (1991) divided customer value into social, functional, emotional, epistemic, and conditional value, whereas Sweeney and Soutar (2001) grouped it into functional, emotional, and social value. Previous studies have considered utilitarian (functional) value (Babin et al. 1994; Prebensen and Rosengren 2016) and hedonic (emotional) value (Chiu et al. 2014). Functionality is the main aspect that drives customer perceived value of and satisfaction with a service. Therefore, it is essential to consider the role of utilitarian value in influencing customers’ behavioral intention. For hedonic value, previous studies have focused on the joyfulness, fun, and relaxation associated with using applications such as social media (Ashraf et al. 2019) and with online shopping (Overby and Lee 2006). However, for OTA, instead of enjoyment, it is more practical to consider the epistemic value that represents the response to customers’ curiosity (Sheth and Sharma 1997). For online booking services, both usefulness and curiosity-driven factors are the main influencers. Hence, we focus on two types of value: utilitarian value and epistemic value. Table 1 summarizes the definitions of utilitarian value and epistemic value defined in previous research.

Table 1.

Definitions of utilitarian value and epistemic value

Customer value	Sources	Definition
Utilitarian value	Babin et al. (1994)	Consumer value acquired from functional dimensions such as convenience, economic and practical value, service speed, and moneysaving
	Chiu et al. (2014)	Value based on functionality including sub-dimensions of convenience, diversity of goods, extensive product knowledge, and financial savings
	Prebensen and Rosengren (2016)	Experience value derived from the elements of functional and monetary quality
Epistemic value	Sheth et al. (1991)	Crucial factor in customer experience that involves consumers’ curiosity and the desires to learn new things and gain new information
	Sheth and Sharma (1997)	Curiosity, novelty, and knowledge formed when using a new or innovative product
	Um and Yoon (2021)	The value obtained from evaluating novelty, uniqueness, and attractiveness

Aligning with Yang et al. (2021), we see utilitarian value as a cognitive reaction to stimuli. In this study, utilitarian value refers to consumers’ evaluation of the ability of service experience to satisfy their needs (Lo and Qu 2015). Similarly, Roy et al. (2019) described utilitarian value as productivity oriented. Service providers need to ensure that adequate utilitarian value is provided to customers before focusing on other aspects of development (Ashraf et al. 2019).

Epistemic value refers to the perceived utility gained from an alternative’s capacity to arouse curiosity, provide novelty, and satisfy a desire for knowledge seeking (Sheth et al. 1991) and the value derived from learning new ways of doing things (van den Berge et al. 2021). Interacting with a technological system (e.g., reading reviews and information and making bookings) assists an individual in gaining epistemic value, as technological features can enhance knowledge. However, epistemic value has not been widely discussed in the tourism and hospitality literature (Phau et al. 2014; Cranmer et al. 2020).

Intention to book online as response (R)

Individuals’ cognitive and affective reactions determine their response to environmental stimuli. Positive responses are reflected in approach behaviors, whereas negative ones are reflected in avoidance behaviors (Hsu et al. 2021). In this study, the response results in behavioral intention associated with customers’ revisiting and future purchases (Chi et al. 2021). In tourism, the behavioral intention to use information technology for travel has been associated with perceived usefulness, identifying the importance of utilitarian value (Jeng 2019). Thus, when customers book services online, they may consider the experience and value in their future intention. Furthermore, the epistemic value derived from customers’ experience positively affects satisfaction and the willingness to continue using online services (El-Adly and Eid 2016). Furthermore, technology-related studies of epistemic value have revealed its impact on behavioral intention (Um and Yoon 2021).

Service experience through OTA platforms

Service experience is a key strategic objective for achieving customer loyalty (Stein and Ramaseshan 2016; Li et al. 2021). Travel customers encounter many self-service technologies, such as online reservations and booking through OTAs. OTAs have been traditionally defined as travel intermediaries that book hotel rooms through their websites and receive a commission for rooms sold (Lee and Chung 2019). To develop strategies, OTAs need to determine the impact of experiential value (e.g., utilitarian and epistemic value) on behavioral intention. We separate service experience into two dimensions: OTA platforms and social media. We then identify four main elements of service experience through OTA platforms: ratings and reviews, information on accommodations, visual appeal, and service bundling.

Ratings and reviews

Customer reviews are critical to the purchase of experiential goods (e.g., destinations, hotels, restaurants) before consumption because they enable consumers to assess the quality of the products as related by experienced customers (Talwar et al. 2020). Ratings and reviews are tools sites provide that enable customers to exchange feedback, give comments, and inform others about their choices (Nam et al. 2020). The ratings and reviews element benefits customers by enabling them to choose the service that satisfies them and benefits hotel managers by improving customer loyalty (Ahani et al. 2019).

Information on accommodations

Information is the main factor in evaluating tourism websites (Talwar et al. 2020). The criteria for assessing information include adequacy, accuracy, and currentness. Information quality, including whether information is accurate, up to date, and concise, is a component of e-travel service quality for OTAs (Chen and Chang 2018). In addition, information on accommodations offered positively influences the perceived usefulness of these platforms, and this information should be appropriately designed and delivered by OTAs (Ye et al. 2016).

Visual appeal

Chopdar and Balakrishnan (2020) proposed that visual appeal matters when one is using an information system. Perceived visual appeal and formality (e.g., a clear design) impact the cognitive, affective, and conative responses of online customers more significantly than perceived esthetic appeal. Thus, a visually pleasant website design may attract more customers and influence impulse buying decisions (Tariq et al. 2019). An esthetically pleasing website may be considered superior to less visually appealing sites (Zheng et al. 2019).

Service bundling

Bundling strategies focuses on passing on product and service cost benefits on to consumers (Saini et al. 2019). Firms bundle services in various ways, such as by combining multiple components of services produced in house or partnering with other firms (Ranaweera and Karjaluoto 2017). Many top travel agencies offer package options with multiple travel-related services, including flights, hotels, car rentals, and activities (Kim et al. 2018). Customers who purchase bundled offerings realize monetary savings, which results in compelling reactions.

Service experience consists of service value that arises from tangible and intangible products or services during the customer experience (Tsou et al. 2019). This value influences customers’ evaluations of products while shopping. Thus, searching on the Internet or booking online may lead to utilitarian and epistemic value in the booking. When it comes to online shopping, a company has to design appropriate experiences that provide convenience, helpful information, visually attractive websites, and beneficial service packages to attract customers’ attention and encourage interaction with the company. We posit that when customers perceive an OTA’s services as relevant to their experience, they find the process beneficial, which helps them achieve their goal and desired outcomes. Therefore, the following hypothesis is proposed:

Hypothesis 1

Service experience through OTA platforms has a positive effect on utilitarian value.

Epistemic value has been widely studied in the research on product purchases (Mäntymäki and Salo 2015). However, the literature on epistemic value in tourism and hospitality is limited. When consumers search for information to satisfy their curiosity about a travel destination or prospective hotels, epistemic value provided by the destination tourism site or hotels is established (Ashton 2015). When an OTA service is perceived as relevant to customers’ service experience, the process will arouse customers’ curiosity and interest in trying something new, learning new ways of doing things, seeking novelty, and satisfying their desire for knowledge. Consequently, positive experiences with perceived novelty and arousal of curiosity (epistemic value) can lead to feelings of pleasure in Internet users (Talwar et al. 2020). Hence, the following hypothesis is posited:

Hypothesis 2

Service experience through OTA platforms has a positive effect on epistemic value.

Service experience through social media

Social media is a software-based digital technology through which users can send and receive content over online social networks (Kwahk and Kim 2017; Appel et al. 2020). The most popular social media platforms, Facebook, Instagram, and Twitter, are often used as digital marketing channels. Utilitarian consumers are concerned about finding applicable content that suits their purpose (Jung et al. 2018; Yang et al. 2021). High-quality content on brands’ social media pages increases the possibility of interaction with brand users, which leads to customer engagement, brand awareness, and future purchase intention. Utilitarian consumers concerned about upcoming purchases prefer to receive useful information by browsing social media (Ashraf et al. 2019). Furthermore, a utilitarian motivation to shop from retail Facebook pages influences customers’ purchase intention (Chen and Tsai 2020). Thus, we propose that an OTA’s presence on social media can help customers obtain beneficial information from posts. We thus posit the following hypothesis:

Hypothesis 3

Service experience through social media has a positive effect on utilitarian value.

Epistemic value drives customers’ online engagement (Zhang et al. 2020). Learning new things is one reason for followers to engage on social media (Appel et al. 2020). Photo or video content can provide new knowledge and satisfy customers’ curiosity. In addition, the novelty of the content or readers’ comments on the original content in a post can initiate online discussion. Thus, social media is an innovative tool that creates a variety of value for customers. We assume that social media can be a practical marketing tool for promoting OTAs via content designed by companies and comments/reviews created by others in the social network, which can arouse customers’ curiosity, provide novelty, and increase their knowledge. The following hypothesis is thus proposed:

Hypothesis 4

Service experience through social media has a positive effect on epistemic value.

Customer value and intention to book online

We define behavioral intention as the intention to revisit, purchase, or recommend (Lo and Qu 2015), in particular whether customers are likely to book through an OTA in the future. Previous studies have shown an association between utilitarian value from consumption experiences and positive behavioral intention, such as repeat purchases and continued use (Chen and Tsai 2020; Shi et al. 2022). In addition, the utilitarian features of online hotel booking, such as accessibility, effective information search, and comprehensive product and service information, increase customer loyalty (Jeon and Yoo 2021; Yang et al. 2021). For these reasons, perceived value positively affects customers’ behavioral intention. The following hypothesis is thus proposed:

Hypothesis 5

Utilitarian value has a positive effect on behavioral intention.

As epistemic value is positively linked to purchase intention in the tourism and hospitality industries in general, it is expected to be positively related to purchase intention in the OTA context (Talwar et al. 2020). Yang et al. (2021) proposed that the effect of epistemic value on Internet shopping intention can arise from curiosity and novelty along with the pursuit of knowledge. OTAs are innovative systems and services on the Internet. The desire to seek new knowledge, curiosity, and an interest in learning something new may influence customers’ behavior and their recommendations to others to try new technologies (Huang 2018). We therefore posit the following hypothesis:

Hypothesis 6

Epistemic value has a positive effect on behavioral intention.

The moderating role of social influence

Social influence describes the effect of the people most important to an individual on that individual’s perceptions, attitudes, and behavior related to using a system such as an OTA (Alzubaidi et al. 2021). As social influence affects behavioral change, we aim to understand the moderating effect of social influence from important people such as family, relatives, and friends on customers’ intention to book using OTAs. A positive attitude can enable one’s behavior only when significant others approve; it has little or no impact if the social context opposes the behavior (Eagly and Chaiken 1993). This statement implies that social influence has a significant moderating effect on the relationship between customer value and behavior (Yang et al. 2021). The social environment of family and friends may critically influence a person’s behavior, and social context may act as a key variable to determine the extensiveness of Internet use (Singh et al. 2020). As many studies have found that social influence positively influences usage intention, we propose the following hypotheses:

Hypothesis 7a

Social influence has a positive moderating effect on the relationship between utilitarian value and behavioral intention.

Hypothesis 7b

Social influence has a positive moderating effect on the relationship between epistemic value and behavioral intention.

The moderating role of political instability

Political instability occurs when a government is likely to collapse or when people’s basic needs are not being stably met (Sönmez 1998; Farmaki et al. 2019). Political instability may seriously impact a country’s image and the sales of travel companies (Seddighi et al. 2001; Ghalia et al. 2019) and may prevent customers from traveling because of insecurity or uncertainty. Hanon and Wang (2020) found that tourism demand is even more sensitive to political instability than to terrorism. Related studies have shown the impact of political instability on perceptions of destinations from both the travel agent and visitor perspectives (Perles-Ribes et al. 2019), on their perceptions of adopting e-commerce (Abualrob and Kang 2016), and on the tourism industry in a country (Ivanov et al. 2017). We argue that political instability impacts utilitarian and epistemic value and behavioral intention when politics are unstable where customers live. As a result, we investigate how domestic political unrest influences customers’ intention to book. The following hypotheses are posited:

Hypothesis 8a

Political instability has a negative moderating effect on the relationship between utilitarian value and behavioral intention.

Hypothesis 8b

Political instability has a negative moderating effect on the relationship between epistemic value and behavioral intention.

Figure 1 shows the proposed S–O–R model.

Fig. 1.

Conceptual research framework. OTA online travel agency

Methodology

Measures

We developed a structured questionnaire survey that included screening questions and questions on the variables and demographics. Items that measured relevant constructs used a typical 5-point Likert scale. The response options were 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, and 5 = strongly agree. Service experience through OTA platforms was modeled as a second-order construct and measured with four reflective indicators: (1) ratings and reviews, from Flanagin and Metzger (2013), Li (2017), and Sparks et al. (2013); (2) information on accommodations, from Herrero and San Martín (2012) and Gao and Bai (2014); (3) visual appeal, from Heijden (2003), De Marsico and Levialdi (2004), and Rose et al. (2012); and (4) service bundling, from Sheng et al. (2007) and Saini et al. (2019). Scales for service experience through social media were from Voorveld et al. (2018). Questions on utilitarian value were from Overby and Lee (2006) and Lo and Qu (2015), whereas questions on epistemic value were from Pihlström and Brush (2008) and Cheng et al. (2009). Measures of political instability were from Sönmez (1998) and Seddighi et al. (2001). Social influence scales were from Lu et al. (2005) and Lu (2014). Finally, behavioral intention scales were from Williams and Soutar (2009) and Lo and Qu (2015). All scales were originally developed in English in previous studies; therefore, we used the double translation protocol (Brislin 1976) to validate the measurement items. The original English version was translated into Thai by two bilingual professional translators. Then the two Thai versions were translated back into English by two other bilingual professional translators. Then two bilingual management professors reviewed each back-translated item to check if the original meaning was retained. After the two professors discussed the differences between the original and back-translated English versions, the items were revised accordingly. Finally, the pilot test was conducted using the revised questionnaires. There were 39 respondents and we adjusted the questionnaire based on their feedback. The English version of the questionnaire is presented in the Appendix.

Research design

We used an online survey to examine Thai customers’ service experiences booking through an OTA. Inclusion criteria included being of Thai nationality and having experience booking hotel accommodations online through any OTA platform. Then we launched the formal questionnaire via Google Forms using a snowball sampling technique. Snowball is a popular convenience sampling method that makes it easy to reach target respondents. With this technique, some groups of people are invited to participate in a survey usually by email or a social network. Each participant is asked to pass on the invitation to others who match the criteria (Heckathorn 2011). Generally, this study targets office workers with experience using an OTA service because of their adequate purchasing power and their information and communication technology (ICT) capability. First, we sent the link of the survey to 200 randomly selected employees who work in travel agencies in Thailand and asked them to forward the link to their friends who have experience booking via an OTA. Simultaneously, we posted the link to Facebook pages and groups related to OTA services to attract more respondents with experience using an OTA. There were no limitations in terms of age or region of residence. Ultimately, we collected 568 responses. After eliminating 160 responses from those without experience using OTAs and 4 responses with low-quality data, we had valid responses from 404 individuals. The valid response rate was 71.13%. Table 2 presents the sample profile. Most respondents were 20–29 years old (60.9%), followed by 30–39 (22.5%). In terms of frequency of visiting an OTA, the majority of respondents visited an OTA 1–5 times a month (60.6%), followed by less than once a month (26.5%).

Table 2.

Sample demographics

Item	Category	Frequency	Percentage rates (%)
Gender	Male	100	24.8
	Female	304	75.2
	Total	404	100
Age	Younger than 20	5	1.2
	20–29	246	60.9
	30–39	91	22.5
	40–49	35	8.7
	50 and older	27	6.7
	Total	404	100
Region of residence	Bangkok	244	60.4
	Central	84	20.8
	Northern	16	4
	Southern	16	4
	Eastern	17	4.2
	Western	5	1.2
	Northeastern	22	5.4
	Total	404	100
Occupation	Office worker	201	49.7
	Full-time student	60	14.8
	Self-employed	47	11.6
	Official	27	6.7
	Freelancer	26	6.4
	Professional	22	5.4
	Cabin crew	9	2.2
	Housewife	5	14.8
	Retired	3	0.7
	Unemployed	3	0.7
	Other	1	0.2
	Total	404	100
Frequency of visit to an Online Travel Agency (average/month)	Less than 1 time	107	26.5
	1–5 times	245	60.6
	6–10 times	28	6.9
	More than 10 times	24	5.9
	Total	404	100
Income per month (USD) | 1 USD = 32.83 THB	No income	17	4.2
	Less than $305	17	4.2
	$305–$610	58	14.3
	$611–$915	122	30.2
	$916–$1220	70	17.3
	$1221–$1525	35	8.7
	$1526 or more	85	21
	Total	404	100

Common method bias

Because the data came from a single source, the current study used Harman’s single-factor test to control for potential common method bias (Podsakoff et al. 2003). The test was conducted with all 41 items to identify the variance explained by a single factor. One factor explained 27.365% of the data, which is less than the 50% threshold, which indicates that common method bias was not a problem.

Data analyses and results

Measurement model

IBM SPSS 22.0 and IBM SPSS AMOS 21.0 were used for data analysis. Exploratory factor analysis and confirmatory factor analysis were used to assess convergent validity and discriminant validity. After the removal of seven items (RR4, SB1, SB2, SM6, UV4, UV6, and BI1), the overall model statistics satisfied the threshold values: chi-square/degrees of freedom = 1.63, goodness-of-fit index (GFI) = 0.87, incremental fit index (IFI) = 0.96, comparative fit index (CFI) = 0.96, and root mean square error of approximation (RMSEA) = 0.04 (Hu and Bentler 1999).

We tested the internal reliability of the measurement model using Cronbach’s alpha and composite reliability (CR). Cronbach’s alpha for all constructs exceeded the recommended level of 0.7, demonstrating high internal consistency (Nunnally and Bernstein 1994). CR was between 0.83 and 0.94, exceeding the recommended standard of 0.7. Convergent validity was validated, as the average variance extracted (AVE) ranged from 0.55 to 0.77, exceeding the threshold of 0.5. The square root of the AVE was greater than the inter-construct correlations, confirming high discriminant validity. Table 3 shows the results of first-order reliability and validity analyses, CR, AVE, and MSV. The overall model statistics were as follows: chi-square/degrees of freedom = 1.82, GFI = 0.85, IFI = 0.94, CFI = 0.94, and RMSEA = 0.04.

Table 3.

Results of first-order reliability and validity analyses, CR, AVE, and MSV

	CR	AVE	MSV	RR	IA	VA	SB	SME	UV	EV	SI	PI	BI
RR	0.888	0.726	0.479	0.852
IA	0.835	0.558	0.511	0.692**	0.747
VA	0.870	0.628	0.368	0.480**	0.607**	0.792
SB	0.910	0.776	0.239	0.138**	0.291**	0.371**	0.881
SME	0.879	0.593	0.254	0.307**	0.425**	0.504**	0.414**	0.770
UV	0.836	0.562	0.511	0.578**	0.715**	0.462**	0.185**	0.377**	0.750
EV	0.942	0.730	0.493	0.050**	0.126**	0.347**	0.454**	0.432**	0.211**	0.854
SI	0.847	0.581	0.493	− 0.036**	0.040**	0.353**	0.489**	0.432**	0.044**	0.702**	0.762
PI	0.914	0.729	0.156	0.230**	0.179**	0.198**	0.249**	0.318**	0.139**	0.332**	0.395**	0.854
BI	0.914	0.728	0.391	0.384**	0.477**	0.376**	0.235**	0.364**	0.625**	0.288**	0.271**	0.200**	0.853

Inter-correlation coefficients are below the diagonal, and square roots of AVE estimates are along the diagonal in bold

CR composite reliability, AVE average variance extracted, MSV maximum shared variance, RR ratings and reviews, IA information on accommodations, VA visual appeal, SB service bundling, SME social media engagement, UV utilitarian value, EV epistemic value, SI social influence, PI political instability, BI behavioral intention

**p < 0.01

Structural model

A structural equation model was used to test the hypotheses. Figure 2 shows standardized path coefficients and the overall model fit for the direct effects model (chi-square/degrees of freedom = 2.057, GFI = 0.87, IFI = 0.94, CFI = 0.94, RMSEA = 0.05). As shown in Table 4, service experience through OTA platforms had a significant positive effect on utilitarian value (β = 0.739, t = 5.349***), so H1 was supported. However, the path from service experience through OTA platforms to epistemic value was not significant (β = 0.038, t = 0.717, p > 0.05), so H2 was not supported. Service experience through social media had significant positive effects on both utilitarian value (β = 0.093, t = 2.021*) and epistemic value (β = 0.419, t = 7.638***). Hence, H3 and H4 were supported. Moreover, utilitarian value (β = 0.598, t = 11.054***) and epistemic value (β = 0.174, t = 3.869***) had significant positive effects on behavioral intention; thus, H5 and H6 were supported.

Fig. 2.

The structural model with moderating effects. Dashed lines indicate unsupported hypotheses. OTA online travel agency. ***p < 0.001; **p < 0.01; *p < 0.05

Table 4.

Results for direct effects

Hypothesis	Standardized path coefficient	t	Result
H1	0.739	5.349***	Supported
H2	0.038	0.717	Not supported
H3	0.093	2.021*	Supported
H4	0.419	7.638***	Supported
H5	0.598	11.054***	Supported
H6	0.174	3.869***	Supported

***p < 0.001; **p < 0.01; *p < 0.05

Moderating effects

As shown in Table 5 and Figs. 2, 3 and 4, the Social Influence × Utilitarian Value interaction had a significant negative moderating effect on the relationship between utilitarian value and behavioral intention (β =  − 0.119, t =  − 2.965**). Therefore, H7a was not supported, as we had hypothesized that the moderating effect would be in a positive direction. The Political Instability × Utilitarian Value interaction had a significant negative moderating effect on the relationship between utilitarian value and behavioral intention (β =  − 0.122, t =  − 2.977**), so H8a was supported. However, the Social Influence × Epistemic Value (β =  − 0.032, t =  − 0.684, p > 0.05) and Political Instability × Epistemic Value (β =  − 0.024, t =  − 0.473, p > 0.05) interactions had nonsignificant moderating effects on the relationship between epistemic value and behavioral intention. H7b and H8b were therefore not supported.

Table 5.

Results for moderating effects

Hypothesis	Path	Standardized path coefficient	t	Result
	UV → BI	0.538	13.504***
	SI → BI	0.218	5.483***
H7a	UV × SI → BI	− 0.119	− 2.965**	Not supported
	EV → BI	0.203	3.365***
	SI → BI	0.107	1.799
H7b	EV × SI → BI	− 0.032	− 0.648	Not supported
	UV → BI	0.528	12.800***
	PI → BI	0.122	3.002**
H8a	UV × PI → BI	− 0.122	− 2.977**	Supported
	EV → BI	0.235	4.702***
	PI → BI	0.120	2.306*
H8b	EV × PI → BI	− 0.024	− 0.473	Not supported

UV utilitarian value, EV epistemic value, SI social influence, PI political instability, BI behavioral intention

***p < 0.001; **p < 0.01; *p < 0.05

Fig. 3.

Plot diagram: The moderating effect of social influence on the relationship between utilitarian value and behavioral intention

Fig. 4.

Plot diagram: The moderating effect of political instability on the relationship between utilitarian value and behavioral intention

A comparison of the effects of OTA platforms and social media

As shown in Fig. 2 and Table 4, service experience through OTA platforms directly influenced utilitarian value, not epistemic value. This is possibly because the four elements of online attributes—ratings and reviews, information on accommodations, visual appeal, and service bundling—are standard features that customers expect when using an OTA website or application; therefore, they do not satisfy the customer’s curiosity to acquire new information. Meanwhile, service experience through social media significantly influenced both utilitarian value and epistemic value.

When we compared the impacts of the two service experience dimensions on utilitarian value, the empirical findings proved that the significant positive effect of OTA platforms (β = 0.739, t = 5.349***) was larger than that of social media (β = 0.093, t = 2.021*). This is in line with the theory that utilitarian value is based on functionality, which can be derived from the four elements of online attributes. In contrast, social media may effectively facilitate the use of OTA platforms, but its direct influence on utilitarian value is less crucial than that of OTA platforms.

Service experience through social media was significantly related to epistemic value in the positive direction (β = 0.419, t = 7.638***), which implies that the flow of information through social networks efficiently satisfies customers’ curiosity to obtain new information. The information spread through social media is abundant and diverse. Thus, it stimulates epistemic value related to the need for novelty. Figure 5 shows the significant relationships and effect size comparisons in the S–O–R model for OTA service experience.

Fig. 5.

Stimulus–organism–response model for OTA service experience. Solid lines indicate strong relationships, and dashed lines indicate weaker ones. OTA online travel agency

Discussion and implications

Research findings

Our proposed model determined the two dimensions of service experience of OTA users—OTA platforms and social media—to evaluate their impacts on utilitarian and epistemic value and future behavioral intention. Service experience through OTA platforms positively affected utilitarian value (H1 was supported). This result suggests that all four elements of OTA platforms play important roles in terms of utilitarian value for customers booking through OTAs. However, the online elements of OTA platforms had no impact on epistemic value (H2 was not supported). Possible reasons for this include customers’ inability to learn new skills from these services and their failure to arouse customers’ curiosity about trying OTA systems, leading to dissatisfaction in knowledge seeking. In contrast, service experience through social media positively affected both utilitarian and epistemic value (supporting H3 and H4, respectively), which indicates that content on OTAs’ social media is necessary for creating utilitarian and epistemic value. In terms of utilitarian value, content on OTAs’ social media may provide extensive information for customers, enabling them to find information online with minimal effort (Liu and Park 2015). The effect of social media engagement on epistemic value was more robust than that of utilitarian value. Social media content may actively stimulate customers to learn something new, arouse curiosity, and support knowledge acquisition. In addition, the results showed that utilitarian and epistemic value had positive effects on behavioral intention (H5 and H6 were supported), which confirms that obtaining high utilitarian and epistemic value from OTAs can encourage customers to continue booking through OTAs.

Social influence and political instability had negative moderating effects on the relationship between utilitarian value and behavioral intention. As social influence was hypothesized to act in the positive direction, H7a was not supported. The negative moderating effects of social influence indicate that the effect of the interaction term is weaker than the sum of single-variable effects. Thus, the effect of utilitarian value is more crucial than that of social influence when considering booking online via OTAs. One plausible reason for this is that because utilitarian-oriented customers probably expect to obtain benefits or appropriate services from companies, undesirable experiences or feedback from family or friends may diminish their intention to book with OTAs. As shown in Fig. 4 and Table 5, the gap between the low-and the high-social influence groups was smaller when utilitarian value was high rather than low. This result suggests that social influence dampened the effect of utilitarian value on behavioral intention even though it had a significant positive effect on behavioral intention (β = 0.218, t = 5.483***). The moderating effect of political instability on the relationship between utilitarian value and behavioral intention turned out to be significant in the negative direction, as hypothesized; therefore, H8a was supported. A possible reason for this is that customers with high utilitarian value are more likely to accomplish their goals; however, high political instability may reduce their sense of security and their intention to travel or book through OTAs to minimize risk. Because inconveniences may arise during violent political situations, the influence of utilitarian value on the intention to book online decreases when the risk of political instability increases. As shown in Fig. 5 and Table 5, the gap between the low- and high-political instability groups was smaller when utilitarian value was high rather than low. This result suggests that political instability weakened the effect of utilitarian value on behavioral intention even though it had a significant positive effect on behavioral intention (β = 0.122, t = 3.002**). However, neither social influence (β =  − 0.032, t =  − 0.648, p > 0.05) nor political instability (β =  − 0.024, t =  − 0.473, p > 0.05) had a moderating effect on the relationship between epistemic value and behavioral intention (H7b and H8b were not supported). A possible reason for this is that epistemic responses are typically described as feelings of authenticity, uniqueness, novelty, and curiosity (Prebensen and Rosengren 2016), so epistemic customers may not let others (social influence) or unusual conditions (including political instability) affect their intention to use a system.

Theoretical implications

The current study contributes to tourism and hospitality research using S–O–R theory in the OTA context. First, the research framework considered two dimensions of service experience—OTA platforms and social media—as stimuli (S). Comparing the two dimensions allowed us to understand the different antecedent effects on customer value. OTA platforms only relate to utilitarian value, whereas social media impacts both utilitarian and epistemic value. We identified four elements of direct OTA platforms: ratings and reviews, information on accommodations, visual appeal, and service bundling. Utilitarian customers would find these elements beneficial, and the elements would drive their future purchase intention via booking online. In contrast, epistemic customers would not consider these elements complements for their curiosity because none of the content would be new. Regarding the antecedent roles of social media, social media encourages both utilitarian and epistemic value via the circulation of information among social networks. Note that its antecedent effects on epistemic value are much stronger than its effects on utilitarian value. This is aligned with the phenomenon that epistemic value involves product-arousing curiosity that provides novelty or the desire to change one’s pace (van den Berge et al. 2021), and social media assists the customer in obtaining unexpected and unstructured information. As shown in the previous literature (Appel et al. 2020), social media has become the central arena in which individuals receive large amounts of data, exchange knowledge, share content and stories from their lives with others, and receive information about the world around them. This makes social media culturally relevant even if the information might be of questionable accuracy. Note that combining both stimuli (S; OTA platforms and social media) would drive better organisms (O; customer value) and therefore lead to a better response (R; the intention to book online).

Second, in addition to utilitarian value, which mainly focuses on functionality, our model also included epistemic value, which pays attention to the novelty of information that satisfies customers’ curiosity. By using S–O–R theory, this study extends the investigation of epistemic value in the tourism and hospitality sectors. The S–O–R paradigm has been applied to various topics, such as pro-environmental behaviors and online shopping. Still, we are among the first to develop an S–O–R model for OTA settings. Our findings prove that utilitarian value plays a more crucial role than epistemic value in stimulating the intention to book online via OTAs, which implies that behavioral intention in our focal context is more functionality driven than curiosity driven. General benefits are still the base of customers’ selection and usage intention. Furthermore, although social media significantly impacts both utilitarian and epistemic value, its direct positive effect on epistemic value is more prominent. Empirically speaking, epistemic value is aroused by service experience through social media, which implies that the flow of unexpected content within a social network effectively responds to customers’ need for new information, mainly WOM. For instance, people may rely more on reviews that spread through social networks than on the official reviews on OTA platforms. Hence, the role of epistemic value as the organism (O) should be encouraged by allowing a greater presence on social media to increase pleasant service experience.

Third, the model revealed negative moderating effects of social influence on utilitarian value and intention to book online. This stands in contrast with our prediction that social influence would positively moderate utilitarian value. However, social influence did not significantly impact epistemic value or behavioral intention. Fourth, we contribute to tourism and hospitality research by including political instability as a moderator of the relationship between customer value and intention to book online. As predicted, political instability negatively moderated the relationship between utilitarian value and intention, which confirms that negative macro-environments, such as political instability, can lessen OTAs’ opportunities. In contrast, political instability did not significantly influence the relationship between epistemic value and intention.

Managerial implications

This study has several implications for management. First, it is essential to enhance service experience and customer value through both OTA platforms and social media. OTA platforms should properly provide the four elements studied here—ratings and reviews, information on accommodations, visual appeal, and service bundling—to encourage utilitarian value. For instance, appropriately designed hotel reviews without biased information would help maintain a positive customer service experience and facilitate future booking decisions. Moreover, active visual design may promote quality, support customers’ service experience, and attract more prospective customers. Apart from focusing on company platforms, OTAs should enhance their presence on social media to stimulate not only utilitarian value but also epistemic value among customers. Furthermore, as customers are coming to rely more on social media as a source of information, advertising with unexpected content via social media is the most powerful strategy for rebuilding after the COVID-19 pandemic, when people tended not to travel because of health and safety preventions.

Second, managers can stimulate customers’ intention to book through OTAs by mainly communicating utilitarian value (e.g., convenience, functional benefits, and a goal-oriented experience) and simultaneously promoting epistemic value. Regarding utilitarian value, posting ratings and reviews by expert reviewers may improve the reliability of content, thereby increasing convenience and favorability ratings among customers. Moreover, clear information on accommodations, user-friendly visual appeal, and attractive service bundling are the basic attributes of OTA platforms that lead to positive utilitarian value. In terms of epistemic value, managers may use social media networks to drive future intention to book. Customers are more likely to be active on social media—by commenting on, liking, and sharing brand-related content—when they are satisfied with and enjoying the content. Thus, presenting useful advice, tips, and ideas on social media may have both utilitarian and epistemic value for customers. For example, posting unknown facts about a destination on social media may provide knowledge and draw people’s attention to that destination. Furthermore, social media allows customers to actively participate in value cocreation with brands based on their knowledge and preferences for information sharing, which contributes to the development of OTAs and thus helps leverage marketing strategies.

Third, managers should avoid negative social influence and WOM by ensuring that customers’ unsatisfactory experiences and complaints are resolved properly. Negative feedback, bad experiences, and negative WOM can decrease customers’ intention to use OTAs. Fourth, OTAs should consider political instability, especially when entering new markets, such as foreign countries. OTAs should also develop business contingency plans that are effective when instability arises and pay attention to relevant government policies (e.g., national/fiscal policies, campaigns). Because people tend to avoid going out when political instability arises, companies can reduce the price of tourism services or promote the safety of facilities to encourage demand. Finally, managers can use these findings to develop marketing plans for the post-COVID-19 era. The COVID-19 pandemic has led to large losses in the tourism industry. As the global pandemic has come under control, many countries have lessened travel restrictions, and there are now more opportunities for OTAs. To boost customer intention to book online, OTAs should clearly promote safety measures in the information on accommodations and reviews shown on their platforms and social media networks.

Limitations and future research directions

This study investigated OTA service experience in general without a focus on any specific OTA. However, features of OTA platforms vary based on travelers’ perceptions, so future research should be limited to one or two OTAs to better understand customer service quality. In addition, this study focused solely on the tourism industry; thus, future research should include other industries. As only Thai respondents were included in this study, future research should attempt to replicate this study in other countries with different cultural settings. Moreover, we did not limit the number of responses to one per IP address when collecting the data via online questionnaires, to allow respondents working in offices and sharing a network to complete the survey. Future research may do this to improve the quality of data collected via online platforms. Also, this study discussed only two types of value, four elements of service experience through OTA platforms, and service experience through social media; it did not consider customers’ overall perceptions. Finally, future research can examine the behavior of customers who have never used OTAs to determine whether they would be willing to use these services in the future.

Appendix: study questionnaire

Constructs		SFL
Service experience via OTA platforms
Ratings and reviews (RR) (α = 0.884)
RR1	I find that the ratings and reviews of hotels from other users is one of the reasons I like this OTA	0.783
RR2	I am interested in ratings and reviews from other users	0.917
RR3	I find that ratings and reviews from other users are useful	0.849
*RR4	My hotel choices depend on information providing in ratings and reviews	0.688
Information on accommodations (IA) (α = 0.859)
IA1	This OTA website provides adequate hotel information (e.g., rooms, facilities)	0.844
IA2	This OTA website provides adequate information on hotel rates and offers	0.810
IA3	I find that the hotel information on this OTA website is accurate	0.738
IA4	I find that the hotel information on this OTA website is up to date	0.715
Visual appeal (VA) (α = 0.866)
VA1	The colors that are used on the website are attractive	0.766
VA2	The lay-out of this OTA website is attractive	0.877
VA3	The design of this OTA website is eye-catching	0.809
VA4	The esthetics of this OTA website promotes a perception of quality	0.708
Service Bundling (SB) (α = 0.884)
*SB1	For cost and convenience, I would likely book a flight and hotel accommodations together as a single package	0.753
*SB2	For cost and convenience issue, I would find such a package/bundle quite attractive	0.840
SB3	I would be satisfied if I bought the package/bundle at the reduced price	0.757
SB4	If I bought the package/bundle, the deal I am getting would be very good	0.889
SB5	I find that buying flights and hotel accommodations as a package is very economical	0.892
Service experience via social media (α = 0.883)
SM1	I find that posts from OTAs on social media sites (e.g., Facebook, Instagram) can enable me to do or share something with others	0.761
SM2	I find that posts from OTAs on social media sites (e.g., Facebook, Instagram) can be the subject of conversation	0.770
SM3	I find that posts from OTAs on social media sites (e.g., Facebook, Instagram) can enable me to gain knowledge of the opinions of others	0.769
SM4	Posts from OTAs on social media sites (e.g., Facebook, Instagram) can make sure I am up to date	0.810
SM5	I find that posts from OTAs on social media sites (e.g., Facebook, Instagram) can give me useful information	0.762
*SM6	I feel that posts from OTAs on social media sites (e.g., Facebook, Instagram, etc.) can fill an empty moment	0.715
Utilitarian Value of Service Experience (UV) (α = 0.833)
UV1	I accomplished what I wanted to do on this OTA website	0.708
UV2	I found the hotel accommodation and/or the services I was looking for	0.794
UV3	The price of the hotel accommodation and/or services I purchased through this OTA website is right, given the quality	0.785
*UV4	When I book hotel accommodations and/or services through this OTA website, I save time	0.685
UV5	The hotel accommodation and/or services I booked from this OTA website were a good buy	0.711
*UV6	This OTA website offers good economic value	0.538
Epistemic Value of Service Experiences (EV) (α = 0.944)
EV1	The reason that I made the bookings through this OTA is to learn new knowledge	0.864
EV2	I used this OTA website to test the new technologies	0.904
EV3	I used this OTA to experiment with new ways of doing things	0.931
EV4	The reason that I make bookings through this OTA website is to learn various search methods and skills	0.833
EV5	Booking through this OTA website makes my life different	0.824
EV6	I used this OTA website out of curiosity	0.787
Political Instability (PI) (α = 0.913)
PI1	Domestic political instability has an impact on my holiday planning and intention to book	0.757
PI2	I would likely postpone or put my holiday plan on hold if some political situation such as a coup d’état, revolution, political protest, happened in my country	0.871
PI3	I find that domestic political instability is one of the important factors in my holiday planning and intention to book	0.909
PI4	Domestic political violence has an impact on my holiday planning and intention to book	0.869
Social Influence (SI) (α = 0.886)
SI1	People who are important to me think that I should use OTAs	0.881
SI2	People who influence my behavior think that I should use OTAs	0.900
SI3	People around me who use OTAs have more prestige than those who do not	0.746
SI4	People who use OTAs have a high profile	0.699
Behavioral Intention (BI) (α = 0.912)
*BI1	I will say positive things about this OTA website to other people	0.693
BI2	I will visit and make bookings through this OTA website again in the future	0.824
BI3	I will recommend this OTA website to others	0.907
BI4	I will encourage friends and relatives to visit and make a booking through this OTA website	0.892
BI5	I will book other services through this OTA website in the future	0.781

SFL Standardized Factor Loading, OTA online travel agency. Asterisks denote items deleted during the data analysis.

Funding

This work was partially supported by the Ministry of Science and Technology, Taiwan with Grant number: MoST 107-2410-H-155-037-MY3.