THE INFLUENCE OF EWOM ON THE PURCHASE INTENTION OF YOUNG CONSUMERS AT ONLINE TRAVEL AGENTS THROUGH THE EXPANSION ON THE INFORMATION ADOPTION MODEL

Siti Balqis¹, Refi Rifaldi Windya Giri²

Study Program S1 Telecommunication and Informatics Business Management, Faculty of Economics and Business, Telkom University

sitibalqiss@student.telkomuniversity.ac.id¹, rifaldi@telkomuniversity.ac.id²

Abstract

e-WOM has become very popular in recent years due to the increasing number of contributors and the proliferation of mobile platforms for applicatios on social media. In online travel agents, it is currently known that tourist is obtain information from online review websites rather than hotel websites to decide their hotel choices, therefore eWOM significantly influences consumers purchasing decisions. This research contributes to IAM and the extended model will be tested in eWOM research on five online travel egents (Traveloka, Tiket.com, Agoda, Pegipegi, Airbnb). This study aims to examine the effect of argument quality, source credibility, information quantity, and emotional word understanding on the percived usefulness of eWOM, to analyze the effect of perceived usefulness on information adoption and to investigate the effect of information adoption on young consumers purchase intentions. The type of research used in this study is descriptive analysis with quantitative approach with CB-SEM data analysis techniques using the Amos24 program with a minimum sample of 385 respondents.The results show that argument quality, source credibility, information quantity and emotive word comprehension have positive effect on the perceived usefulness of eWOM. Perceived usefulness has positive influence on the information adoption of eWOM, which in turn predicts the young consumers’ purchase intentions. This study makes several contributions to the literature on marketing communications, in particular to eWOM research and IAM theory. Practically, this research provides knowledge and insight to online travel agent site marketers in order to increase the usability and effectiveness of eWOM to attract more young consumers.

Keywords: eWOM, online travel agents, young generation consumers, IAM, argument quality, source credibility.

Introduction

According to (Chong et al., 2018) electronic word of mouth (eWOM) is currently gaining popularity as the most influential source of information on digital consumers in the travel industry. Electronic word of mouth is any positive or negative statement made by a potential customer, former customer towards a product or service available on the internet. In this case customers prefer as well as utilize interpersonal communication with other customers to build their trust or confidence in the product or service rather than the advertising carried out by the company.

E-WOM has become very popular in recent years due to the increasing number of contributors as well as the rise of mobile platforms for applications on social media (Lee et al., 2009). The way social networks work through a particular social media site can allow users to create profiles, share information and be able to interact and communicate with other users on the same site (Rudyanto, 2018). In online travel agents it is currently known that travelers prefer to obtain information from online review websites rather than hotel websites to decide on their hotel choice therefore eWOM significantly influences consumer purchasing decision making (Dedeoglu, 2019). Some of the most popular eWOM platforms among consumers today are social media websites, review websites, blogs, discussion forums and shopping websites. Through these various eWOM platforms, travelers can share information related to their travel, personal experiences, and opinions in the form of comments via text, photos and videos (Pihlaja et al., 2017).

According to research conducted by [28] 80% of tourists use the internet to search for hotel information. In addition to the hotel's official website, bookings through Online travel agents (OTAs) have become very popular. Since the launch of Expedia in 1996, Priceline in 1997, Hotwire in 2000, and the introduction of other OTAs, the hotel's distribution channels have changed drastically. OTAs have taken a big part of traditional ordering channels (Inversini & Masiero, 2014)

According to (Antara et al., 2017)Online travel agents (OTAs) are travel agents that currently play a very important role as an online marketing medium through websites. Currently the company has a website to promote to consumers and this type of website also varies according to the purpose and content. Currently, the website can be used by companies as "shops" to offer their tourism products to potential tourists.

This research focuses on the younger generation of online travel agent consumers with an age range of 18-35 years. Based on a survey conducted by shows that the majority of respondents from 37.8% are domestic tourists aged 25-39 years, followed by 26.4% aged between 15-24 years. Young travelers prefer to use SNS to get travel information. SNS play an important role in influencing behavior in the buying of young consumers as well as contributing greatly to the growth of the tourism industry (Statista, 2019)(Jamaludin et al., n.d.). Therefore, many companies have turned their attention and invested heavily to understand the attitudes and behaviors of young consumers in the tourism industry. (Liu et al., 2019)

Furthermore, this study will explore popular travel SNS such as Traveloka, Tiket.com, Agoda, Pegipegi.com and Airbnb as young Indonesian travelers mostly use this site for their purchasing decision-making. According to a survey of online travel agencies in Indonesia conducted in November 2021, 86 percent of respondents used Traveloka the most, followed by Tiket.com at 57 percent, Agoda at 37 percent, Pegipegi.com at 33 percent and Airbnb at 6th at 13 percent. (Rakuten Insight, 2021)

According to research conducted by Google, Temasek in the 2019 SEA e-conomy report found that Indonesia's online travel market is still ahead in Southeast Asia. This is driven by the growth trend of the tourism industry in Indonesia. The search rate for online Travel such as traveloka, tiket.com was very large in 2019 so that the growth reached 20% from 2018 so that this affected the increase in transaction value (Gross Merchandise Value / GMV) so that it reached US $ 10 billion (Databoks, 2019).

At the beginning of 2020, Indonesia experienced a decrease in tourist visits by 13.5 percent. Indonesia's borders have been closed to international tourists since April and many attractions from museums to hiking trails have been closed. As of February 2020, the Indonesian government has set aside 443 billion rupiah to encourage domestic tourists to visit one of the ten tourist destinations promoted in Indonesia. Globally, the tourism sector is greatly affected by the spread of the virus (Japutra & Situmorang, 2021), (Statista, 2021) shows that in 2020 there were around 4.05 million foreign tourist visits to Indonesia, this number decreased significantly compared to the previous year due to travel restrictions due to COVID-19. However, over time, the positive number of Covid-19 in Indonesia has begun to decline. So that the tourism sector in Indonesia has the potential to develop again. This fact can be seen from the following chart:

Figure 1.1 Statistic of Corona Virus (COVID-19)

Source : Covid19.go.id || https://covid19.go.id/

From the figure above, it can be seen that the highest number of covid-19 virus cases in July 2021 was 954,940 people. However, it is currently seen that in October 2021 the covid rate in Indonesia has dropped drastically due to the government's PPKM (enforcing restrictions on community activities) policy and requiring vaccines to all communities in various regions. (Badan Pusat Statistik, 2021) explained that the number of tourist visits to Indonesia in October 2021 reached 151.03 thousand visits so that when compared to September 2021, it increased by 21.73 percent. In addition, the room occupancy rate (TPK) of star classification hotels in Indonesia in October 2021 reached 45.62 percent or can be said to have increased by 8.14 points compared to the TPK in October 2020. The average length of stay of foreign guests and domestic guests also increased by 0.02 points compared to october 2020.

According to a survey conducted by JAKPAT, it shows that 55% of tourism respondents vacation by staying in lodging. Villas and hotels are one of the people's favorite options for vacation. The survey also showed that 65% chose a secluded villa to enjoy their leisure time while 38.7% chose an inner-city hotel. (Databoks, 2021) The Information Adoption Model (IAM) is something that affects individuals motivated by e-WOM that will depend on how individuals adopt information IAM consists of four constituents namely the quality of the argument, the credibility of the source, the perceived usefulness and adoption of the information. (Song et al., 2021) There is currently an increase in competitiveness towards Online travel agent (OTA) sites to attract customers to their sites. Although consumers are exposed to high exposure to eWOM information from various platforms, the effectiveness of eWOM information in influencing consumers' purchasing decisions remains ambiguous. There are increased concerns about the credibility of the sources and the quality of eWOM information. (Hussain et al., 2020)

Research conducted earlier revealed that the sources of information on tourism SNS are said to be less credible because the level of anonymity in SNS varies which can lead to a lack of personal accountability and identity fraud problems. On the other hand (Tham et al., 2013), it is stated that the credibility of high sources of information on social media is a very important role in influencing decisions related to tourism selection because there is travel content that attracts the attention of other users. Therefore, the issue of credibility of sources and the quality of eWOM in SNS related to tourism(Agag & El-Masry, 2017) is still a debate of previous researchers.

In the development of e-WOM, e-WOM information sources such as the quantity of information and the understanding of emotive words have played an important role influencing in the adoption of individual information. Based on previous research conducted by (Gopinath et al., 2014) information in the eWOM should take into account "how many people say" and the emotions associated with "what people say" (Agag & El-Masry, 2017) asserts that there is a gap in research on the role and emotional impact in eWOM and there is a need to investigate the emotional impact of eWOM communication through social networks on tourism decisions.

The research gaps outlined need to be addressed by proposing further research on the quantity of information and the understanding of emotive words in the eWOM perspective. Therefore in this study, it proposes to test the expanded IAM by incorporating two additional constructs namely the quantity of information and the understanding of emotive words into the original IAM. The research carried outn by proposes that future research should add more variables in the original IAM or combine IAM with other models to improve its application in different contexts (Tseng & Wang, 2016).

(Hussain et al., 2020) said that there is a need to integrate IAM and TRA to investigate the characteristics of the use of information that affect consumer behavior intentions. To enhance the significance of the study, researchers will link eWOM information to young consumer behavior towards information by adding a construct of consumer repurchase intent as a result. The main objective of the study is to examine the influence of the quality of arguments, the credibility of sources, the quantity of information and the understanding of the emotive word on the perceived benefits of eWOM aimed at investigating the influence of information adoption on the repurchase intentions of young consumers.

Based on the foregoing, this study will contribute to the expansion of IAM and the model to be expanded by conducting an eWOM investigation in SNS Online travel agents on the purchase intentions of young consumers. Therefore, researchers are interested in conducting a research entitled "The Effect of eWOM on the Buying Intentions of Young Generation Consumers on Online travel agents through the Expansion of information adoption models"

Research Methods

In this study, researchers used quantitative methods. According to the quantitative method, which is a method based on the philosophy of positivism used to examine certain populations or samples, data collection in this study uses research instruments and data analysis is quantitative or statistical, which has the aim of testing the hypothesis that the researcher has set. Based on its objectives, this study uses descriptive analysis. As descriptive analysis says it can be used to find out the value of a standalone variable, either one or more variables (independent) without making comparisons or attributing them to other variables. In this study, the data used and studied were data derived from samples contained in the population. (Sugiyono, 2017) (Rochaety et al., 2019) In this study, the minimum number of samples taken based on the Cochran formula was 384.16 respondents and could be rounded to 385 respondents. The research unit in this study was an individual who used online travel agents (Traveloka, Tiket.com, Agoda, Pegipegi, and Airbnb) with characteristics in young consumers with an age range of 18-35 years. The data analysis technique in this study used descriptive analysis and CB-SEM using AMOS 24 software. Descriptive analysis can be used to find out the value of a standalone variable, either one or more variables (independent) without making comparisons or linking them to other variables. The independent variables in this study are e-WOM and the dependent variable in this study is purchase intention. In this case, this study uses covariance based matric structural equation modeling (CB-SEM) or can be called hard modeling. Hard modeling aims to provide a statement about the relationship of causality or it can be called a causal relationship. The main purpose of covariance based SEM (CB-SEM) Hard modeling is to test whether causality has been built on the theory and whether this model can be confirmed with its empirical data (Ghozali, 2017)

Result and Discussion

This research is sourced from primary data obtained through the distribution of questionnaires that are shared online, using G-forms through social media such as Instagram, Twitter, WhatsApp, and Line. Where the criteria for respondents are internet users who already have social networking sites aged over 18 years to 35 years. The number of respondents obtained through the distribution of questionnaires was 285 respondents.

Outer Model

The use of outer models aims to show the specific relationship between the estimated indicator or parameter and the latent variable. That is by approaching validity tests and reliability tests.

Convergent Validity Test

The convergent validity test according to (Ghozali, 2017) is an indicator of a latent construct that must converge or share a high proportion of variants. To measure the validity of the construct can be seen from the value of the loading factor. This test is carried out to see each indicator that has a high loading factor value with the construct and has a high extracted variance value. The validity test criteria in this study are that the indicator is declared valid if the value of the loading factor and variance extracted is greater than 0.5.

Table 1

Construct (Latent Variable)	Indicators	Standardized Loading Estimate	Standardized Loading Estimate²	error	Variance Extracted
Quality of Arguments	KA1	0.741	0.549	0.451	0.516
	KA2	0.730	0.533	0.467
	KA3	0.694	0.482	0.518
	KA4	0.706	0.498	0.502
	Σ	2.871	2.062	1.938
	Σ²	8.243
Source Credibility	KS1	0.692	0.479	0.521	0.518
	KS2	0.668	0.446	0.554
	KS3	0.720	0.518	0.482
	KS4	0.771	0.594	0.406
	KS5	0.742	0.551	0.449
	Σ	3.593	2.588	2.412
	Σ²	12.910
Understanding the Emotional Word	PKE1	0.663	0.440	0.560	0.536
	PKE2	0.775	0.601	0.399
	PKE3	0.732	0.536	0.464
	PKE4	0.754	0.569	0.431
	Σ	2.924	2.145	1.855
	Σ²	8.550
Quantity of Information	KI1	0.683	0.466	0.534	0.526
	KI2	0.776	0.602	0.398
	KI3	0.761	0.579	0.421
	KI4	0.674	0.454	0.546
	Σ	2.894	2.102	1.898
	Σ²	8.375
Perceived Uses	KYD1	0.731	0.534	0.466	0.520
	KYD2	0.748	0.560	0.440
	KYD3	0.730	0.533	0.467
	KYD4	0.674	0.454	0.546
	Σ	2.883	2.081	1.919
	Σ²	8.312
Information Adoption	AI1	0.749	0.561	0.439	0.553
	AI2	0.720	0.518	0.482
	AI3	0.756	0.572	0.428
	AI4	0.749	0.561	0.439
	Σ	2.974	2.212	1.788
	Σ²	8.845
Purchase Intent	NP1	0.721	0.520	0.480	0.505
	NP2	0.698	0.487	0.513
	NP3	0.713	0.508	0.492
	Σ	2.132	1.515	1.485
	Σ²	4.545

Source: Amos 24 Data Processing (Author, 2022)

Based on table 3.1 above, it is known that each indicator in each variable has a loading factor value of > 0.5 so that it can be concluded that all indicators are declared valid.

Uji Reliabilitas

According to reliability can be related to the degree of data consistency and data stability in a study (Sugiarto, 2017) Reliability measurement can be done with composite (construct) reliability with cut-off-value criteria for construct reliability of at least 0.70, but reliability values of 0.60-0.70 are still acceptable provided that the validity of the indicators in the model is good

Table 2

Construct (Latent Variable)	Indicators	Standardized Loading Estimate	Standardized Loading Estimate²	error	Reliability
Quality of Arguments	KA1	0.741	0.549	0.451	0.810
	KA2	0.730	0.533	0.467
	KA3	0.694	0.482	0.518
	KA4	0.706	0.498	0.502
	Σ	2.871	2.062	1.938
	Σ²	8.243
Source Credibility	KS1	0.692	0.479	0.521	0.843
	KS2	0.668	0.446	0.554
	KS3	0.720	0.518	0.482
	KS4	0.771	0.594	0.406
	KS5	0.742	0.551	0.449
	Σ	3.593	2.588	2.412
	Σ²	12.910
Understanding the Emotional Word	PKE1	0.663	0.440	0.560	0.822
	PKE2	0.775	0.601	0.399
	PKE3	0.732	0.536	0.464
	PKE4	0.754	0.569	0.431
	Σ	2.924	2.145	1.855
	Σ²	8.550
Quantity of Information	KI1	0.683	0.466	0.534	0.815
	KI2	0.776	0.602	0.398
	KI3	0.761	0.579	0.421
	KI4	0.674	0.454	0.546
	Σ	2.894	2.102	1.898
	Σ²	8.375
Perceived Uses	KYD1	0.731	0.534	0.466	0.812
	KYD2	0.748	0.560	0.440
	KYD3	0.730	0.533	0.467
	KYD4	0.674	0.454	0.546
	Σ	2.883	2.081	1.919
	Σ²	8.312
Information Adoption	AI1	0.749	0.561	0.439	0.832
	AI2	0.720	0.518	0.482
	AI3	0.756	0.572	0.428
	AI4	0.749	0.561	0.439
	Σ	2.974	2.212	1.788
	Σ²	8.845
Purchase Intent	NP1	0.721	0.520	0.480	0.810
	NP2	0.698	0.487	0.513
	NP3	0.713	0.508	0.492
	Σ	2.132	1.515	1.485
	Σ²	4.545

Source: Data Processing of AMOS 24 (Author, 2022)

Based on table 3.2 above, it is known that each indicator in each variable has a construct reliability value above 0.70, so it can be concluded that all indicators are declared reliable.

Structural Model Evaluation (Inner Model)

According to The Show the specifics of the causal relationship between the latent veriabels. In this study, the evaluation of structural models used goodness of fit.

Goodness of Fit

Goodness- of-fit (GOF) is performed to measure the suitability of an observational or actual input in a covariant matric or correlation with a prediction of the proposed model. In goodness of fit there are three types of measures, namely absolute fit measure, incremental fit measures and parsimonious fit measure. Absolute fit measure measures overall fit (both structural models and measurement models together).

Table 3 Result of Goodness of Fit

Category	Fit Measurement	Result	Fit Criteria	Conclusion
About Fit Measures	CMIN/DF	2.055	0 ≤ CMIN/DF ≤ 5.00	Good Fit
	GFI	0.885	>0.90	Marginal Fit
	RMSEA	0.052	0.05≤ RMSEA ≤0.08	Good Fit
Incremental Fit Measures	AGFI	0.862	≥ 0.90	Marginal Fit
	TLI	0.921	≥ 0.90	Good Fit
	NFI	0.872	≥ 0.90	Marginal Fit
Parsimonious Fit Measures	PNFI	0.78	0.60 ≤ PNFI ≤0.90	Good Fit
	PCFI	0.831	0 ≤ PGFI ≤ 1.0	Good Fit

Sumber: Olah Data AMOS 24 (Penulis, 2022)

Based on table 3. 3 it is known that all criteria in the Goodness- of-fit (GOF) test already have a good fit value, only AGFI has a marginal fit value. Revealed that the use of 4-5 goodness of fit test criteria is considered sufficient to assess the feasibility of a model, as long as each criterion of goodness of fit, namely absolute fit indices, incremental fit indices, and parsimony fit indices is represented. Based on this, it is concluded that the research model carried out has met the feasibility test.

Hypothesis Testing

Hypothesis testing in this study uses the t-value contained in the AMOS 24 program is the critical ratio (CR) value of the overall fit model with a significance level of 0.05(Ghozali, 2017). This study consists of 6 one-tailed hypotheses with criteria if CR ≥ 1.65 or P ≤ 0.05 then H0 is rejected. Vice versa, if cr ≤ 1.65 or P ≥ 0.05 then H0 is supported. The results of hypothesis testing are presented as follows:

Table 4 Test Result of Hipotesis

Hipotesis	Description	Estimate	S.E.	C.R.	P	Conclusion
H1	The quality of the argument positively affects the perceived usefulness	0.206	0.085	2.415	0.008	Supported
H2	The credibility of the source positively affects the perceived usefulness	0.148	0.08	1.856	0.031	Supported
H3	The quantity of information positively affects the perceived usefulness	0.182	0.05	3.651	***	Supported
H4	Understanding the word emotional positively affects perceived usefulness	0.363	0.077	4.688	***	Supported
H5	Perceived usefulness positively affects the adoption of information	0.887	0.073	12.142	***	Supported
H6	Information adoption positively affects consumers' purchase intentions	0.674	0.065	10.317	***	Supported

Sumber: Olah Data AMOS 24 (Penulis, 2022)

Conclusion

Based on the results and discussion in the study, it can be concluded as follows: The quality of the arguments has a positive effect on the perceived usefulness, meaning that the better the quality of the arguments given, the more the usefulness felt by potential consumers will increase, The credibility of the source has a positive effect on the perceived usefulness, meaning that the more credible the source that provides information, the more usefulness felt by potential consumers will increase, The quantity of information has a positive effect on the perceived usefulness, meaning that the more information provided, the more usefulness felt by potential consumers will increase, Understanding emotional words has a positive effect on the perceived usefulness, meaning that the better the understanding of emotional words in the information provided, the more usefulness felt by potential consumers will increase, The perceived usefulness has a positive effect on the adoption of information, meaning that the higher the benefits felt by potential consumers, the intention of adopting information will increase, The adoption of information has a positive effect on purchase intentions, meaning that the higher the use of information, the purchase intention will increase.

References

Agag, G. M., & El-Masry, A. A. (2017). Why Do Consumers Trust Online Travel Websites? Drivers and Outcomes of Consumer Trust toward Online Travel Websites. Journal of Travel Research, 56(3), 347–369. https://doi.org/10.1177/0047287516643185

Antara, M., Astuti, L. S., & Madiun, I. N. (2017). E-MARKETING PEMASARAN VILLA MENGGUNAKAN ONLINE TRAVEL AGENTS (1st ed.). Pelawa Sari.

Badan Pusat Statistik. (2021). Perkembangan Pariwisata dan Transportasi Nasional Oktober 2021.

Chong, A. Y. L., Khong, K. W., Ma, T., McCabe, S., & Wang, Y. (2018). Analyzing key influences of tourists’ acceptance of online reviews in travel decisions. Internet Research.

Databoks. (2019). Pasar Travel Online Indonesia Tertinggi di Asia Tenggara.

Databoks. (2021). Stay on Vacation, Metode Populer Berlibur Semasa Pandemi. https://databoks.katadata.co.id/datapublish/2021/01/27/stay-on-vacation-metode-populer-berlibur-semasa-pandemi

Dedeoglu, B. B. (2019). Are information quality and source credibility really important for shared content on social media? The moderating role of gender. International Journal of Contemporary Hospitality Management, 31(1), 513–534.

Ghozali, I. (2017). Model Persamaan Struktural: Vol. VII (7th ed.). Badan Penerbit Universitas Diponegoro.

Gopinath, S., Thomas, J. S., & Krishnamurthi, L. (2014). Investigating the relationship between the content of online word of mouth, advertising, and brand performance. Marketing Science, 33(2), 241–258. https://doi.org/10.1287/mksc.2013.0820

Hussain, S., Song, X., & Niu, B. (2020). Consumers’ Motivational Involvement in eWOM for Information Adoption: The Mediating Role of Organizational Motives. Frontiers in Psychology, 10. https://doi.org/10.3389/fpsyg.2019.03055

Inversini, A., & Masiero, L. (2014). Selling rooms online: The use of social media and online travel agents. International Journal of Contemporary Hospitality Management. https://doi.org/10.1108/IJCHM-03-2013-0140

Jamaludin, M., Sultan, P., Shah, I., Aziz, A., & Mariapan, M. (n.d.). Millennial Travelers Decision Making Influenced Through User-Generated Contents And Psychological Attributes On Destination Loyalty To A Tropical Island Human Willife Management View project Education View project. www.jthem.com

Japutra, A., & Situmorang, R. (2021). The repercussions and challenges of COVID-19 in the hotel industry: Potential strategies from a case study of Indonesia. International Journal of Hospitality Management, 95. https://doi.org/10.1016/j.ijhm.2021.102890

Lee, M., Rodgers, S., & Kim, M. (2009). Effects of valence and extremity of eWOM on attitude toward the brand and website. Journal of Current Issues & Research in Advertising, 31(2), 1–11.

Liu, Q., Zhang, · Xiaoyu, Zhang, L., Zhao, · Yang, & Zhao, Y. (2019). The interaction effects of information cascades, word of mouth and recommendation systems on online reading behavior: an empirical investigation. Electronic Commerce Research, 19, 521–547. https://doi.org/10.1007/s10660-018-9312-0

Pihlaja, J., Saarijärvi, H., Spence, M. T., & Yrjölä, M. (2017). From electronic WOM to social EWOM: Bridging the trust deficit. Journal of Marketing Theory and Practice, 25(4), 340–356. https://doi.org/10.1080/10696679.2017.1345593

Rakuten Insight. (2021, January). Most popular online travel agencies among consumers in Indonesia as of November 2020. Rakuten Insight.

Rochaety, E., Tresnati, R., & Latief, A. M. (2019). Metodologi Penelitian Bisnis (2nd ed.). Mitra Wacana Media.

Rudyanto, R. (2018). Pengaruh Pemasaran jejaring media sosial dan keterkaitan konsumen terhadap niat beli konsumen. Jurnal Manajemen Dan Pemasaran Jasa, 11(2), 177–200. DOI: https://doi.org/10.25105/jmpj.v11i2.3126

Song, B. L., Liew, C. Y., Sia, J. Y., & Gopal, K. (2021). Electronic word-of-mouth in travel social networking sites and young consumers’ purchase intentions: an extended information adoption model. Young Consumers, 22(4), 521–538. https://doi.org/10.1108/YC-03-2021-1288

Statista. (2019). Travel and tourism in Indonesia.

Statista. (2021). Number of International Visitor Arrivals Indonesiia 2011-2020. https://www.statista.com/statistics/707633/number-of-international-visitor-arrivals-in-indonesia/

Sugiarto. (2017). Metodologi Penelitian Bisnis. Penerbit Andi.

Sugiyono. (2017, September). Metode Penelitian Bisnis. Alfabeta.

Tham, A., Croy, G., & Mair, J. (2013). Social Media in Destination Choice: Distinctive Electronic Word-of-Mouth Dimensions. Journal of Travel and Tourism Marketing, 30(1–2), 144–155. https://doi.org/10.1080/10548408.2013.751272

Tseng, S. Y., & Wang, C. N. (2016). Perceived risk influence on dual-route information adoption processes on travel websites. Journal of Business Research, 69(6), 2289–2296. https://doi.org/10.1016/j.jbusres.2015.12.044