Abstract
Introduction:
This study examines how different digital environments influence user experience and consumer behavior when selecting accommodation in tourism, comparing Booking, ChatGPT, and general internet search.
Methods:
A controlled experiment was conducted with 165 participants from Croatia, Serbia, and Bosnia and Herzegovina. User experience was assessed through pragmatic and hedonic dimensions, while decision-making outcomes included booking likelihood and perceived simplicity.
Results:
Significant differences were found between digital environments. Booking performed better than general internet search in pragmatic user experience, booking likelihood, and perceived simplicity. ChatGPT achieved results comparable to Booking but did not surpass it. No significant differences were found in accommodation price or decision time.
Discussion:
The findings show that digital environment characteristics shape tourism decision-making and user experience. Artificial intelligence demonstrates considerable potential as a decision-support tool, although specialized booking platforms still offer advantages. The study contributes to understanding the role of artificial intelligence in digital tourism and consumer behavior.
1 Introduction
Digital transformation of tourism in the last decade has significantly changed the way consumers search for, compare, and choose tourism services. The development of information and communication technologies has led to the dominance of online platforms, but also to the emergence of new tools based on artificial intelligence (AI), which are increasingly taking over the role of intermediaries between users and market offerings. Here, systems based on generative artificial intelligence stand out, such as conversational agents, which enable interactive and personalized communication with users during the decision-making process ().
In the contemporary tourism context, artificial intelligence is no longer limited to operational support but actively shapes user experience and consumer behavior. AI systems enable the analysis of large amounts of data on user preferences, search behavior, and previous decisions, thereby contributing to the development of personalized recommendations and the optimization of tourism services (). In this context, Big Data analytics plays a crucial role by enabling the collection, processing, and interpretation of large volumes of user-generated and behavioral data. These data-driven insights support the development of predictive models and personalized services, allowing tourism providers to better understand user preferences and anticipate decision-making patterns. In addition, modern AI applications, including chatbots and virtual assistants, simplify the process of searching and booking and can significantly improve user experience through faster and more relevant delivery of information. At the same time, research increasingly emphasizes the importance of understanding consumer behavior in digital environments. Previous studies have shown that artificial intelligence affects various aspects of user behavior, including trust, perception of value, and decision-making, but also that these effects are often context-dependent (). Despite the growing interest in this topic, the literature still points to the fragmentation of theoretical approaches and the lack of integrated models that would explain the relationship between technology, user experience, and consumer decisions in tourism ().
It is important to point out that most existing studies analyze artificial intelligence from the perspective of service providers or technological capabilities, while considerably fewer papers are focused on actual user behavior in specific decision-making situations. Furthermore, although the literature increasingly mentions the need to compare different digital environments, such as traditional platforms, general internet search, and AI tools, such studies are still relatively rare (). This leaves open the question of the extent to which artificial intelligence can replace or complement existing digital channels in the process of selecting tourism services. In this context, this research aims to contribute to the existing literature by analyzing consumer behavior in three different digital environments: by using a specialized accommodation booking platform, by using artificial intelligence (ChatGPT), and by using general internet search without advanced digital tools. From a theoretical perspective, this study views digital tourism decision-making as a technology-dependent process in which the characteristics of the digital environment shape user experience and subsequent decisions. Recent studies show that generative AI and conversational agents influence tourism information search and decision-making, although these mechanisms remain theoretically fragmented (; ). AI-supported decisions may depend on personalization, authenticity, emotional response, information structure, and cognitive demands (; ). Accordingly, this study distinguishes between pragmatic and hedonic user experience and experimentally compares a structured booking platform, conversational AI, and general internet search.
The focus of the research is placed on user experience and decision-making outcomes, with the aim of understanding the role of artificial intelligence in contemporary tourism ecosystems. The contribution of the paper is reflected in several aspects. Also, the study contributes to the understanding of digital innovation in tourism by examining how emerging AI-based tools interact with traditional digital platforms in shaping consumer decision-making processes. This perspective aligns with broader discussions on technological transformation and the evolving role of intelligent systems in service innovation. The research provides experimental insight into consumer behavior under controlled conditions, which is rarely represented in existing studies. The paper enables a direct comparison of different digital tools in the accommodation selection process. Also, the results contribute to a better understanding of the potential of artificial intelligence as a tool for personalization and support in decision-making in tourism, thereby directly addressing contemporary challenges of digital transformation of the tourism sector.
The value of this research lies in the experimental approach which enables a direct comparison of three different digital decision-making environments in controlled conditions such as specialized accommodation booking platforms, artificial intelligence systems, and general internet search. By doing so, the study provides insights into how digital technologies reshape interactions between users, platforms, and information sources, contributing to the understanding of evolving digital tourism systems.
2 Theoretical background and hypothesis development
2.1 Digital transformation, AI and big data in tourism
In digital tourism environments, decision-making increasingly takes place through various technological intermediaries, including specialized platforms, artificial intelligence systems, and general internet search. Each of these environments is characterized by different levels of structure, information availability, and personalization, which can significantly influence user experience and consumer behavior. In this context, understanding the differences between these approaches is essential for analyzing the role of artificial intelligence in the digital transformation of tourism. The tourism industry has undergone significant transformation with the rise of digital technologies. Initially, travel agencies provided basic information and options to travelers through static online brochures. Early digital methods for travel planning and customer service were manual, offering limited personalization. The next phase introduced online platforms with more flexible and accessible tools, enhancing user convenience and experience ().
Artificial intelligence (AI)-enabled technologies and applications have changed various aspects of travel, communication, and everyday life over the previous decade (). Tourism ranks among the world's largest industries, and its development has increasingly followed rapid technological advancement. AI is now recognized as a transformative force in tourism because it supports decision-making, service automation, operational efficiency, and personalized customer experiences, thereby fostering innovation and competitiveness (). In recent years, AI has driven innovations such as chatbots, recommender systems, service robots, and smart destination management, while also raising important questions about AI adoption frameworks and its wider impacts on consumers, businesses, and society. Advances in data mining, machine learning, and natural language processing have further supported the intelligent transformation of tourism (). By analyzing large volumes of data, AI can provide customized recommendations, improve efficiency through automated processes, support chatbots and security measures, and help tourism organizations make data-driven decisions and target their marketing strategies more effectively ().
The development of new technologies has led to remarkable digital transformations in tourism, changing tourist destinations, products, business experiences, and broader tourism systems. This transformation requires new business relationships, business models, and competencies in the tourism industry (). Technological advancement and digitalisation have reshaped the relationship between supply and demand, changed customer travel patterns, and created new forms of interaction between destinations, businesses, and visitors. At the same time, digital technologies generate large amounts of data, information, and knowledge, enabling more innovative tourism products, services, and experiences. As a result, tourists increasingly value personalized and meaningful experiences, which have become an important factor in destination choice and consumer behavior (; ). Changing interactions in the tourism market have therefore increased the importance of experience as a key factor in travel planning and destination choice, while ICT-based tourism tools can support travelers in planning and managing trips.
The rapid advancement of digital transformation technologies is reshaping tourist experiences, decision–making processes, and service interactions. Tourism platforms increasingly rely on user data to provide personalized and tailored services, while technologies such as AI, VR, AR, and smart recommender systems enhance tourist engagement and satisfaction by enabling more seamless and intelligent travel experiences (). At the same time, these technologies influence how travelers search for information, evaluate options, and make decisions, particularly through personalization and data-driven insights, which are central to contemporary digital tourism systems (; ).
2.2 Digital decision environments in tourism
Artificial intelligence is increasingly used by smart tourist destinations to develop new business models, improve marketing services, and enhance tourist experiences, sales, productivity, and operational efficiency. In this context, AI can also be understood as an intermediary and decision-support tool, as it connects tourists with relevant information, personalized recommendations, and service providers while helping destinations make more informed marketing and management decisions. However, the adoption of AI-based platforms requires significant financial investment, which may represent a challenge for destinations seeking to integrate AI into their long-term tourism development strategies (). The application of AI in tourism can be understood through three main categories: robotic process automation, cognitive insight, and cognitive engagement. These forms of AI support the automation of routine travel agency processes, improve the analysis of customer data, and enable more personalized communication with tourists. In this sense, AI acts as an intermediary and decision-support tool, connecting travelers with relevant services, simplifying decision-making, and helping tourism businesses optimize sales, operations, and customer relationship management ().
Embracing digital technologies is essential for enhancing competitiveness and ensuring the sustainable development of the tourism sector. Technologies such as AI, virtual reality, and big data can help firms improve service quality, optimize business operations, and foster innovation, thereby supporting sustainable competitive advantage (; ). In the context of accommodation search, specialized booking platforms enable structured search, filtering, and comparison of offers, thereby reducing users' cognitive effort. In contrast, artificial intelligence systems enable interactive communication and personalized recommendations based on the analysis of user data. General internet search, on the other hand, requires greater user engagement in processing information and making decisions. Online booking technologies have simplified the reservation process and changed how consumers search for, compare, and choose accommodation services. The growing use of hotel and resort applications has increased customer awareness, convenience, and satisfaction, while digital technology continues to shape booking preferences and support the growth of online reservations in tourism (; ). As travelers increasingly use online platforms for information, inspiration, and planning, digital marketing, personalized content, user-generated reviews, social media, and influencer marketing influence destination choice, strengthen engagement, and help tourism businesses adapt to changing consumer expectations (). Structured decision environments in tourism platforms help users navigate large amounts of information by organizing options through filters, rankings, reviews, and comparison tools. In this way, platforms reduce uncertainty and cognitive effort, while guiding tourists toward decisions that better match their preferences and needs ().
Conversational agents in tourism, such as AI chatbots and virtual assistants, support travelers by providing real-time information, answering questions, and helping with trip planning or booking processes. AI recommendation systems analyse user preferences, previous behavior, and contextual data to suggest destinations, accommodation, activities, or services that are more relevant to individual tourists. Through personalization, AI improves the user experience by reducing information overload and making decision-making faster, easier, and more tailored to user needs. As a result, these technologies strengthen engagement, satisfaction, and trust in digital tourism platforms (; ). In tourism, information overload during online search can make decision-making more complex, as travelers are often exposed to many destinations, accommodation options, reviews, prices, and recommendations. Therefore, digital tools that organize, filter, and personalize information play an important role in reducing uncertainty and supporting more efficient consumer decision-making (; ).
2.3 User experience and decision-making in digital tourism
In digital tourism environments, user experience is a key factor that influences users' satisfaction, trust, and willingness to continue using a particular platform or technology. Since tourists rely on digital tools when searching for information, comparing options, making reservations, and planning travel activities, the quality of their interaction with these tools can significantly shape their decision-making process. Therefore, user experience should be observed not only through technical functionality and ease of use, but also through the emotional value, engagement, and sense of enjoyment that technology creates during the travel planning process (; ). In this context, digital environments influence not only users' perceptions of experience, but also decision-making outcomes, including booking intention. Previous research suggests that structured and reliable environments increase the likelihood of deciding, while personalization can further enhance the relevance of the options presented (; ). Differences in digital environments may therefore lead to variations in both user experience dimensions and decision-making outcomes, which forms the basis for the hypotheses tested in this study. Based on these theoretical insights, differences between digital environments are expected to manifest in both user experience dimensions and decision-making outcomes.
Despite the growing literature on AI in tourism, several theoretical issues remain unresolved. Existing studies often examine AI adoption, trust, personalization, or information quality separately, while fewer studies directly compare AI-based tools with established booking platforms and conventional internet search within the same decision context (; ). Moreover, recent research shows contradictory implications: generative AI may enhance personalization, inspiration, and engagement, but it may also increase uncertainty through inaccurate or confusing information (; ). Therefore, it remains unclear whether conversational AI provides superior user experience and decision support or merely represents an alternative digital interface. This unresolved debate provides the theoretical rationale for the present experimental comparison.
2.4 Hypothesis development
Given the differences between the analyzed digital environments, user experience and decision-making outcomes are expected to vary depending on the tool used. Therefore, the following hypotheses are proposed to examine how different digital environments influence users' perceptions, experiences, and booking-related decisions.
2.4.1 H1: pragmatic quality
Structured digital environments enable efficient searching and comparison, leading to variations in the pragmatic quality of user experience across the conditions studied (; ; ). Such environments organize information through filters, rankings, and comparison tools, which reduces cognitive effort and helps users process information more efficiently. Previous research indicates that usability and system structure play a key role in improving task performance, clarity, and perceived ease of use, all of which are core elements of pragmatic user experience. In contrast, less structured environments, such as general internet search, may require greater user involvement and information processing, potentially leading to lower efficiency and increased decision complexity. The following hypothesis is proposed:
H1: Different digital environments (accommodation booking platform, artificial intelligence system, and general internet search) will result in significant differences in the pragmatic quality of user experience.
2.4.2 H2: hedonic quality
Artificial intelligence systems enable a more interactive and potentially more engaging user experience, which may influence the hedonic dimension of perception in tourism. Previous studies suggest that AI-based catboats and conversational agents can enhance tourist engagement through personalized, real-time and emotionally richer interactions, while enjoyable conversational styles may increase hedonic involvement and customer engagement in hotel booking contexts. Research also shows that personalization and hedonic motivation positively affect customer experience, supporting the idea that digital environments differ not only in functional usability but also in the level of enjoyment, stimulation, and emotional engagement they provide (; ). Through conversational interaction, personalized responses, and recommendation-based support, AI can increase user engagement, enjoyment, stimulation, and emotional involvement during the decision-making process. Since hedonic user experience in digital systems is related to pleasure, novelty, and the perceived attractiveness of interaction, it can be assumed that different digital environments will not provide the same level of experiential value (; ). Accordingly, the following hypothesis is proposed:
H2: Different digital environments (accommodation booking platform, artificial intelligence system, and general internet search) will result in significant differences in the hedonic quality of user experience.
2.4.3 H3: booking likelihood
User experience and the perceived reliability of a digital environment can influence users' final decisions, including the likelihood of booking the selected accommodation. Previous studies show that online travel platforms can shape purchase intentions through interactivity, user-friendly interfaces, personalized recommendations, and trust-building features. Trust is particularly important in online tourism contexts, as it reduces perceived uncertainty and strengthens users' willingness to complete a booking. Similarly, website quality, e-service quality, perceived usefulness, and online reviews have been found to influence booking intention by improving satisfaction, confidence, and perceived value in digital environments (; ). Accordingly, the following hypothesis is proposed:
H3: Different digital environments (accommodation booking platform, artificial intelligence system, and general internet search) will result in significant differences in the likelihood of booking the selected accommodation.
The study also includes additional research questions aimed at analyzing specific aspects of the decision-making process:
RQ1: Do digital environments differ in the perceived ease of accommodation selection?
RQ2: Do digital environments differ in the price of the selected accommodation and the time required to decide?
Based on the theoretical framework and proposed hypotheses, the conceptual research model is presented in Figure 1.
Figure 1
Figure 1 presents the theoretical framework underpinning this study. Rather than proposing a new behavioral theory, the framework integrates existing knowledge on digital tourism, user experience, and AI-assisted decision-making into a unified perspective for experimental investigation. It conceptualizes specialized booking platforms, conversational AI, and general internet search as alternative digital decision environments that differ in the way they organize information and support consumer decisions. Within this framework, digital environment is treated as the independent variable influencing two dimensions of user experience, namely pragmatic quality and hedonic quality, which are subsequently reflected in booking-related decision outcomes. The framework therefore provides theoretical rationale for the hypotheses and research questions examined in this study by linking established concepts from previous literature into a coherent structure suitable for empirical testing under controlled experimental conditions.
3 Materials and methods
3.1 Sample
A total of 165 participants took part in the study, of which 83 were male and 82 females. The average age of participants was M = 24.72 years (SD = 5.92), with a range from 19 to 54 years. Participants were collected from three countries of Southeast Europe: Bosnia and Herzegovina (n = 40), Croatia (n = 68), and Serbia (n = 55).
Participants were recruited from the target population in Croatia, Serbia, and Bosnia and Herzegovina and were allocated across the three experimental conditions while maintaining a broadly comparable demographic composition of the groups. The groups did not differ significantly with respect to age, F(2, 162) = 2.20, p =. 0114, or gender, χ2(2) = 0.41, p > 0.05. Mean age was 25.51 years (SD = 7.27) in the Booking group, 23.25 years (SD = 4.54) in the ChatGPT group, and 25.08 years (SD = 4.81) in the Internet group. These results indicate that age and gender were reasonably balanced across experimental conditions.
Participants were assigned to three experimental groups according to a between-subjects design: the group using an accommodation booking platform (Booking; n = 67), the group using artificial intelligence (ChatGPT; n = 48), and the control group using general internet search (n = 50). Although the group sizes are not completely balanced, this does not represent a methodological limitation for the applied statistical analyses.
3.2 Instruments
User experience was measured using the short version of the User Experience Questionnaire (UEQ-S), which includes two dimensions: pragmatic quality (focused on efficiency and usability) and hedonic quality (focused on experience and enjoyment of use). Confirmatory factor analysis confirmed a two-factor structure of the questionnaire (pragmatic and hedonic dimension), without a hierarchical model. Model fit indices indicate satisfactory fit after introducing two error covariances (according to modification indices between items 7 and 8 and 2 and 4): χ2(17) = 34.58, p < 0.01, CFI = 0.977, TLI = 0.962, RMSEA = 0.079 and SRMR = 0.049. The reliability of the measurement scales was satisfactory. The pragmatic dimension showed high reliability (Cronbach α = 0.90; McDonald ω = 0.91), while the hedonic dimension had acceptable reliability (α = 0.78; ω = 0.78).
In addition to UEQ-S, additional variables related to behavior and decision-making were collected:
total price of the selected accommodation (in euros)
likelihood of booking the selected accommodation (self-assessment)
perceived simplicity of the selection process
time required to make the decision (in seconds)
3.3 Procedure
The study was conducted during April 2026 in Croatia, Serbia, and Bosnia and Herzegovina as a controlled experiment with the aim of simulating a realistic situation of accommodation selection. The experiment was conducted in the presence of researchers who supervised the task implementation and ensured uniformity of experimental conditions. Participants were presented with a standardized scenario in which they were asked to imagine that they were planning a short business trip (e.g., a seminar or training) and to select accommodation that they would book. The task parameters were predefined: destination Graz (Austria), stay of two nights (Friday–Sunday), maximum budget of 300 € for two persons, and location within approximately 3 km from the city center (Hauptplatz). Participants could choose between a hotel room, private room, studio, or apartment, while hostel dormitory-type accommodations were excluded.
Before the start of the task, the researcher read standardized instructions to all participants to ensure uniformity of experimental conditions. Participants were informed about the purpose of the study, the voluntary nature of participation, and the possibility to withdraw at any time. They were also instructed not to communicate with each other during the task or use other participants' screens. Following recruitment, participants were allocated to one of the three experimental conditions before beginning the experimental task. Allocation was conducted with the aim of maintaining comparable group composition in terms of age and gender, while each participant completed only one experimental condition.
Participants were then randomly assigned to one of the three experimental groups:
Booking group (E1): participants used a specialized accommodation booking platform and were not allowed to use additional tools or sources of information. Booking.com was selected as the representative specialized accommodation booking platform because of its widespread use and familiarity among users in Croatia, Serbia, and Bosnia and Herzegovina. Its structured interface, filtering options, reviews, price comparison, and standardized accommodation information also made it suitable for comparison with conversational AI and general internet search under controlled experimental conditions.
ChatGPT group (E2): participants used an artificial intelligence system to find accommodation, without using additional tools or opening external links suggested by the tool.
Control group (K): participants used general internet search (e.g., search engines) but were not allowed to use AI tools or specialized accommodation booking platforms.
The task was performed on a computer, and participants had a maximum of 15 min to decide. After selecting accommodation, participants returned to the survey questionnaire and answered questions related to user experience and perception of the decision-making process. The experimental procedure was designed to ensure a high level of control of conditions while maintaining ecological validity, allowing a realistic simulation of consumer behavior in a digital environment. All participants provided informed consent before participating in the study, and participation was voluntary. Participants were guaranteed anonymity and confidentiality of data.
4 Results
Descriptive statistics for all outcome variables across groups are presented in Table 1. The hypotheses were tested using one-way ANOVA, with the independent variable dividing the sample into the Booking (E1), ChatGPT (E2), and control (C) groups.
Table 1
| Variable | M | SD | F(2, 162) | ω2 | Post-hoc | |
|---|---|---|---|---|---|---|
| UEQ-S Pragmatic | Booking | 1.84 | 1.07 | 5.90** | 0.056 | Booking > Interneta |
| ChatGPT | 1.62 | 1.42 | ||||
| Internet | 1.03 | 1.39 | ||||
| UEQ-S Hedonic | Booking | 0.76 | 1.27 | 2.99 | 0.024 | – |
| ChatGPT | 1.22 | 0.92 | ||||
| Internet | 0.77 | 1.03 | ||||
| Total accommodation price (in euros) | Booking | 203.70 | 62.36 | 1.00 | 0.000 | – |
| ChatGPT | 214.42 | 51.95 | ||||
| Internet | 217.68 | 51.69 | ||||
| Likelihood for booking the chosen accommodation | Booking | 5.52 | 1.16 | 4.59** | 0.042 | Booking > Internetb |
| ChatGPT | 5.15 | 1.22 | ||||
| Internet | 4.80 | 1.49 | ||||
| Simplicity | Booking | 6.40 | 0.91 | 12.91** | 0.126 | Booking > Interneta |
| ChatGPT | 5.90 | 1.45 | ||||
| Internet | 5.28 | 1.23 | ||||
| Time | Booking | 211.25 | 197.59 | 0.08 | 0.00 | – |
| ChatGPT | 212.81 | 133.63 | ||||
| Internet | 223.42 | 173.51 |
Descriptive statistics and ANOVA results.
**p < 0.01.
aGames Howell post-hoc test.
bBonferroni post-hoc test.
To examine whether digital environments differ in the perceived ease of accommodation selection, we analyzed differences between the groups on the UEQ-S Pragmatic and UEQ-S Hedonic scales. The ANOVA results indicated a statistically significant effect of moderate size, according to , on the UEQ-S Pragmatic scale, but not on the UEQ-S Hedonic scale.
The Bonferroni post hoc test showed that the Booking group achieved a statistically significantly higher score on the UEQ-S Pragmatic scale compared to the Internet group. All other comparisons were non-significant. As shown in Figure 2a, the ChatGPT group did not differ significantly from the other two groups.
Figure 2
Therefore, H1 was confirmed, whereas H2 was rejected.
Within the third hypothesis and both research questions, we examined differences between the groups in total accommodation price (in euros), likelihood of booking the selected accommodation, ease of booking, and the time required to complete the booking. A one-way ANOVA was again used to assess this set of hypotheses.
As shown in Table 2 and illustrated in Figures 2c, d, participants who used the Booking platform reported a higher likelihood of booking the accommodation they identified in the study compared to participants who searched for accommodation via an internet search. In addition, participants found the Booking platform more convenient and easier to use than those in the control group, who relied solely on internet search for the same task. Both observed effects were of moderate size, according to . No significant differences were found in accommodation price or in the time spent finding the accommodation.
Table 2
| Dependent | Independent | B | SE | t | p | LLCI | ULCI |
|---|---|---|---|---|---|---|---|
| UEQ Pragmatic [R2 =0.28, F(7, 157) = 8.59**] | Constant | 1.25 | 0.18 | 6.82 | 0.00 | 0.88 | 1.61 |
| Easiness of usage | 0.31 | 0.13 | 2.36 | 0.02 | 0.05 | 0.58 | |
| Booking vs. ChatGPT & Internet (W1) | 0.52 | 0.24 | 2.15 | 0.03 | 0.04 | 1.00 | |
| ChatGPT vs. Booking & Internet (W2) | 0.39 | 0.25 | 1.58 | 0.12 | −0.10 | 0.88 | |
| Easiness of usage X W1 | −0.18 | 0.21 | −0.89 | 0.38 | −0.59 | 0.22 | |
| Easiness of usage X W2 | 0.29 | 0.18 | 1.65 | 0.10 | −0.06 | 0.64 | |
| Gender | 0.46 | 0.18 | 2.57 | 0.01 | 0.11 | 0.81 | |
| Easiness of usage X Gender | −0.25 | 0.14 | −1.80 | 0.08 | −0.53 | 0.03 | |
| R2 | 0.28 | ||||||
| Likelihood [R2 = 0.17, F(2, 162) = 16.78**] | Constant | 4.79 | 0.15 | 31.39 | 0.00 | 4.49 | 5.09 |
| Easiness of usage | 0.23 | 0.08 | 2.82 | 0.01 | 0.07 | 0.39 | |
| UEQ Pragmatic | 0.27 | 0.08 | 3.37 | 0.00 | 0.11 | 0.42 |
Hierarchical regression analysis of Model 9.
Bold values indicate statistically significant predictors (p < 0.05). **Indicates p < 0.01.
Therefore, we confirm the third hypothesis, while both research questions are rejected; that is, no statistically significant differences between the groups were found in terms of price or time.
4.1 Additional analysis
Because previous research provides limited and inconsistent evidence on the role of demographic characteristics in AI-supported tourism decision-making, no confirmatory hypotheses were formulated for age and gender. Therefore, the following moderation and mediation analyses were conducted as exploratory analyses aimed at examining whether demographic characteristics may influence user experience and booking-related outcomes across different digital environments. Besides the main analysis conducted at the group level, we examined the potential role of age and gender in the use of ChatGPT and Booking when selecting a travel destination. Given the sample size (n = 165), these additional moderation and mediation analyses were treated as exploratory rather than confirmatory. Accordingly, their results should be interpreted cautiously, particularly for subgroup-specific effects. To improve the robustness of the mediation estimates, indirect effects were evaluated using bootstrap confidence intervals based on 5,000 resamples.
First, we investigated whether participants of different ages might differ in their preference for ChatGPT and Booking when searching for travel accommodation. To examine this issue, we conducted a hierarchical regression analysis in which membership in experimental groups was treated as a moderator variable, while age was entered as a predictor. The analysis was performed using Model 1 in Hayes' PROCESS macro for SPSS ().
Figure 3 shows that older participants tend to perceive ChatGPT as more pragmatic for use in accommodation booking. A similar, though weaker, effect is observed for Booking; however, it does not reach statistical significance. In contrast, Internet browsing is perceived as the least pragmatic method of booking accommodation as participant age increases. Thus, the perceived pragmatic quality of all three booking methods depends on participant age, with the youngest participants showing the least variation across groups, whereas differences are more pronounced among older participants in the sample. The effect is substantial, as the regression model explains 16.43% of the variance in the UEQ Pragmatic scale [R2 = 0.16, F(5, 159) = 6.25, p < 0.001], while the interaction between age and experimental group membership accounts for as much as 9.55% [ΔR2 = 0.10, F(2, 159) = 9.09, p < 0.001]. For the UEQ Hedonic scale, the model was not statistically significant [R2 = 0.07, F(5, 159) = 2.30, p = 0.05].
Figure 3
Furthermore, we examined the likelihood that participants would actually book the accommodation they identified in the study under real-life conditions (Likelihood). We started from the assumption that the ease of application use (Easiness of usage) is associated with the perceived pragmatic quality of the application (UEQ Pragmatic), which in turn may contribute to a higher likelihood of booking the selected accommodation. This mediation relationship was tested at the subsample level; specifically, we examined whether the model is consistently present for both male and female participants (Gender) across all three subgroups defined by the application used (Application). Model 9 from Hayes' PROCESS macro, used in this study, is presented in Figure 4, and the results of the analysis are shown in Tables 2, 3.
Figure 4
Table 3
| Application | B | BootSE | BootLLCI | BootULCI |
|---|---|---|---|---|
| Booking | ||||
| Males | 0.07 | 0.05 | 0.00 | 0.19 |
| Females | 0.00 | 0.04 | −0.08 | 0.07 |
| ChatGPT | ||||
| Males | 0.19 | 0.07 | 0.06 | 0.35 |
| Females | 0.13 | 0.05 | 0.04 | 0.24 |
| Internet | ||||
| Males | 0.12 | 0.07 | 0.01 | 0.28 |
| Females | 0.05 | 0.05 | −0.05 | 0.16 |
Indirect effect of UEQ Pragmatic on the relationship between Easiness of usage and likelihood to make a reservation.
Statistically significant indirect effects are bolded. B, unstandardized regression coefficient; BootSE, bootstrap standard error; BootLLCI, lower bound of the confidence interval; BootULCI, upper bound of the confidence interval.
According to Table 2, Easiness of usage is positively associated with both UEQ Pragmatic and the likelihood of making a reservation, while the UEQ Pragmatic scale is also positively associated with the likelihood of making a reservation. Additionally, Booking achieves higher scores on the UEQ Pragmatic scale compared to ChatGPT and Internet browsing. These findings refer to the model tested at the level of the full sample. To improve transparency, regression results are reported together with confidence intervals and relevant effect-size indicators, including R2 and ΔR2 where applicable.
However, when examining the model in which Easiness of usage contributes to UEQ Pragmatic, which in turn leads to a higher likelihood of booking the identified accommodation, this indirect model is not statistically significant under all conditions. Specifically, as shown in Table XY, ease of use contributes to a higher perceived pragmatic quality of the application, which in turn increases the likelihood of booking accommodation among male participants, regardless of the application used. In contrast, in the female subsample, this pattern holds only in the case of ChatGPT, while for Booking and Internet browsing the model is not statistically significant. To test the statistical significance of the indirect effect, a bootstrap method with 5,000 resamples was used, according to Table 3 and Figure 5.
Figure 5
These findings suggest that pragmatic user experience represents an important mechanism through which perceived ease of use may translate into stronger booking intentions, but this mechanism is not equally stable across all user groups and digital environments. The results therefore indicate that gender and the type of digital application should be considered when interpreting how usability-related perceptions influence consumer decision-making in digital tourism contexts.
5 Discussion
This study examined how different digital environments shape user experience and decision-making during accommodation selection. By comparing a specialized booking platform, an artificial intelligence system, and general internet search, the research offers empirical insight into the role of AI in digital tourism. The findings indicate that digital environments do not affect all aspects of user experience equally. Their influence is most visible in pragmatic quality, perceived simplicity, and booking likelihood, while hedonic experience, selected accommodation price, and decision time appear less sensitive to the type of tool used. Although several statistically significant differences were identified, the corresponding effect sizes were generally small to moderate. Therefore, the practical significance of these findings should be interpreted cautiously, as the observed differences indicate meaningful but not large effects of digital environment on user experience and booking-related outcomes. The findings should also be interpreted in light of the rapid development of generative AI technologies. Improvements in model accuracy, real-time data access, personalization, multimodal interaction, and integration with external booking systems may substantially change the comparative position of conversational AI within a relatively short period. Therefore, the present results represent the technological conditions under which the experiment was conducted rather than a permanent hierarchy between AI systems and specialized booking platforms.
The confirmation of the first hypothesis shows that pragmatic user experience differs significantly across digital environments. Participants who used the Booking platform reported higher pragmatic quality than those who relied on general internet search. This result is expected, considering that booking platforms are designed to support efficient comparison and decision-making through filters, rankings, reviews, price visibility, location information, and standardized accommodation descriptions. Such structure reduces cognitive effort and helps users complete the task with greater clarity and confidence. The intermediate position of ChatGPT can be explained by the different functions provided by conversational AI and specialized booking platforms. ChatGPT can reduce information-processing effort by offering interactive, personalized, and context-sensitive recommendations, which may explain why it performed similarly to Booking.com on several experiential and decision-related outcomes. However, Booking.com provides a more structured transactional environment through verified availability, standardized accommodation information, price transparency, user reviews, filtering tools, and direct booking functionality. These features likely strengthen pragmatic quality, perceived simplicity, and decision confidence. The findings therefore suggest that generative AI currently functions more effectively as a complementary decision-support tool than as a complete substitute for specialized booking platforms.
General internet search, by contrast, leaves more responsibility to the user, who must independently identify relevant sources, compare offers, and evaluate reliability. This interpretation is also consistent with previous empirical evidence showing that travel website design, information–task fit, and perceived service quality influence online purchase intentions, which supports the importance of structured and functionally clear digital environments in tourism decision-making (). Dedeke's study found that website design influenced perceived product quality and online purchase intentions in the travel website context. The position of ChatGPT is particularly important: it did not significantly differ from either Booking or Internet search. This suggests that AI can already provide meaningful functional support in accommodation selection, although it has not yet reached the level of specialized platforms in terms of structured task completion.
Contrary to expectations, the second hypothesis was not supported. No statistically significant differences were found in hedonic quality, even though the ChatGPT group achieved a somewhat higher mean score. This finding suggests that the interactive and conversational nature of AI does not automatically translate into a significantly more enjoyable or emotionally engaging user experience. One possible explanation is that the task was primarily practical and goal-oriented. Participants were asked to find suitable accommodation under predefined conditions, which may have directed their attention toward efficiency, reliability, and usefulness rather than enjoyment or stimulation. In this context, hedonic experience may be less important than pragmatic support, especially when users are making a decision that involves money, location, and perceived risk. Nevertheless, previous research on AI chatbots in travel tourism indicates that chatbot experience, including ease of use, information quality, security, anthropomorphism, and omnipresence, can influence satisfaction and consumer engagement, suggesting that the hedonic and relational value of AI may become more visible in contexts with stronger interactional involvement (). Magano et al. examined AI chatbot experience on travel websites and found that chatbot attributes affect satisfaction and subsequent consumer engagement behaviors.
Evidence for the third hypothesis confirms that the digital environment also affects booking-related intentions. Participants in the Booking group reported a higher likelihood of booking the selected accommodation compared with participants in the Internet group. This result highlights the importance of trust and decision confidence in digital tourism. Booking platforms do not merely provide information; they also create a decision environment in which users can immediately evaluate availability, reviews, prices, cancellation conditions, and other relevant cues. These elements likely reduce uncertainty and strengthen the feeling that the selected option is safe and appropriate. This finding is supported by research on online hotel booking showing that easy-to-process information, numerical ratings, positively framed reviews, and trust-related cues can increase booking intentions and consumer trust in hotel selection contexts (). Sparks and Browning found that positively framed review information together with numerical rating details increased both booking intentions and consumer trust. Similarly, recent research on online accommodation booking found that perceived ease of use, perceived usefulness, and tourist trust are significantly related to online booking intention, further confirming the relevance of trust and usability for booking-related decisions (). Chouykaew et al. investigated perceived ease of use, perceived usefulness, tourist trust, and online booking intention using survey data from 450 tourists. ChatGPT again occupied an intermediate position, indicating that AI can assist users in narrowing down options and making recommendations, but may still lack some features that are central to final booking confidence, such as verified availability, transparent user reviews, and direct transactional functionality.
The additional research questions further clarify the nature of these differences. Perceived simplicity differed significantly between groups, with Booking again being evaluated more positively than general internet search. This aligns with the result for pragmatic quality and confirms that users experience structured platforms as easier and more convenient. However, the absence of significant differences in accommodation price and decision time is also informative. It suggests that different digital tools may lead to similar objective outcomes under controlled conditions, even when the subjective experience of the process differs. In other words, users may find accommodation of comparable price and spend a similar amount of time searching, but the process may feel more or less simple, reliable, and efficient depending on the environment used.
Additional analyses offer a more nuanced interpretation of the main findings. Age moderated the relationship between digital environment and pragmatic quality, with differences between tools becoming more pronounced among older participants. Older users appeared to perceive ChatGPT, and to some extent Booking, as more pragmatic than general internet search. This may indicate that guided and structured digital environments are especially useful for users who benefit from clearer support when navigating complex online information. Accordingly, the age and gender findings should be viewed as exploratory boundary-condition evidence rather than as central confirmatory results of the study.
These analyses were included as an exploratory extension of the primary research objective. While the main hypotheses focused on differences between digital environments, the additional analyses examine whether these effects vary across demographic characteristics that may shape technology use and decision-making. Given the limited prior evidence in this specific context, age and gender were not included in the confirmatory hypothesis framework but were examined to identify potentially relevant boundary conditions for the observed effects.
The mediation analysis also showed that ease of use contributes to pragmatic quality, which then increases the likelihood of booking. However, this indirect effect varied depending on gender and application, suggesting that the relationship between usability, user experience, and behavioral intention is not uniform across all user groups. These findings should be interpreted as exploratory evidence that may inform the development of more specific demographic hypotheses in future research.
The findings regarding specialized booking platforms should be interpreted primarily in relation to Booking.com. Although other platforms such as Expedia, Agoda, Trip.com, or Airbnb provide comparable search and decision-support functionalities, differences in interface design, recommendation systems, review mechanisms, and accommodation supply may influence user experience and decision-making. Future studies should therefore replicate the experiment across multiple booking platforms to examine the generalisability of the present findings.
The findings suggest that digital tourism technologies should not be treated as equivalent decision-support tools. Recent research indicates that AI effects depend on information structure, personalization, authenticity, and interaction characteristics (; ; ). Research results extend this literature by showing that differences between digital environments are particularly evident for pragmatic user experience and booking-related outcomes, while hedonic effects are less pronounced.
Taken together, the results indicate that artificial intelligence has a relevant but still developing role in digital tourism decision-making. ChatGPT performed comparably to other environments in several aspects, which confirms its potential as a decision-support tool. Nevertheless, specialized booking platforms remain stronger where structure, transparency, and booking confidence are essential. The findings therefore suggest that AI should not be understood simply as a replacement for existing platforms, but rather as a complementary technology that can improve personalization, interaction, and recommendation support. Its practical value would likely increase if combined with reliable real-time data, verified reviews, price transparency, and direct booking options.
From a theoretical perspective, this study contributes to the literature by showing that the effects of digital tourism technologies depend on the dimension of user experience being observed. Pragmatic quality appears more sensitive to differences between digital environments than hedonic quality, at least in the context of accommodation selection. From a practical perspective, the findings provide several implications for different stakeholder groups within digital tourism. First, tourism platforms should priorities interface simplicity, transparent information presentation, and reliable decision-support functions, as these characteristics were associated with higher pragmatic user experience and stronger booking intentions. AI-powered conversational features may be most effective when integrated into established booking platforms rather than operating as standalone systems. Second, destination management organizations should consider AI-assisted recommendation tools as complementary communication channels that help visitors identify accommodation and destination-related services more efficiently. Integrating conversational AI into official destination websites may improve visitor support while maintaining access to verified and up-to-date tourism information. Third, AI developers should focus on improving the practical usability of conversational systems through more accurate recommendations, greater transparency regarding information sources, integration with real-time booking data, and stronger support for complete booking processes. These improvements may enhance user confidence and further strengthen the role of AI as a decision-support technology in digital tourism.
6 Conclusion
This study examined the role of different digital environments in shaping user experience and consumer decision-making in the context of accommodation selection. By comparing a specialized booking platform, an artificial intelligence system, and general internet search, the research provides experimental evidence on how emerging AI-based tools relate to established digital tourism channels. The findings show that digital environments differ primarily in their pragmatic value and in their ability to support decision confidence. Participants who used the Booking platform reported higher pragmatic quality, greater perceived simplicity, and a stronger likelihood of booking the selected accommodation compared with participants who relied on general internet search. These results confirm that structured digital platforms remain highly effective in reducing cognitive effort, organizing information, and supporting users in making booking-related decisions.
At the same time, the results indicate that artificial intelligence represents a relevant and increasingly important decision-support tool in digital tourism. Although ChatGPT did not outperform the specialized booking platform, it achieved comparable results across several outcomes and occupied an intermediate position between Booking and general internet search. This conclusion should be interpreted cautiously because ChatGPT was evaluated as a conversational decision-support tool rather than as a fully integrated booking system. Unlike Booking.com, it did not provide direct access to live availability, real-time prices, verified reservation data, or transactional functionality during the experiment. Therefore, the observed differences reflect not only the characteristics of AI-based interaction but also the broader functional asymmetry between conversational AI and a dedicated booking platform.
This suggests that AI systems can assist users through conversational interaction, personalized recommendations, and flexible information processing, but their effectiveness is still limited when compared with platforms that provide verified availability, user reviews, transparent pricing, and direct booking functionality. The absence of significant differences in hedonic quality, accommodation price, and decision time further suggests that the impact of digital environments is not uniform across all decision outcomes, but is most evident in users' perceived efficiency, simplicity, and confidence. The sample size also limits the statistical power of the exploratory moderation and moderated mediation analyses, particularly for effects estimated within smaller subgroups. Future studies should therefore replicate these relationships using larger samples specifically powered by interaction and indirect effects. Another limitation concerns the inherent asymmetry between the digital environments compared in this study. ChatGPT was examined as a conversational AI decision-support tool, whereas Booking.com represents a specialized accommodation booking platform with integrated transactional functionality. Unlike Booking.com, ChatGPT did not provide direct access to real-time availability, verified reservation systems, price updates, customer reviews, or immediate booking completion. Consequently, the comparison should not be interpreted as a direct evaluation of equivalent digital services, but rather as a comparison of alternative decision-support environments available to consumers during accommodation search. Future research should compare AI systems integrated with real-time booking infrastructure to enable a more functionally equivalent assessment. A further limitation concerns the demographic composition of the sample. Participants were predominantly younger adults, which may limit the generalizability of the findings to older tourist populations with different levels of digital literacy, technology acceptance, and online booking experience. Older users may interact differently with conversational AI, specialized booking platforms, and general internet search, potentially leading to different patterns of user experience and decision-making. Future studies should therefore include more age-diverse samples to examine whether the observed relationships remain consistent across different generations of tourists. Another limitation relates to the experimental scenario, which focused exclusively on accommodation selection for a short business trip under predefined conditions. Decision-making processes may differ in other tourism contexts, such as leisure travel, family holidays, luxury tourism, or international trips, where travelers may place greater emphasis on emotional experience, destination attributes, family needs, or travel risk. Consequently, the findings should be generalized with caution beyond the specific decision context examined in this study. Future research should replicate the experiment using different travel scenarios to determine whether the observed relationships remain consistent across various tourism contexts.
The study contributes to the literature on digital tourism, artificial intelligence, and consumer behavior by demonstrating that the effects of AI-based tools should be evaluated in relation to both user experience dimensions and behavioral outcomes. The findings also have practical implications for tourism providers, platform developers, and destination managers. Rather than treating AI as a replacement for existing booking platforms, tourism stakeholders should consider its potential as a complementary technology that can enhance personalization, interaction, and decision support when integrated with reliable data, transparent information, and trustworthy booking infrastructure. For tourism businesses, the results suggest that the greatest value may come from integrating conversational AI into established booking infrastructures rather than treating AI and booking platforms as competing alternatives.
Future research should extend these findings by including larger and more diverse samples across different age groups, countries, travel motives, and levels of digital literacy. It would also be useful to examine other tourism decisions, such as destination choice, activity planning, transport selection, or restaurant booking, where AI may play a different role. Longitudinal studies could provide deeper insight into how repeated use of AI tools influences trust, satisfaction, and actual booking behavior over time. Future research will examine generative AI integrated with real-time reservation databases and transactional systems to enable a more functionally equivalent comparison with specialized booking platforms. Future studies should replicate similar experiments periodically and across newer generations of AI systems in order to assess whether technological progress changes user experience and booking-related outcomes over time.
Finally, future studies should compare different types of AI systems, including integrated travel assistants, platform-based recommendation engines, and generative conversational agents connected to real-time booking data, in order to better understand the conditions under which artificial intelligence can create superior value in digital tourism ecosystems.
Statements
Data availability statement
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.
Author contributions
MŠ: Writing – review & editing. BA: Writing – original draft. VR: Writing – review & editing.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This research was funded by the Croatian Science Foundation (HRZZ) research project and Tourism 5.0: Customer Satisfaction through the Digital Transformation of Tourist Destinations, grant no. IP-2025-02-1899 (Croatian Science Foundation).
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1
AbashidzeI. (2024). The influence of online platforms on decision-making process and behavioural traits of international travelers. Eur. Sci. J. ESJ20:51. doi: 10.19044/esj.2024.v20n37p51
2
AlharbiA.PanditA.WilkV.Rosenberger IIIP. J.MiahS. (2026). Artificial intelligence-enabled conversational agents in tourism and hospitality: a systematic literature review and future research directions. Asia Pac. J. Tour. Res.31, 21–52. doi: 10.1080/10941665.2025.2555204
3
AlmeidaF.AlmeidaJ.MotaM. (2019). Perceptions and trends of booking online payments in tourism. J. Tour. Serv.10, 1–15 doi: 10.29036/jots.v10i18.39
4
AzharN. B. A.ZahariM. S. M.FerdianF.HanafiahM. H. (2025). Unpacking the power of trust: how relative advantage, compatibility, ease of use and usefulness drive hotel self-directed bookings. Consum. Behav. Tour. Hosp.20, 148–163. doi: 10.1108/CBTH-05-2024-0176
5
BakerJ.NamK.DuttC. S. (2023). A user experience perspective on heritage tourism in the metaverse: Empirical evidence and design dilemmas for VR. Inf. Technol. Tour.25, 265–306. doi: 10.1007/s40558-023-00256-x
6
BalakrishnanJ.DwivediY. K.MalikF. T.BaabdullahA. M. (2023). Role of smart tourism technology in heritage tourism development. J. Sustain. Tour. 31, 2506–2525. doi: 10.1080/09669582.2021.1995398
7
BhattacharjeeD. R.KuanrA.PradhanD.MoharanaT. R. (2025). Fun or warm: How conversational style boosts customer engagement. J. Retail. Consum. Serv.85:104293. doi: 10.1016/j.jretconser.2025.104293
8
ChakrabortyU.BiswalS. K. (2026). Smart tourism systems: transforming travel through mixed-methods insights. J. Hosp. Tour. Insights9, 728–746. doi: 10.1108/JHTI-02-2025-0238
9
ChenT. (2024). Digital Transformation and Information Management in the Tourism Industry. Tourism Management and Technology Economy.
10
ChouykaewT.KimL.IssayevaG. (2024). How perceived ease of use, trust, and perceived usefulness influence tourists' decision to book homestay services online. GeoJournal Tour. Geosites56, 1609–1616. doi: 10.30892/gtg.56417-1331
11
ChristensenJ.HansenJ. M.WilsonP. (2024). Understanding the role and impact of generative artificial intelligence (AI) hallucination within consumers' tourism decision-making processes. Curr. Issues Tour.28, 545–560. doi: 10.1080/13683500.2023.2300032
12
CuomoM. T.TortoraD.ForoudiP.GiordanoA.FestaG.MetalloG. (2021). Digital transformation and tourist experience co-design: big social data for planning cultural tourism. Technol. Forecast. Soc. Change162:120345. doi: 10.1016/j.techfore.2020.120345
13
DaryananiS. N. D.GonzálezC. S. G.RalloA. R. (2026). A user experience–based evaluation model for AI-enabled digital booking engines in tourism. Int. J. Mark. Commun. New Media. 230–252. doi: 10.54663/2182-9306.2026.specialissuecmth.230-252
14
DedekeA. N. (2016). Travel web-site design: Information task-fit, service quality and purchase intention. Tour Manag54, 541–554. doi: 10.1016/j.tourman.2016.01.001
15
DhaigudeA. S.KamathG. B. (2025). Mapping responsible artificial intelligence in business and management: trends, influence, and emerging research directions. J. Open Innov. Technol. Mark. Complex.11:100640. doi: 10.1016/j.joitmc.2025.100640
16
DoganS.NiyetI. Z. (2024). Artificial Intelligence (AI) in Tourism. Leeds: Emerald Publishing. doi: 10.1108/978-1-83753-970-320241001
17
ErdosF.ThinakaranR.FiruzaB.KoloszárL. (2025). The rise of ai in tourism - a systematic literature review. GeoJournal Tour. Geosites60, 1254–1265. doi: 10.30892/gtg.602spl22-1498
18
FergusonC. J. (2009). An effect size primer: A guide for clinicians and researchers. Prof. Psychol.: Res. Pract.40, 532–538. doi: 10.1037/a0015808
19
Florido-BenítezL.del Alcázar MartínezB. (2024). How artificial intelligence (AI) is powering new tourism marketing and the future agenda for smart tourist destinations. Electronics13:4151. doi: 10.3390/electronics13214151
20
GursoyD.SfoderaF.PiccioniN.NosiC.ChiC. G.ChiO. H. (2025). Consumers' willingness to use the metaverse for information search: an investigation of the underlying mechanism and critical determinants. Int. J. Hosp. Manag. 124:103957. doi: 10.1016/j.ijhm.2024.103957
21
HadinejadA.NoghanN.Moyle BD. (2021). Future research on visitors' attitudes to tourism destinations. Tour Manag83:104215. doi: 10.1016/j.tourman.2020.104215
22
HakimI. N.AntariksaB.HamidahS.RusataT. Faisyal. (2025). “Measuring the user experience toward ChatGPT for sustainable tourism purposes,” in Handbook of Sustainability in Tourism and Hospitality in Indonesia, eds. P. D. S. Pitanatri and A. Hassan (Singapore: Springer Nature Singapore), 1181−1208. doi: 10.1007/978-981-96-3379-1_43
23
HaysW. L. (1963). Statistics for Psychologists.New York, NY: Holt, Rinehart and Winston.
24
HinaM.IslamN.LuoX. R. (2025). Towards sustainable consumption decision-making: Examining the interplay of blockchain transparency and information-seeking in reducing product uncertainty. Decis. Support Syst.189:114370. doi: 10.1016/j.dss.2024.114370
25
KaurA.SinghJ.KaurS. (2025). “Enhancing tourism website usability via automated web parsing and user feedback analytics,” in Proceedings of the Indian National Science Academy, 1–28. doi: 10.1007/s43538-025-00609-6
26
KhoshrooM.SoltaniM. (2025). Digital transformation of tourism: towards a model of technology acceptance by tourists in the Industry 5.0. Eur. J. Innov. Manag.28, 2101–2124. doi: 10.1108/EJIM-11-2023-1018
27
KleinA.SharmaV. M. (2022). Consumer decision-making styles, involvement, and the intention to participate in online group buying. J. Retail. Consum. Serv.64:102808 doi: 10.1016/j.jretconser.2021.102808
28
KonstantinovaS. (2019). Digital transformation in tourism. Knowl. Int. J.35, 188–193.
29
KrečkováR.ŠálkováD.ProcházkováR.TrnkaR. (2025). determinants of accommodation choice on digital platforms: price, cleanliness, and trust. J. Tour. Serv.16, 173–194. doi: 10.29036/s08nkw26
30
MaX.ChenX.YuanC. (2025). “Crossroads of AI and tourism: Enhancing destination management and traveler engagement,” in Proceedings of the 2025 2nd international conference on generative artificial intelligence and information security (New York, NY), 31–36. doi: 10.1145/3728725.3728731
31
MaganoJ.QuintelaJ. A.BanerjeeN. (2025). Driving consumer engagement through AI chatbot experience: the mediating role of satisfaction across generational cohorts and gender in travel tourism. Sustainability17:7673. doi: 10.3390/su17177673
32
MickM.KovaleskiJ.ChiroliD. (2024). Sustainable digital transformation roadmaps for SMEs: a systematic literature review. Sustainability.16:8551. doi: 10.3390/su16198551
33
MishraD.DasS.PatnaikR. (2024). Application of AI technology for the development of destination tourism towards an intelligent information system. Econ. Aff.69, 1083–1095. doi: 10.46852/0424-2513.3.2024.31
34
OECD (2024). Artificial Intelligence and Tourism: G7/OECD Policy Paper (OECD Tourism Papers, 2024/02).Paris: OECD Publishing, Paris. doi: 10.1787/3f9a4d8d-en
35
OuaddiC.BenaddiL.SouhaA.AbdelmalekH.JakimiA.SaadaneR. (2025). A model-driven approach employing dsl and generation templates to accelerate the development of conversational agents for smart tourism. IEEE Access. 28:e02649. doi: 10.1109/ACCESS.2025.3574641
36
PrasannaA.PushparajP.KushwahaB. P. (2025). Conversational AI in tourism: a systematic literature review using TCM and ADO framework. J. Hosp. Tour. Manag.64:101310. doi: 10.1016/j.jhtm.2025.101310
37
RatherR. A.RasulT.LadeiraW. J.de Oliveira SantiniF.RamkissoonH. (2026). Artificial intelligence (AI) in tourism, hospitality, and service research: a new digital frontier for consumer engagement, psychology, and behavior. J. Hosp. Mark. Manag. 1–18. doi: 10.1080/19368623.2026.2616769
38
SparksB. A.BrowningV. (2011). The impact of online reviews on hotel booking intentions and perception of trust. Tour Manag32, 1310–1323. doi: 10.1016/j.tourman.2010.12.011
39
SwadhiR.GayathriK.DimriS.BalakrishnanA.JyothiP. (2025). “Role of digital marketing in shaping travel decisions: consumer behavior in tourism,” in Intersections of Niche Tourism and Marketing, eds. B. B. Sousa and V. R. Santos (London: IGI Global Scientific Publishing), 153–176. doi: 10.4018/979-8-3693-8417-6.ch007
40
TangL.WangX.KimE. (2022). Predicting conversion rates in online hotel bookings with customer reviews,J. Theor. Appl. Electron. Commer. Res.17, 1264–1278. doi: 10.3390/jtaer17040064
41
TyrväinenO.KarjaluotoH.SaarijärviH. (2020). Personalization and hedonic motivation in creating customer experiences and loyalty in omnichannel retail. J. Retail. Consum. Serv.57:102233 doi: 10.1016/j.jretconser.2020.102233
42
VinodB. (2024). “The impact of emerging technologies on intermediaries,” in Mastering the Travel Intermediaries: Origins and Future of Global Distribution Systems, Travel Management Companies, and Online Travel Agencies (Cham: Springer Nature Switzerland), 363–394. doi: 10.1007/978-3-031-51524-8_13
43
WangS.WangQ.CuiQ.LanT. (2025). Artificial intelligence in tourism: a systematic literature review and future research agenda. Sustainability17:9080. doi: 10.3390/su17209080
44
YangB.LiuY.ChenW. (2023). A twin data-driven approach for user-experience based design innovation. Int. J. Inf. Manage.68:102595. doi: 10.1016/j.ijinfomgt.2022.102595
45
YuJ.MengT. (2026). Image generative AI in tourism: trends, impacts, and future research directions. J. Hosp. Tour. Res.50, 288–301. doi: 10.1177/10963480251324676
46
ZhangM.YiF.GursoyD. (2026). The effects of generative artificial intelligence on consumers in hospitality and tourism: a systematic review and future research directions. Int. J. Hosp. Manag.133:104452. doi: 10.1016/j.ijhm.2025.104452
47
ZhangY.Papp-VáryÁ.SzabóZ. (2025). Global influences of digital transformation on behavioral factors in tourism: a systematic literature review. Cogent Bus. Manag.12:2536101. doi: 10.1080/23311975.2025.2536101
48
ZhangZ.LiC.ZhangH. (2026). Digital confusion: Comprehending the impact mechanisms of artificial intelligence-generated content and user-generated content on tourism decision making. Tour Manag112:105269. doi: 10.1016/j.tourman.2025.105269
49
ZhuS.SongM.DuanY. (2025). Emotional arousal: how artificial intelligence-generated content influences tourism decision-making. Tour. Recreat. Res.50:1639–1650 doi: 10.1080/02508281.2024.2439321
Summary
Keywords
artificial intelligence, consumer behavior, decision-making, digital tourism, personalization, user experience
Citation
Šostar M, Andrlić B and Ristanović V (2026) Artificial intelligence in digital tourism: evidence from an experimental study of consumer decision-making. Front. Sustain. Tour. 5:1924194. doi: 10.3389/frsut.2026.1924194
Received
30 June 2026
Revised
28 July 2026
Accepted
04 August 2026
Published
26 August 2026
Volume
5 - 2026
Edited by
Xiling Xiong, Southwestern University of Finance and Economics, China
Reviewed by
Lewis T. O. Cheung, York St John University, United Kingdom
Beenish Shameem, City University College of Ajman, United Arab Emirates
Updates
Copyright
© 2026 Šostar, Andrlić and Ristanović.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Berislav Andrlić, bandrlic@ftrr.hr
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.