The determinants of the usefulness of online reviews in the tourist offer selection

: The dynamics of the development of tourism sector is reflected, among other things, in the fact that an increasing number of tourists choose an offer based on online reviews. Although it is true that due to the intensive development of communication via the Internet, online reviews are one of the dominant sources of information, their level of perceived usefulness may differ. The aim of the research is to examine the extent to which textual comments, photos and ratings of tourist offers determine the usefulness of online reviews. In order to analyze the results, we used exploratory factor analysis. Moreover, by using regression analysis, it was confirmed that usefulness affects the level of trust in online reviews, as well as that trust has implications for the intention to purchase a tourist offer. Finally, we came to conclusions that may have numerous implications for the decisions of marketing managers in the field of tourism.


Introduction
The Internet has drastically affected the way of exchanging information regarding a tourist offer. Unlike traditional approach, where users" attitudes about a tourist offer were formed mainly under the influence of oral communication, in digitalized environment, communication is carried out through various forms, which are all based on the use of web content (Gellerstedt & Arvemo, 2019;Martins et al., 2018). Hence, reviews of other tourists are one of the most important sources of information when it comes to choosing a tourist offer (El-Said, 2020; Filieri & Mc Leay, 2013;Guo et al., 2021;Zhao et al., 2015). The fact that in the era of digitalization, a large amount of information is available to tourists additionally emphasizes the question of whether all user-generated content is equally useful (Filieri, 2015;Zhao et al., 2015). Research on the usefulness of online reviews in the field of tourism, in accordance with textual, numerical and visual indicators, is not sufficiently represented in domestic literature.
The aim of this paper is to get relevant knowledge about the extent to which the textual content, rating and visual context of online reviews influence their usefulness. Another goal is to examine whether the usefulness of online reviews has an impact on trust towards a tourist offer, as well as whether that trust implies purchase intention.
The paper is structured in the following way. The theoretical overview refers to basic concepts, such as electronic interpersonal communication and online reviews, as observed in the field of tourism. Furthermore, we analyze the importance of studying the usefulness of online reviews and trust in online environment, in order to empirically examine numerous relationships through an integrative research framework. As for the qualitative research methods, we applied the content analysis method, deductive method, as well as analysis and synthesis. In order to test the hypotheses, a quantitative methodology was used, specifically, reliability analysis, exploratory factor analysis and regression analysis. After elaborating the key results, concluding remarks were given, and within them the most significant implications, limitations and directions of future research were presented.
Online reviews of other users are one of the fundamental forms of electronic interpersonal communication (Filieri, 2015 (Filieri, 2016), potential or current tourists gain insight into additional information that can be significant for final decision (Filieri, 2015;Hlee et al., 2018). The users of the tourist offer very often perceive the information provided by other tourists as more honest and reliable compared to those originating from business entities (Borisavljević, 2021;Zhao et al., 2015). For this reason, it is very important to examine the various determinants of the usefulness of online reviews and their role in the process of creating tourist trust (Khare et al., 2020;Zelenka et al., 2021). As stated by Hlee et al. (2018), online reviews will reflect the potential demand for tourist products, which is one of the basic reasons for conducting this research.  (Filieri et al., 2018). This is particularly emphasized when collecting information related to destinations, accommodation facilities, restaurants, attractions, and the overall tourist offer (Hlee et al., 2018). In this paper, the importance of online reviews will be examined in relation to the tourist offer in general, without highlighting the differences in terms of its specific features.
Although there are numerous papers in the field of tourism which aim to examine online reviews (Chan et al., 2017;Rianthong et al., 2016), only few of them are based on the key question that precedes the intention to purchase a tourist offer: How useful are online reviews? The time and effort that tourism users invest in analyzing online reviews has further contributed to examine key characteristics when analyzing the content of online reviews (Jiménez & Mendoza, 2013). Although the usefulness of online reviews may depend on those who in fact leave the reviews ( The diversity of determinants that establish usefulness can be reduced if we observe some of the previous research that linked the question of the usefulness of online reviews to the textual, numerical or pictorial contents of the reviews (Filieri et al., 2018). When examining the influence of online reviews on the intention to book a hotel offer, Chan et al. (2017) state that the crucial components of online reviews are rating in the form of numbers and descriptive comments. One of the studies that included over two hundred thousand reviews was aimed at a comparative analysis of the relationship between textual comments and numerical ratings (Alantari et al., 2022). Equivalent to the conclusions of this study, in which we applied a method of machine learning by which the usefulness of online reviews of hotel offers was covered, it was observed that the rating is related to some of the keywords given in the form of textual or descriptive reviews. Based on the classification provided by Le et al. (2022), online reviews include textual comments, ratings in the form of numbers, i.e. stars, and photos of other users. Bearing in mind that most websites base their offer on a combination of textual and visual content (Kim et al., 2021), examining the influence of that content in choosing a tourist offer can provide significant scientific and practical implications.

Textual comments within online reviews
Textual comments can be very significant in the process of profiling certain offer, determining the rate of conversation (Ludwig et al., 2013) and determining the purchase intention (Li et al., 2019;Wu et al., 2021). When writing textual comments about a particular offer, users not only describe its features, but also express emotions and emphasize their opinions and attitudes as well (Kim et al., 2022). The qualitative content of online reviews allows a more detailed insight into the experience of other users in relation to the tourist offer and is often indicator of their quality (Ludwig et al., 2013). According to previous research, it has been determined that textual features can be key determinants of the usefulness of reviews (Wu et al., 2021). Textual comments within online reviews refer to opinion, the value of reading, comprehensiveness and quality of reviews (Le et al., 2022). Some authors go a step further, where the sentimental aspects of the reviews are analyzed more thoroughly within the textual comments (Ristova, 2020), which are brought into relation with sales (Li et al., 2019).

The significance of photos within online reviews
The impressions of potential or current tourists about a certain tourist destination will be more complete if user reviews, in addition to textual and numerical data, contain photos as well. Since it is not possible to test a tourist offer prior to using it, the photos that other travelers leave have a significant impact on the perception of the level of quality ( 2021) found an association between online reviews containing photos and intentions to purchase a travel product.

Rating as a form of online reviews
Information overload in online environment often results in tourists basing their choice on the rating of a particular offer rather than on textual comments (Bigné et al., 2020). A tourist offer can be expressed numerically, in the form of certain numbers, in the form of stars or graphic representations that categorize the given offer, i.e. indicate its rating (Filieri,  Research has confirmed that tourists expect a high quality of service and better interaction with users from a hotel offer that is highly categorized, compared to those tourist offers that are rated low, where the priority is cleanliness, safety and security of tourists (Kim et al., 2021). Some studies have confirmed that when choosing a tourist offer, extreme rating values (one or five stars) are more useful compared to offers categorized with three stars (Liu & Park, 2015;Park & Nicolau, 2015). Rianthong et al. (2016) point out that the rating of the hotel offer has positive influence on both the number of reservations and the profit generated on that basis. In previous research, it was found that when examining purchase intention, it is important to observe the congruence between the rating and the textual comments of the reviews (Jiménez & Mendoza, 2013), and that this type of relationship contributes to increasing the value of online reviews (Wu et al., 2021).

The usefulness of online reviews
When potential tourists consider online reviews, they create certain beliefs that are actually created as a consequence of the extent to which such reviews are useful, informative and upto-date (Sparks et al., 2013). In this sense, the usefulness of reviews is one of the crucial prerequisites for online reviews to influence behavioral intentions ( . Higher usefulness of online reviews is often one of the potential sources for providing additional value to consumers, especially in the area of tourism (Liu & Park, 2015). By examining the usefulness of online reviews when assessing websites for booking tourist offers, Liu and Park (2015) have analyzed quantitative and qualitative characteristics. The conclusions made in the paper written by Wu et al. (2021) imply that the interaction between content and rating of reviews implies more usefulness, i.e. it creates higher value for consumers. In accordance with the available literature, another variable whose effect on perceived usefulness will be observed is visual content. Similar to previous theoretical conclusions, we defined the following hypotheses in the paper: H1: Textual comments within online reviews have a positive and statistically significant effect on the usefulness of those reviews related to a tourist offer.
H2: Photos within online reviews have a positive and statistically significant effect on the usefulness of those reviews related to a tourist offer.
H3: Rating within online reviews has a positive and statistically significant effect on the usefulness of those reviews related to a tourist offer.

Trust towards online reviews
A large number of online reviews does not always imply that those reviews will evoke trust in potential tourists. The matter of trust in the online environment is particularly significant if considering the pronounced risk and insecurity due to the inability to have insight into the tourist offer itself (Ladhari & Michaud, 2015). In that sense, trust towards online reviews is understood as truthful, non-commercial opinion of those who have experience with a certain product or service (Filieri, 2016, p. 48). Users" trust in the area of tourism is one of the phenomena that is often studied (Filieri,  For that reason, in order to identify false reviews, we applied textual analysis, as well as numerous mathematical and statistical methods (Zelenka et al., 2021). Useful online reviews tend to result in a higher level of trust (Liu & Park, 2015). However, few studies aim to establish interdependence between the usefulness of online reviews and trust towards those reviews. In accordance with that, we defined the following hypotheses in the paper: H4: The usefulness of online reviews has a positive and statistically significant effect on the trust towards those reviews related to a tourist offer.
H5: Trust towards online reviews has a positive and statistically significant effect on the intention to purchase a tourist product.

Methodology
In order to conduct the empirical research, we used the survey method, and we used a questionnaire as an instrument to gather data, which we distributed electronically in the period from April to May 2022. The respodents were contacted by sending the online survey link to their social media accounts. The survey included only those respodents who were interested in reading online reviews before making the reservation of a tourist offer. The respondents expressed the degree to which they agreed with the statements in the questionnaire using a five-degree Likert scale. The variables used in the questionnaire were measured by statements that were adjusted according to the aim of the research. Statements related to textual comments within online reviews were adjusted based on the paper by Wu et al. (2021). Contextual photos were already examined by statements in the paper by Le et al. (2022). The significance of rating as a means of expressing online reviews was examined by statements in papers by Filieri (2015) and Le et al. (2022). The usefulness of online reviews was measured based on papers by Zhao et al. (2015). The level of online trust towards reviews was measured by statements that were adjusted based on the paper by Su et al. (2021), while the intention to purchase a tourist product was measured by statements based on the paper by El-Said (2020).
The analyses needed to make valid conclusions were made in the statistical software for social sciences SPSS, version 21. We interpreted the results of the research based on the results of the explorative factor analysis, reliability analysis and regression analysis. The research model is presented in Figure 1.

Research results
As stated, we used explorative factor analysis to group statements into factors. In order to test whether using this analysis is justified, we showed Bartlett"s Test of Sphericity and Kaiser-Meyer-Olkin -KMO measure of sampling adequacy ( Table 1). The value of the KMO measure is 0.892, which is significantly higher than the recommended limit value of 0.6 (Coakes, 2013). Bartlett"s Test of Sphericity has a statistically significant value (Sig=0.000), which shows that there is a statistically significant correlation between the original variables. .000 Source: Authors" research By using Principal Component Analysis, and after the factor rotation (varimax), three factors were singled out. The first factor pertains to textual comments; the second factor pertains to online reviews expressed by photos, while the third factor pertains to the rating of a tourist offer. The given matrix is interpretable, and only those statements whose values of absolute factor weights are above 0.5 were kept.  Table 3 includes the obtained factors, the values of factor weights in each factor, as well as the value of the Cronbach"s alpha coefficient. The high value of the Cronbach"s alpha coefficientabove 0.7 (Nunnally, 1978) indicates that the factors have good internal consistency. When deciding on a certain number of factors, it is important to point out that we kept the factors whose characteristic value was higher than 1, which was the case with the first three factors (it was 5.377 for the first factor, 1.798 for the second factor and 1.006 for the third factor). The total percentage of variability that these three factors explain is 62.93%, which is in line with recommendations given by Tabachnick and Fidell (2013). The rating of the tourist offer within online reviews enables me to learn more about the offer. 0.829 The rating of the tourist offer within online reviews enables me to understand the quality of the offer. 0.805 When searching for tourist offers, I focus on those that have a high rating expressed in online reviews. 0.678

Source: Authors" research
In order to examine the shared effect of the observed factors on the usefulness of online reviews, we applied multiple regression analysis (Table 4). The model is representative (Sig=0.000); the value of the Adjusted Coefficient of Determination equals 0.623, which shows that 62.3% of the varability of the usefulness of online reviews is explained by the defined factors.
The value of the growth factor variance which is lower than five for all defined factors confirms that the problem of multicollinearity does not exist in the regression model (Table  5). When observing them separately, textual reviews contribute most to the usefulness of online reviews (ß=0.609, p=0.000). Consequently, it can be concluded that the effect of photos within online reviews on the usefulness of online reviews is pronounced and very strong (ß=0.484, p=0.000). Rating within online reviews also has a positive and statistically significant effect on the usefulness of those reviews, but its effect has lower intensity (ß=0.152, p=0.000). We thereby conclude that hypotheses H1, H2 and H3 can be confirmed.  (Table 6), we examined whether the usefulness of online reviews has any implications on the level of trust towards those reviews, as well as whether the level of trust affects the intention to purchase a tourist product. We can conclude that 47.4% of the variability of trust towards online reviews as the dependent variable is explained by the usefulness of online reviews, and that 43.9% of the variability of the intention to purchase a tourist product online as the dependent variable is explained by online trust. Therefore, based on the above, hypotheses H4 and H5 can be confirmed.

Conclusion
The evident complexity of the tourist offer, the existence of a large number of criteria which tourists consider when making their choice, as well as the omnipresent insecurity in the online environment emphasize the significance of empirical analysis in this paper. In accordance with the conclusions of the empirical research, marketing managers should bear in mind that various forms of online reviews have certain implications on the perceived usefulness of online reviews. By adopting a proactive approach and by improving the exchange of online reviews, this form of communication can be used as an effective marketing means.
Although, this study represents a valuable resource for the tourism sector, its limitations should be explained. The limitation reffering to the representativeness of the sample could be overcome by enlarging the number of respodents in future research. This study examines the intention to purchase a tourist product in general, without specifying its components. In addition, most of conclusions are based on the analysis of online reviews, and future papers could also examine the significance of expertise, competence and credibility of the tourists who post reviews. Moreover, the research examined the significance of the online reviews regardless of the platform where the reviews were posted. Some future research should develop an integrative research framework that would identify the similarity of online reviews on different platforms, whether on websites of hotels and tourist agencies, or reviews on social media websites.
One more variable that could be included is the country of origin of tourists that post comments about a tourist offer. This could help get a more complete insight into additional factors that characterize the choice of the tourist offer made by Serbian tourists. Furthermore, the details contained in online reviews, as well as the ratio of positive to negative reviews, could also be a useful marketing tool in the area of promoting and selling a tourist offer. The paper is based on the analysis of the quality of the content of online reviews in terms of their rating, photos and textual comments. The quantity of such content could be examined based on related research in order to determine whether perceived usefulness changes if there are more reviews posted by other tourists in the online environment. The paper examines the direct link between textual content, rating and photos within online reviews and their usefulness. Some future papers could also examine the effect of the moderator, for instance, in order to determine whether some extreme rating supported by photos of a tourist offer provides higher usefulness for tourists.
Although the role of online reviews is more than obvious with the development of the Internet, it would also be desirable to perform a comparison of the degree to which tourists in our country give advantage to online reviews compared to recommendations received from family and friends, i.e. by traditional word-of-mouth communication. Research has shown that a higher level of usefulness of online reviews for tourists is also a potential source of trust, which later results in the purchase intention. Accordingly, a systematic approach and constant analysis of online reviews should be the imperative in order for business entities in the area of tourism to be able to do business successfully.