Factors Affecting Online Reservation Decisions Through Hotel Websites: An Empirical Study from Can Tho City, Vietnam

Abstract

Many consumers are opting for online booking over traditional booking systems. Customers can actively seek out information about hotels and lodging services, as well as book rooms, at any time and from any location. Customers also feel more supported when they interact with virtual assistants or professionals. Recognizing this issue, several hotels have focused on improving their websites by incorporating aspects that encourage customers to book directly through the hotel's website. The study's goal is to discover what factors impact people's decisions to book a hotel stay through the hotel's website. Therefore, hotel managers and owners can make decisions to improve the hotel website to attract residents to Can Tho City. The factors are website quality, affective commitment, social presence, and e-trust that affect customers' decision to book through the hotel website. The study uses quantitative methods to collect data from 180 residents living in Can Tho. Through data analysis on SPSS and Amos software, the research results show that three factors considered, namely website quality, affective commitment, and social presence, positively influence customers' booking decisions. This finding also suggests that e-trust is less critical to residents in Can Tho City, different from what the study had predicted.

1. Introduction

The tourism and hotel industry is a vital and growing industry in Vietnam. Today, along with the rapid development of information technology, the trend of popularizing electronic devices connected to the Internet has changed customers’ consumption habits. Therefore, the trend of online hotel booking is also gradually becoming popular. According to a survey by Indochina Capital Group on the habits and desires of Vietnamese tourists, domestic tourism from 2015 to 2019 has increased from 57 million to 90 million visitors; in which 64% of tourists tend to use online channels to book rooms (Thanh, 2020). Tourists will be more proactive and independent in finding hotels on online booking platforms such as hotel websites and online travel agencies. It is a tourism method with solid interaction between the hotel and the customer, between the customer and the customer through two parties’ agreement and trust. Moreover, it also brings convenience and the ability to access the website anytime, anywhere, making it easy for users to review many objective reviews from customers who have experienced services in the hotel.

As human income and needs increase day by day, hotel choices will become vital. Online booking is growing increasingly popular in Vietnam, and it has an advantage over traditional booking in terms of ease and speed, as seen by the expansion of websites such as hotel websites, travel agent booking websites, and smartphone booking applications. As the popularity of booking platforms grows, so does competition in various areas of the website. Hudson (2020) admits renowned brands (TripAdvisor, Booking.com) are more likely to attract users to book, besides pointing out that only 7% of bookings are made on the hotel’s official website. As a result, the focus of this study will be on the elements that impact booking decisions on hotel websites. Customers can get more benefits from booking a room through a hotel website, such as direct communication with the hotel, a full description of the hotel property and location, browsing more photos and videos, comparing better rates, and avoiding paying additional booking fees (O’Connor & Frew, 2004; Sparks & Browning, 2011). Moreover, it also explores the important factors influencing hotel website booking choices, which in turn can assist in the development of more comprehensive and optimized hotel websites.

This research looks at how elements such as website quality (website functionality, usability, and security and privacy), e-trust, affective commitment, and social presence in hotel websites influence the booking decisions for accommodation services of Can Tho residents. Research objects are both men and women from 18 to 35 years old, will survey quantitative questions, and the total number of survey samples collected is 180 samples. Collected quantitative data will be numerically processed on Excel software, then analyzed and evaluated on SPSS and Amos software.

2. Literature Review and Hypotheses

2.1. Booking Decision

In the hospitality industry, hotel booking intention is related to tourists’ willingness to book a room via a hotel website for accommodation services (Lien et al., 2015). In addition, the purchase decision reflected a consumer’s desire to book via the hotel website. According to Wang et al. (2015), online purchasing decisions were related to procedures for evaluating website quality and product/ service information. Li et al. (2017) showed that the benefits of the official hotel website compared to other booking systems (OTAs) played an important role in increasing the online booking decision of tourists. Online reviews had a significant influence on consumers’ booking decision-making process (Nguyen et al., 2021; Mongkhonvanit, 2020). Besides, the effectiveness of the website quality element also had a positive impact on the guest’s booking tendency via hotel websites (Kouzmal et al., 2020). Gefen and Straub (2000) defined trust and website quality as determinants of purchasing through e-commerce. Because of the high level of trust, consumers preferred to choose the form of booking through the hotel website (Baki, 2020). According to Bilgihan and Bujisic (2015), an affective commitment was a precursor to e-loyalty and influenced customers’ e-trust in online suppliers. A hotel website that was transparent and responsive to guest satisfaction would gain trust and return. And lastly, the social presence of the hotel website increased e-trust and affected the hotel booking of tourists (Amin et al., 2021).

2.2. Website Quality

Liu and Zhang (2014) proposed a comprehensive model for measuring website quality relating to the choice of online booking channels for accommodation services, including four elements: information quality, service quality, accessibility, and trust and privacy. Additionally, Wang et al. (2015) developed a theoretical framework for assessing hotel website quality and defined that functionality and usability constituted a significant dimension of website quality’s effectiveness. Inheriting the factors of Wang et al.’s (Wang et al., 2015) hotel website quality, Ali (2016) developed the constructs by adding the “security and privacy” element. These factors were previously explicitly designed to consider hotel website quality in developing countries. Thus, the components of the “website quality” factor in this study got from Ali (2016) and Wang et al. (2015). The hotel website quality includes functionality, usability, security, and privacy. Nguyen et al. (2020) urged online commodity traders to improve their client satisfaction levels. Customers must be able to simply search for product information and navigate to other pages on a user-friendly website.

Website quality was a crucial driver of e-trust since web indications boosted customers’ perceived control over an online vendor’s activities by directly altering their perceived safety and confidentiality (Hoffman et al., 1999). According to Shelat and Egger (2002), providing accurate and helpful information on websites might increase users’ perceived trustworthiness. Tourism and hospitality academics were increasingly interested in the relationship between website quality and e-trust (Fam et al., 2004; Kim et al., 2011; Sparks & Browning, 2011). Customer satisfaction and purchase intentions were influenced by aspects such as hotel websites, system quality, information quality, and security and privacy (Ali, 2016). E-trust was built, according to Ladhari and Michaud (2015), when passengers had good expectations about a hotel’s website. Travelers anticipated the hotel to deliver on its commitments during this period, and this expectation originated from their trust in the hotel website’s accuracy (Lien et al., 2015). De Wulf et al. (2006) discovered that the enjoyment gained from visiting a website led to a higher strong commitment to online purchases. The honesty of the hotel website and its ability to supply tourists with the promised services and amenities when staying at the hotel determines a traveler’s confidence (Ponnapureddy et al., 2017; Rather, 2018). Most academics agreed that passengers who trust online hotel websites were more likely to devote themselves to their connection and create online booking intentions (Agag & El-Masry, 2016).

H1: Hotel website quality has a positive effect on e-trust.

H2: Hotel website quality has a positive effect on affective commitment.

H3: Hotel website quality has a positive effect on online hotel booking decisions.

2.3. E-trust

In business exchanges, trust was a vital factor (Akrout, 2019). Trust was defined as a party’s belief that its needs would be met in the future by actions taken by the other party (Anderson & Weitz, 1989). According to The Economist (2016), it was easy to forget that even the most trivial commercial transactions relied on small acts of trust. Studies on the relationship between measures of confidence and economic growth have found a strong association between these two factors. More complex partnerships, of the kind that helped long-term economic growth, required a much higher level of trust. One of the sectors most strongly influenced by the trust was the service sector (Grayson & Ambler, 1999). According to Gefen et al. (2003), consumer trust was also important for online commerce, and building trust with customers was seen as a critical benefit. Online trust could be distinguished from the offline trust based on the physical distance between buyers and sellers, the absence of a salesperson, and the distance between buyers and products (Horppu et al., 2008). Online trust was built through the belief that the provider had nothing to gain by cheating, the belief that there were safety mechanisms built into the website, and by having a typical interface, which meant, moreover, ease of use (Gefen et al., 2003). For hotel websites, when consumers felt comfortable interacting in online transactions, they were more likely to develop trust (Cheng et al., 2019). Ratnasingham (1998) argued that beliefs allowed people to take risks. Therefore, travelers’ state of trust towards the hotel website was a significant determining factor when they considered interactions with hotel websites. When making transactions on the website, customers would communicate their needs and personal information. They expected the website to be a reliable medium for transactions, and that the supplier would behave honestly and professionally when fulfilling a customer’s request (Bauman & Bachman, 2017). The more a traveler trusts a hotel website, the lower the perceived online transaction risk, and the higher the decision to book a hotel room (Lien et al., 2015).

H4: E-trust has a positive effect on commitment.

H5: E-trust has a positive effect on online hotel booking decisions.

2.4. Affective Commitment

According to Mercurio (2015), the “affective commitment” factor might mediate and influence client behavior and reactions to transactions with the company. In the pieces of literature on organizational and buyer behavior, a commitment was also seen as crucial. Organizational commitment was one of the oldest types of relational commitment, and it was vital to the hotel’s internal ties (Becker, 1960).

Customers’ emotional attachment and feelings toward a network operator were quantified by affective commitment, but guests’ determination to stay with the company owing to a lack of better options was measured by continuance and calculative commitment. Furthermore, the affective commitment had been demonstrated to have a beneficial effect on repurchase intention (Li et al., 2006; Jones et al., 2007; Chih et al., 2014). This study confirmed that affective commitment could improve employee creativity and performance (Astuty & Udin, 2020). Customers would gain trust as a result of commitments, which was defined as “consumer confidence in the quality and reliability of the service provided” (Garbarino & Johnson, 1999). The commitment made here about the service or product that the customer would receive as agreed or the rewards that would be offered when the consumer returned. In general, couponing was used to raise short-term sales (Taylor, 2001), gain new customers (Taylor & Long-Tolbert, 2002; Varadarajan, 1984), and encourage repeat purchases of a brand (Krishna & Shoemaker, 1992). Customer returning from previous experiences was described as having a positive attitude and dedication toward an online shop that led to repeat purchase behavior based on previous occasions (Srinivasan et al., 2002).

H6: Affective commitment has a positive effect on online hotel booking decisions.

2.5. Social Presence

According to Algharabat et al. (2018), the concept of “social presence” represented people’s perception of a specific content that was communicated and presented through an intermediary interface such as a website or a search engine. Furthermore, Gao and Li (2018) stated that social presence also reflected the interactions between buyers and sellers. Gefen and Straub (2004) conducted a study to discover that social presence also influenced consumers’ purchasing decisions through three issues: a sense of human contact, a sense of personalness, and a sense of human sensitivity to the website. Firstly, a sense of human contact means that customers perceive and interact with buyers or sellers on the website (Choi, 2016). Hence, customers collected the necessary information and made a booking decision via hotel websites (Jiang et al., 2010). Secondly, Ye et al. (2019) admitted that a social presence was created through personalized amenities such as greetings, personalized information, and providing tailored booking recommendations to customers based on the above facilities. Furthermore, Head and Hassanein (2004) showed that conveying a social presence was high through providing personalized images and text related to products, thus potentially able to create customer interest and intention to book. Thirdly, a sense of human sensitivity was the third characteristic that social presence contained, which was measured by the level of communication feeling, and the emotional experiences of customers when manipulating an intermediary interface such as a website (Lu et al., 2016). Therefore, a website with a good social presence would easily connect and interact with customers, then the customer’s purchase decision would be increased (Gao & Li, 2018). Researchers had also demonstrated that a higher level of social presence would have a positive effect on the reliability and truthfulness of a website, thereby increasing online trust in customers booking intentions of customers (Head & Hassanein, 2004).

H7: Social presence has a positive effect on e-trust.

H8: Social presence has a positive effect on online hotel booking decisions.

The proposed research model of this study is depicted in Figure 1.

Figure 1: The Proposed Research Model

3. Research Methods and Materials

3.1. Research Design

In this study, a quantitative research method has been applied and implemented to determine the impact of factors on the decision to make a reservation through the hotel websites of residents in Can Tho. Quantitative data collection through a questionnaire was adjusted and sent to the participants. In the profile analysis, participants were asked if they had ever used the hotel’s website to make a reservation. Results completed with “Have never” answers will be discarded as unsuitable for further study. In the survey of the affecting factors, the participants were asked to rate the booking experience through the hotel website using a 5-point Likert scale format from 21 reviewed and developed items. The survey items were developed based on the results from the review of literature related to hotel online booking, identified theories, frameworks, models, and results of other researchers and authors.

3.2. Data Collection

Firstly, the data were transformed into questions and affirmations so that participants could openly share their views. Then the opinions of 180 people were collected as the primary data for the study. To collect data, participants were presented with Likert or questions statements, usually with 5 or 7 items indicating their level of agreement on each issue. Each item is assigned a score, allowing for quantitative analysis of the data.

Secondly, during data processing, the collected data must be preliminarily processed and filtered based on the total number of people surveyed. Primary data were generated using the statistical software package SPSS (Statistical Package for the Social Sciences) version 25.0 produced by IBM Corporation.

Thirdly, to analyze and create relationships between variables, the study uses analytical techniques: descriptive statistics, Cronbach’s alpha, EFA, CFA, and SEM.

Finally, the collected data will be analyzed by comparing and contrasting similarities and differences. Then practical conclusions are drawn in accordance with the research objectives set out.

3.3. Data Analysis Methods

After collecting data from the questionnaire survey, conducting data analysis by using the software package SPSS Statistics (Statistical Package for the Social Sciences) version 25.0 and Amos program produced by IBM Corporation is necessary. Firstly, descriptive statistics was used to survey participants’ demographic characteristics and online booking behavior. The mean score and standard deviation of each variable was calculated. Secondly, reliability analysis was taken for each factor by calculating Cronbach’s alpha Coefficient value. Thirdly, Exploratory Factor Analysis (EFA) was conducted to group observable variables into groups of factors, on the principle of ensuring module and convergence. The validity of the questionnaire was measured by Kaiser-Meyer-Olkin (KMO) Bartlett’s Test of Sphericity. Fourthly, Confirmatory Factor Analysis (CFA) was conducted to realize if the research model data was satisfactory, the scales met the requirements of a good scale, and which observed variables contributed to the model. Accordingly, the newly found results would be reviewed and evaluated on the model fit, the quality of CFA observed variables, reliability, convergence, and discriminability. Finally, Structural Equation Modeling (SEM) analysis was for finding multidimensional relationships between multiple variables in the research model. More specifically, SEM is used to test the relationship between concepts, latent variables, and observed variables. With the appropriate factor analysis technique to estimate the variance, the error was significant or not.

4. Results

4.1. Profile of Respondents

Table 1 clearly shows the demographic profiles of all respondents. The results focus on the residents in Can Tho, including males (57.8%) and females (41.7%), which illustrates both genders take interest in making online reservations. Specifically, most of the respondents are young people between 18 and 25 years old (55%) with university degrees (75.0%). In addition, the freelancer occupation makes up the majority, accounting for 32.2%, and the average income is from 5 to 10 million/ month (53.3%).

Table 1: Profile of Respondents

4.2. Behavior of Using Hotel Websites for Reservations

The popular platform that currently accounts for a high percentage of users is booking through hotel websites (95.6%), and most residents in Can Tho have travel needs and frequency from 1 to 3 times per year (79.4%) (Table 2).

Table 2: Traveling Behavior of Respondents

4.3. Testing the Reliability Coefficient Through Cronbach’s Alpha

According to the results in Table 3, the study employed Cronbach’s alpha analysis method to examine the reliability of the scale of the components in the theoretical model, as well as the correlation between the observed variables and the total variable. Which, the total variables in the model include Website functionality (WF), Website usability (WU), Website security and privacy (WSP), Affective commitment (AC), E-trust (ET), Social presence (SP), and Online hotel booking decisions (OD).

Table 3: Cronbach’s Alpha Test Results on Scales

The Cronbach’s alpha coefficients of the scales range from 0.773 to 0.857, all greater than 0.6, indicating that the association between the observed variables and the total variable is very reliable. Furthermore, the correlation coefficients of each observed variable compared to the total variable are all above 0.3 (Nunnally & Burnstein, 1994), indicating that the component variable and total variable have a high correlation, and the scale measure has a high level of dependability.

4.4. Exploratory Factor Analysis (EFA)

For the independent variables, the EFA results show that the KMO coefficient is 0.790 > 0.5, which proves that the data used for factor analysis is completely appropriate. Bartlett test results with Sig level. = 0.000 < 0.05, the test results are statistically significant. Thus, the variables are correlated with each other and satisfy the conditions of factor analysis. Factor analysis is extracted by Principal axis factoring with Promax rotation. The results show that the original 18 observed variables are grouped into 6 groups. The total value of variance extracted = 60.924% > 50%. It represents the six factors extracted in the EFA that reflect 60.924% of the variability of all included observed variables. Eigenvalues of all factors exceed 1; the sixth factor has the lowest Eigenvalues of 1,016 > 1.

For the dependent variable, the results show that the KMO index = 0.726 > 0.5, sig Bartlett’s Test = 0.000 < 0.05, so factor analysis is applicable. The analysis results show that there is a factor extracted at eigenvalue equal to 2,179 > 1. The factor can explain 63.947% of the data variation of three observed variables participating in EFA. Therefore, the EFA results are reliable and can be used for Amos analysis in the next step.

4.5. Confirmatory Factor Analysis (CFA) of Quadratic Variables

Confirmatory factor analysis (CFA) was used to determine the fit of the variables in the measurement model according to Hu and Bentler (1999). In which the P-value indicates whether the observed variables exhibit the properties of latent variables or not. When analyzing the quality of observed variables, quadratic variables become first-order variables, and partial first-order variables become observed variables. Therefore, it is suitable to use the first-order variables evaluation parameters in this case. The results of quality testing of quadratic observed variables show that all have p-values of 0.000 < 0.05, so all quadratic observed variables are significant in the model. In addition, six indicators were used to consider and the results are shown in Table 4. These indicators are suitable for evaluating the measurement model. The results show that there is a good fit between the model and the research data, there is an agreement, and the observed variables are considered highly appropriate and significant.

Table 4: CFA Measurement Model Fit Indices of Quadratic Variables

4.6. Confirmatory Factor Analysis (CFA) of the Whole Research Model

Carrying out a model fit test from the point of view of authors Hu and Bentler (1999), from the relevant indicators of the results, it is shown that the hypothesized model formed fits well with the collected data (Table 5). Additionally, the results of the quality test of observed variables show that all observed variables have a p-value of 0.000 < 0.05, so all observed variables are significant in the model.

Table 5: CFA Measurement Model Fit Indices of All Observed Variables

In conclusion, the results after CFA analysis show that the representative variables have a reciprocal relationship. The observed variables all achieve convergent value, and discriminant value and meet the requirements of value and reliability.

4.7. Structural Equation Modeling (SEM)

In particular, SEM models often offer a variety of relationships between the independent and dependent variables. We use statistical data to confirm or disprove the hypothesis theory after it has been assessed and recognized (Table 6 and Figure 2).

Table 6: Results of the Integrating Mode

Note: ***p-value < 0.001. Significant at the 0.05 level.

Figure 2: Direct and Indirect Effects on Online Hotel Booking Decisions

Notes: CMIN/DF: 1.361, GFI: 0.893, CFI: 0.960, RMSEA: 0.045 and TLI: 0.953. ***p < 0.001.

Using the 95% confidence standard, the current general trend is to look closely at the values of the 95% confidence interval (Nguyen, 2010). P (sig) of SP affecting ET is 0.182 > 0.05, SP has no effect on ET; sig of ET that affects OD is 0.083 > 0.05, variable ET has no effect on OD. The remaining variables all have sig equal to 0.000 (AMOS sign *** is sig equal to 0.000) or less than 0.05, so these relationships are all significant. Thus, there is one variable that affects ET, which is WQ. There are two variables affecting AC including WQ and ET. There is one variable that affects WF, which is WQ. There is one variable that affects WSP, which is WQ. There are three variables affecting OD including WQ, AC, and SP. Out of the eight possible hypotheses, two are unlikely: H5. E-trust has a positive effect on online hotel booking decisions, and H7. Social presence has a positive effect on e-trust. Besides that, the study shows among the two variables affecting AC, the order of decreasing variables is as follows: ET, WQ. Of the three variables affecting OD, the order of effects is descending: WQ, AC, SP.

Estimate (R-squared) for ET is 0.38 = 38%, so the independent variables affect 38% of variation in ET. Similarly, R squared of AC is 0.406 = 40.6%, so the independent variables affect 40.6% variance of AC. R squared of OD is 0.687 = 68.7%, so the independent variables affect 68.7% variance of OD.

5. Discussion and Recommendations

Accordingly, the SEM method is applied to examine and analyze data collected from residents in Can Tho to determine the factors affecting the decision-making of online booking through the hotel website.

For this purpose, a literature review-based model was developed and tested with proposed hypotheses such as website quality, affective commitment, and social presence that have a positive influence on online hotel booking decisions.

According to research results, website quality (including website functionality, website usability, and website security and privacy) is the factor that has the most considerable direct influence on customer intentions. In there, website usability has a strong impact on website quality, followed by website functionality and website security and privacy. It shows that users tend to focus on the interface and approach to using the website more, then consider the information that the hotel provides. Next, the research results proved the hypothesis that website quality is the driving force for the development of e-trust, affective commitment, and customer booking decisions (H1, H2, H3). The study also shows the importance of e-trust and affective commitment in booking decision-making (H4). Scholars also emphasize that e-trust and commitment are fundamental factors in enhancing tourists’ booking decisions (Lien et al., 2015). Affective commitment is also a factor that has a great direct influence on the booking decision of Can Tho people through hotel websites (H6). It ranks just behind website quality with a normalized regression weight of 0.322. In addition, social presence has a positive influence on a customer’s decision to book through a hotel website (H8). Customers prefer to choose a hotel website that already has a reputation and provides a sense of interaction. Based on the findings of this study, it is clear that developing characteristics such as website quality, affective commitment, and social presence is crucial to support the evolution of booking behavior through the hotel website of the residents in Can Tho.

Besides, the study also rejects two hypotheses that e-trust has a positive effect on online hotel booking decisions (H5), and social presence has a positive effect on e-trust (H7). Future research may expand the research model as consumers’ perceptions of e-trust may change over time. Accordingly, it is recommended to re-examine the changing factors and the influence of many different variables (such as presence on social networks) on the concept of trust to increase the reliability of the study.

6. Conclusion

Nowadays, to bring convenient and time-saving services to customers, many online booking channels have been created. Online booking simplifies booking operations and can help customers proactively book rooms anytime, anywhere. Realizing that, many customers have switched from direct booking to booking through online channels such as OTAs, travel agents, and hotel websites. In particular, booking through hotel websites is preferred by many people because of benefits such as direct contact with the hotel, providing honest and accurate information, a clear privacy policy, and meeting personalization needs. However, not all hotels comprehend what elements of the website can attract customers, increase usability and encourage booking on the website. To find out those influencing factors, the research paper sent questionnaires to 180 people living and working in Can Tho. Questionnaire on four factors affecting online booking decision through hotel website including website quality, affective commitment, social presence, and e-trust. Through the data collection and data processing on SPSS and AMOS, the research results have shown that there are only three factors affecting the booking decision of Can Tho residents: website quality, affective commitment, and social presence. Besides, the results also reject two hypotheses H5. E-trust has a positive effect on online hotel booking decisions and H7. Social presence has a positive effect on e-trust. Through this, hotels can take high contributing factors and develop them to improve website quality and increase booking performance.