Technology and Innovation: Analyzing the Heterogeneity of the Hotel Guests’ Behavior

Abstract

The study intends to identify and analyze different consumer segments. For this purpose, we examine why customers turn to electronic word-of-mouth (eWOM) before making a purchase and how they perceive a hotel’s information and communication technology (ICT) and relational innovation after making a purchase. The objective was empirically tested with data from a panel of consumers who stayed at hotels during the post-pandemic recovery period in Spain. In total, 393 valid questionnaires were obtained. The estimation of a finite mix model was applied to identify guest profiles. Estimation identified three guest profiles where the perceptions of the hotel’s relational innovation and ICT present a high discriminant power in the first two segments. Moreover, compared to the second segment, the first group is characterized by the low impact level of these variables. On the other hand, the motivation to consult eWOM in the prebooking stage significantly influences all three groups; however, the guests of the third segment present less motivation than the rest. Hotel managers may consider ICT, relational innovation, and eWOM factors when segmenting consumers. Understanding this would enhance the company’s service delivery and the hotel’s competitiveness. The contribution of this study lies in considering ICT, relational innovation, and eWOM as novel factors that help identify different guest profiles.

1. Introduction

In recent years, intense competition within the hospitality sector has increased awareness about the importance of a more accurate understanding of key tourists’ needs and desires to predict different behavior patterns. In this line, marketing segmentation as a tool has gained broad acceptance when it comes to understanding tourists and their specific needs [1]. Recent studies continued to call attention to the heterogeneity in the perceptions and behavior of tourists, e.g., [2,3], especially in the hospitality industry [4].

There is still an open debate regarding the most effective criteria for market segmentation that is highly important to all businesses [4,5,6], particularly to the hospitality market [7,8]. Indeed, customer segmentation has evolved into a central concept in marketing [1], with numerous companies employing segmentation strategies to enhance customer satisfaction. The present study focuses on lodging services, such as those in the hospitality industries, in which the success of the company–customer relationship is closely tied to an understanding of consumer profile [2,3]. Analyzing potential heterogeneity in customer perceptions is particularly important for such services’ segmentation efforts.

Recent studies have still focused more on observed, e.g., [9,10,11], rather than delving into the subtleties of unobserved heterogeneity. Yet, scant attention has been given to exploring how distinctive latent variables are able to impact the segmentation process. This proposal aims to advance in this line, analyzing in detail the usefulness of customers’ motivations to consult eWOM about the hotel (prebooking phase) and their perception of the implementation of information and communication technologies and relational innovation policies developed by service providers. These three variables have been singled out as significant for modeling consumer behavior [12]. Hence, the challenge lies in examining how these factors can effectively segment the market. Drawing from insights on the relevance of ICT, relational innovation, and the motivation behind consulting eWOM throughout the customer journey in tourism, it is plausible to consider these variables for possible market segmentation purposes. This highlights the novelty and potential importance of this topic, not only for refining variable conceptualization but also for generating both theoretical and practical implications for segmentation within the tourism and hospitality domains. Consequently, faced with these open research gaps, this study attempts to identify the unobserved heterogeneity of tourists regarding their motivations to consult eWOM before booking the hotel stay, perception of company ICT facilities, and relational innovation perceptions during the staying period for the purpose of customer segmentation. More specifically, this study seeks to answer the following research question: How can unobserved heterogeneity in terms of customer perceptions be effectively applied to identify clients’ heterogeneity in the hotel industry?

On one hand, the consumer purchasing journey has evolved drastically during the last decade, especially in the last three years. The global crisis and pandemic rushed digitalization, the expanse of the Internet, and the development of ICT. This situation provoked unreversible changes in how clients look for prebooking information, share the hotel stay experience with others [7,13,14], and relate with the company [15,16]. In the prepurchase stage in the lodging industry, clients search for information to reduce the potential post-booking risks, considering eWOM as a source facilitating future decision making [4,7,13,17,18]. In this sense, eWOM represents a digital tool for consumer-generated and consumer-related communication about the brand crucial to companies [8]. The hospitality industry is, indeed, one of the biggest providers of consumer-generated communication thanks to modern platforms such as TripAdvisor, Trivago, Facebook, and Booking. In this study, we focus on customers’ motives to consult eWOM during the prepurchase stage to better understand the consumer decision-making process on the prepurchase stage.

On the other hand, faced with the challenges after the global crisis caused by COVID-19, many companies accelerated the innovation processes on all levels in order to successfully adapt to the new environment and develop lasting competitive advantages, i.e., [16,19]. Furthermore, the hospitality industry’s service nature led hotel companies to adopt a service-dominant logic paradigm [20]. This supposes that all consumers are active participants in value creation during all stages of service consumption; moreover, value is determined by the clients and depends on the context [21]. Evidently, the company is urged to develop initiatives to improve and build long-lasting company–customer relationships during the entire decision-making process [10]. With this purpose, this study introduces an emerging service innovation concept called relational innovation, which belongs to the nontechnological service innovation category [10,16,19,22]. However, very little research has been conducted on the relational innovation that allows conceptualizing this phenomenon.

To analyze the effectiveness of these subjective variables in the segmentation process, the research undertakes finite mixture modeling (FMM), assigning probabilities of belonging to each latent class according to the importance guests attach to company ICT facilities, relational innovation implementation, and motivations to consult eWOM before booking tourist accommodation. To the best of our knowledge, only recently have certain applications of the method been identified in tourism research, e.g., [3,7,9], and there are scarce examples that specifically target segmentation related to prepurchase eWOM, ICT, and innovation.

Hence, the outcomes of this study are expected to hold practical significance, particularly for hotel managers. This research’s theoretical contributions primarily revolve around shedding light on the latent diversity among tourist customers, focusing on ICT, relational innovation, and eWOM as emerging factors helping to detect distinct guest profiles. This theoretical background represents a crucial starting point for further research in how the interrelated factors of consumers’ perceptions of innovation, technologies, and their eWOM consultation behavior contribute to the segments formation, a phenomenon of interest to the management of companies in the lodging sector.

2. Literature Review and Research Question

2.1. eWOM and Motivation to Consult eWOM

In modern times, consumers seek new sources of information in order to decide about a product or service. In place of traditional word of mouth, considered a highly trusted face-to-face and direct exchange of information or experience, a novel marketing tool known as electronic word of mouth (eWOM) has emerged. eWOM offers potential consumers detailed, experiential-based, and up-to-date information online [13,23,24,25,26]. In the hospitality business context, eWOM might be represented by various forms such as online forums, online communities, social media, and, most important, online reviews [13,24].

The body of literature in eWOM research usually adopts one or both approaches, based on the consumer role (sender or receiver), where eWOM is studied as either antecedent or a consequence of consumer shopping behavior [27]. For the purpose of this research, we adopt the perspective of receivers of eWOM, where eWOM influences consumers in the prepurchase stage [4,7,17,18,25]. Indeed, recent studies showed that eWOM significantly impacts customers’ buying behavior, e.g., [13,14,24,25,28,29]. Moreover, in their review works, Donthu et al. [29] and Verma and Yadav [30] highlighted the eWOM antecedent nature as an emerging research topic. According to the authors, eWOM’s impact on sales and consumer behavior, on the one hand, and “the role of eWOM in experience-dominated sectors such as hospitality and tourism” [30] (p. 122), on the other hand, are central topics to be discussed in marketing. Thus, Appendix A shows some recent empirical research on the antecedent role of eWOM in tourism- and hospitality-related literature.

2.2. Information and Communication Technologies (ICT)

The recent literature has pointed out the many advantages that ICT adoption might bring to businesses: more customer satisfaction; higher loyalty; better communication and interaction; enhanced knowledge management; improved customer services and company performance, e.g., [8,11,16,31,32]. Furthermore, digitalization, as ICT implementation, has been recognized as a relevant aspect of the innovation process of companies and one of the main priorities on the agendas of most businesses [15,16], including hospitality ones [8]. In Appendix A, we review several recent empirical studies on ICT perception by the consumer to draw insights into technology’s impact on tourism, particularly in the hotel context.

The literature mainly supports the essential role of ICT adoption in consumer responses as positive perception of service quality, communication and company image [31,32], and satisfaction and loyalty between hospitality clients [11]. Nevertheless, ICT adoption is still ongoing in the tourism sector, and many emerging ICT aspects, such as robotization and artificial intelligence implementation, will be integrated into tourism and hotel operations [33]. The recent turbulent times, including global pandemic consequences, intensified “the speed at which it pushes along a wave of tourism and hospitality innovations that influences consumer cultures, preferences, choices, and identities” [34]. Gössling [34] points out that these changes might elevate the degree of disruptiveness and upset long-established economic models. The author stated that although the tourism system has, indeed, grown due to ICT, it has added risks and increased instability. Therefore, some scholars, e.g., [33], agree that the ICT topic calls for further research relying on modern-time incoming novelties in information and communication technologies.

From the marketing perspective, ICT plays a key role in the way companies interact and relate with their clients [15,19] and simplifies the information searching by consumers, especially in the tourism sector [8,10]. Notwithstanding, further refinement of the ICT concept is needed regarding its perception by consumers [10,33].

2.3. Relational Innovation

The COVID-19 outbreak pushed hospitality firms to discover new service offerings and ideas to fit into the post-pandemic new normal. Service innovation is recognized as an important strategy for hospitality businesses to achieve differentiation and respond to new consumer trends in sustainable development [35]. Thus, there is a growing interest in hospitality research on service innovation following a combination of technological and nontechnological innovations [36]. Shin and Perdue [35] in their study demonstrate the empirical importance of customer engagement for open innovation in hospitality services in a pandemic and post-pandemic era. Along the same line, highlighting the importance of customer participation, Hameed et al. [37] introduced open innovation as a mediator element of the effects of external knowledge and internal innovation on service innovation. Open innovation uses purposive inflow and outflow of knowledge from different stakeholders, where the company’s clients play a crucial role in accelerating innovation inside the firm through technology [8]. From the perspective of the open innovation paradigm, key ideas can emerge from both internal and external sources, thereby enabling firms to leverage by combining in new ways to create value [38]. In line with the resource-based view approach [39], firms’ internal resources may include the open innovation performance mechanism, such as internal innovation and external knowledge acquisition, which ultimately leads to service innovation in hotels [37]. Furthermore, a dynamic process of collaboration with external resources, including collaborators, networks, and communities, results in the co-creation of value by the parties involved, who work in a coordinated manner to achieve superior value [40]. However, most scholars agree that much remains to be learned about service innovation in hospitality and tourism and the optimal roles of customers in this process [36]. In Appendix A, we gathered several recent empirical studies on service innovation in hospitality from a consumer perspective.

Giving the hospitality business nature, service-dominant logic implementation suggests developing close relationships with the clients, treating them as active participants in service development and consumption at all stages [20,21]. Indeed, various studies’ findings evidenced that hospitality companies and their guests should establish close relationships to facilitate an open innovation process [41,42]. In this regard, relationship marketing is critical to establish relational bounds that might lead to co-creation activity to increment innovation [41] and client loyalty [42].

Therefore, a relational type of innovation has increasingly attracted the attention of scholars and practitioners, i.e., [10,19]. Relational innovation is a nontechnological innovation related to the company’s organizational and/or marketing system [19]. In this line, Marín-García et al. [16] suggest that a company’s marketing innovation (including relational innovation) aims to improve how companies deal with their consumers. Nonetheless, further examination is needed, as the existing literature on relational innovation as part of marketing service innovation is minimal [22], especially in the tourism and hospitality sector.

2.4. Research Question

Regarding customer heterogeneity, segmentation appears to be an answer when explaining differences in consumer behavior, especially in the hospitality industry [3,7,9]. Market segmentation analysis provides insights into existing consumer profiles and helps companies understand which market segments they should target [1]. Nevertheless, despite the significance of unobserved heterogeneity in research, it remains a topic that has not received sufficient attention within the tourism field research [3,6,7]. Although identifying unobservable heterogeneity provides valuable information on homogeneous subgroups of customers that report different responses to stimuli [1,3], it is still unclear what are the most effective criteria for market segmentation in tourist services [3,7]. Recent studies usually investigate observable characteristics such as socioeconomic factors like gender, age, and income, while often neglecting to explore unobserved variables, e.g., [43,44]. Moreover, those few studies that attempt to examine unobserved heterogeneity in tourism often do so by employing criteria such as distinguishing between the level of ecological concerns, e.g., [45], or leisure and business tourism preferences [9,11,46]. Therefore, scarce attention has been given to numerous other variables relating to consumer perceptions, such as those concerning ICT, relational innovation, and eWOM.

The prevailing body of research concerning customer behavioural responses operates under the assumption that the way clients perceive technological and nontechnological innovations from a company, alongside eWOM, impacts all buyers in a uniform manner. Consequently, it is presumed that buyers react similarly concerning outcome variables such as customer satisfaction and loyalty, e.g., [10,25,29,32]. However, this assumption appears to be somewhat unrealistic in numerous instances of behavioural research. Indeed, some researchers [47] claim that a considerable amount of heterogeneity must exist, as some customers perceive greater value in a service offering than others do. Motives to consult eWOM, ICT, and perception of relational innovation by guests are relevant variables to better understand tourists’ decision-making process in hospitality. Yet, only a few or no studies have directly addressed the diversity among consumer heterogeneity in relation to ICT [48], relational innovation perception, or eWOM consultation, e.g., [7,12].

Moreover, an examination of research in the tourism domain reveals that investigations into segmentation can be delineated into four distinct categories. The first category involves studies that have used some threshold values or simply binary coding to identify tourist segments, for example, leisure vs. business tourists or the degree of eco-concerns of the customers, e.g., [45,46], among others. The second category includes studies that have used clustering techniques, predominantly K-means, for segmentation purposes, e.g., [43,49]. The third category encompasses studies employing alternative segmentation methods or simply focusing on factors influencing customers’ perceptions and outcomes, e.g., [11,35,49]. Lastly, few studies were found that have used the FMM approach that considers unobserved heterogeneity when segmenting visitors based on their perceptions, e.g., [3,7,9].

This study attempts to examine how tourists’ perceptions-based segmentation can be performed, based on unobserved heterogeneity, through FMM. This procedure identifies the unobserved heterogeneity inherent to subjective variables and is framed within latent segmentation modeling [5]. Each individual is associated with a segment based on their likelihood of belonging to it, providing a much more realistic view of market segmentation than deterministic methods such as hierarchical/nonhierarchical cluster procedures [50].

Therefore, following all the above, we aim to implement the FMM method to evaluate the ability of customer perception regarding ICT, relational innovation, and motivation to consult eWOM in order to segment the hospitality market. Thus, the following research question is proposed:

RQ: Are there different guest segments based on the evaluation of the ICT, relational innovation, and motivations to consult eWOM?

3. Research Design

3.1. Data Collection and Measures

The data were collected through a consumer panel in Spain using a structured questionnaire based on the tourist guests’ opinions about their hotel stay during the post-pandemic recovery period. The measurement instrument included scales selected from the reviewed literature and adapted for the hotel environment (see Appendix B). The perceived relational innovation was measured using the proposal of Oke and Idiagbon-Oke [51]. The items for the perception of ICT implementation scale were adapted from the work of Wu et al. [15]. Finally, to measure motivation to consult eWOM, the scale adapted from the work of Kim et al. [28] was used. A 7-point Likert scale was applied to measure all items, where 1 was “strongly disagree”, and 7 was “strongly agree”. Moreover, data on the guests’ age, gender, trip purpose (business/leisure), and hotel type (stars) were collected to use to outline the potential latent segments.

The fieldwork was conducted by a specialized panel company between October and November 2020, coinciding with the time when the governments of Spain (both national and regional) started lifting restrictions on citizen mobility. An initial list of hotels was compiled based on the official Spanish Hotel Guide and the databases SABI (Sistema Ibérico de Análisis de Balances) and DUNS100,000. Participants were selected from Spanish tourists who had stayed at the hotel after the restriction was lifted. The sample was delimited by quotas based on the respondents’ gender, age, and the location of their hotel, specifically by Autonomous Community within Spain. In order to complete the questionnaire, permission was requested from the hotel to conduct the interviews at the reception and access areas. The questionnaire was carried out in the mornings and evenings. A response rate of 70% was achieved at the end of the sample collection. In total, 393 valid questionnaires were obtained, which implies a sampling error of 5.04% (p = q = 0.5 and infinite population), indicating that the sample represents the target population accurately. Most respondents were men (50.9%), the average respondent age was around 44 (±15) years old, the primary trip purpose was for leisure (93.9%), and more than half of the guests stayed in hotels with four stars (58%) (see

Table 1. Sample profile.

Gender	%	Occupation	%
Male	50.9%	Student	7.1%
Female	49.1%	Employee	62.6%
Age (years)		Self-employed	8.9%
18–25	14.5%	Unemployed	6.9%
26–35	19.6%	Housekeeper	2.0%
36–45	21.4%	Retired	12.5%
46–55	19.3%	Education level
56–65	14.5%	Primary education	6.6%
>65	10.7%	Secondary education	14.0%
With a partner?		Vocational school	18.6%
Alone	57.3%	Higher education	54.7%
In pairs	28.0%	Postgraduate education	6.1%
With family	5.1%	Travel reason
With a group	4.3%	Leisure/vacations	93.9%
With others	57.3%	Business	5.6%
Hotel rating		Both	0.5%
2-star	2.8%	Number of nights spent
3-star	32.3%	2	31.8%
4-star	58.0%	≥3	68.2%
≥5-star	6.9%	Advance booking
		<1 week	30.3%
		1–2 weeks	48.9%
		2–3 weeks	13.7%
		>3 weeks	7.1%

).

3.2. Data Analysis and Results

To assess the psychometric properties of the measurement scales, a first-order measurement model (confirmatory factor analysis) was estimated with ML Robust using EQS6.2. An item from Motivations to consult eWOM (“I read online reviews about hotels because I like being part of a virtual community”) was eliminated since it showed a factor loading < 0.6. The results shown in

Table 2. Measurement model estimation.

Latent Construct	Item	Factor Loading	t-Stat	R2
F1. Motivations to consult eWOM α = 0.949 CR = 0.957 AVE = 0.760	MOT1	0.881 a		0.776
	MOT2	0.907 **	25.14	0.823
	MOT3	0.910 **	30.94	0.827
	MOT4	0.883 **	32.14	0.780
	MOT5	0.886 **	28.61	0.784
	MOT6	0.892 **	27.98	0.795
	MOT8	0.733 **	19.81	0.538
F2. Relational Innovation α = 0.948 CR = 0.951 AVE = 0.886	RINN1	0.920		0.846
	RINN2	0.930 **	39.24	0.864
	RINN3	0.932 **	36.79	0.869
F3. ICT implementation α = 0.951 CR = 0.951 AVE = 0.828	ICT1	0.886		0.786
	ICT2	0.928 **	35.35	0.861
	ICT3	0.925 **	29.65	0.855
	ICT4	0.905 **	30.90	0.820

Fit indexes: χ²_Sat-B/df = 194.19/74 = 2.62; RMSEA = 0.064; CFI = 0.970; IFI = 0.971; BB-NFI = 0.953; BB-NNFI = 0.964. ^a In italics, parameters fixed before estimation. α: Cronbach’s alpha; CR: composite reliability; AVE: average variance extracted. **: p < 0.01.

indicate that the three scales reached optimal levels of reliability and internal consistency (Cronbach’s α > 0.7; composite reliability > 0.7 [52]; AVE > 0.5 [53]). The convergent validity was confirmed since all the standardized loadings were higher than 0.7 and significant to their latent construct [54], as depicted in Table 2.

According to Fornell and Larcker’s [53] criterion, the discriminant validity was also verified since the linear correlation between each pair of latent constructs was lower than the square root of the AVE of the scales involved (see

Table 3. Descriptive and correlations between latent factors (discriminant validity).

Latent Construct	Mean	SD	F1	F2	F3
F1. Motivations to consult eWOM	5.13	1.40	0.872
F2. Relational Innovation	4.67	1.51	0.811 **	0.941
F3. ICT Implementation	4.33	1.44	0.289 **	0.308 **	0.910

SD: Standard deviation. In Italics, diagonal values represent the square root of the AVE. Below the diagonal: correlations between constructs. **: p < 0.01.

). A χ² difference test was also performed to compare the estimation of the unrestricted model with the estimation where the correlations between latent factors were restricted to the unit. The results (Δχ²_Sat-B(df = 3) = 52.49, p < 0.001) indicated that the unrestricted model estimation showed a significantly better fit, confirming that each scale measures a different latent construct.

A finite mixture model was estimated with Mplus 8 to determine whether hotel guests have heterogeneous motivations to consult eWOM, perception of the ICT implementation and relational innovation. To avoid a local maximum that did not coincide with the absolute optimum, the estimation process of the segments assumed the increase in the initial random values and the iterative limits in the estimation–maximization algorithm [55]. The model was estimated from s = 1 (no heterogeneity) to s = 4 (4 latent segments or classes). The best option was determined concerning the lowest levels in the Bayesian information criterion (BIC = 2267.9) and the Akaike information criterion (AIC = 2212.3), suggesting three latent classes model in this way (see

Table 4. Fit indices to determine the number of latent segments (s = 1 to s = 4).

	-LL	AIC	BIC	CAIC	Entropy	R2	N Free Par
1-Class	1261.72	2535.44	2559.28	2565.28	1.00	1.00	6
2-Classes	1108.15	2236.30	2276.04	2286.04	0.86	0.88	10
3-Classes	1092.15	2212.31	2267.94	2281.94	0.72	0.73	14
4-Classes	1090.57	2217.14	2288.67	2306.67	0.61	0.57	18

-LL: -Log-likelihood; AIC: Akaike information criteria; BIC: Bayesian information criteria; CAIC: consistent Akaike information criteria; N free par: Number of free parameters.

).

Therefore, the following groups compose the 3-classes model: π1 = 49.1% (193 customers), π2 = 34.9% (137 customers), and π3 = 16.0% (63 customers) (see

Table 5. 3-class model estimation.

Predictors	Seg. 1 (n = 193; 49.1%)		Seg. 2 (n = 137; 34.9%)		Seg. 3 (n = 63; 16.0%)		Wald	R2
	Parameter (s.e.)	z-Stat	Parameter (s.e.)	z-Stat	Parameter (s.e.)	z-Stat
Intercept	0.33 * (0.13)	2.44	0.02 (0.16)	0.13	−0.35 * (0.18)	−1.96
Moti. Consult eWOM	−0.27 + (0.17)	−1.65	0.85 ** (0.15)	5.47	−0.57 ** (0.19)	−2.99	29.96 **	0.16
Relational Innovat.	−3.32 ** (0.81)	−4.11	3.25 ** (0.77)	4.23	0.07 (0.36)	0.20	18.36 **	0.79
ICT implement.	−2.98 ** (0.67)	−4.45	2.73 ** (0.56)	4.87	0.25 (0.32)	0.80	23.83 **	0.73

s.e.—standard error. +: p < 0.1; *: p < 0.05; **: p < 0.01.

).

As shown in Table 5, the significant effect of estimated parameter variables (motivations to consult eWOM, relational innovation perceptions, and ICT implementation perceptions) on the segments can be identified by z-statistic values. Regarding the discriminatory capacity of the three variables, the Wald statistic and the R² coefficient indicate significant differences at the global level in the position of each parameter variable between the three segments. At the local level, relational innovations and ICT implementation perceptions significantly impact the first and second segments. However, this impact is negative in the first segment, while in the second it is positive. In contrast, motivation to consult eWOM significantly affects all segments, being the only variable that impacts segment 3.

The composition of the groups was described with the categories of the different characteristics such as gender, age, trip purpose, and hotel type (see

Table 6. Segments profile.

Variable		Seg. 1	Seg. 2	Seg. 3
Motivations to consult eWOM F(df = 2) = 41.974 **		4.793 (±1.40)	5.917 (±0.93)	4.447 (±1.49)
Relational Innovation F(df = 2) = 399.5 **		3.461 (±1.02)	6.166 (±0.62)	5.111 (±0.81)
ICT Implementation F(df = 2) = 328.4 **		3.216 (±1.02)	5.701 (±0.79)	4.758 (±0.52)
Covariable
Gender χ2(df = 2) = 6.22 *	Male	46.1%	46.7%	63.5%
	Female	53.9%	53.3%	36.5%
Age F(df = 2) = 1.61	Mean (±sd) year	43.3 (±14.8)	43.2 (±14.8)	46.9 (±15.3)
Motive for traveling χ2(df = 4) = 6.98	Leisure	91.2%	97.8%	93.7%
	Business	7.8%	2.2%	6.3%
	Both	1.0%	0%	0%
Number of stars χ2(df = 6) = 19.31 **	2 *	5.2%	0.7%	0%
	3 *	35.8%	29.9%	13.4%
	4 *	56%	60.6%	16.2%
	≥5 *	3.1%	8.8%	33.3%

*: p < 0.05; **: p < 0.01.

). In order to determine differences depending on the segments, ANOVA tests were used to analyze significant differences when the dependent variable is quantitative. In contrast, contingency tables were used when the dependent variable is qualitative.

As observed in Table 6, the groups’ profile is as follows: Segment 1 shows the respondents’ lowest perception of ICT implementation (3.22 ± 1.02) and relational innovation (3.46 ± 1.02) compared to the rest of the segments. This group is formed by men and women, predominantly traveling by leisure (91.2%) and staying in three- and four-star hotels (91.8%). In contrast, the hotel guests in segment 2 show the highest perception of relational innovation (6.17 ± 0.62) and technology perceptions (5.70 ± 0.79). Moreover, in this group are the most motivated consumers that consult eWOM about the hotel (5.92 ± 0.93). Compared to other groups, guests from the second segment show the highest percentage of travel for leisure (97.8%) and choose to stay preferably in four-star hotels (61%). In segment 3, the motivation to consult eWOM is the only variable significantly and negatively impacting latent class formation (4.45 ± 1.49), presenting in this way a lower mean value than other segments. This last group is formed mainly by males, with the oldest age segment (46.9 years old ±15.3), and the chosen hotel type is primarily five-star and higher (33.3%).

4. Conclusions and Implications

As stated by Dolincar [1], marketing segmentation in tourism is still persisting and is a highly relevant topic in helping tourism companies shape their offer to specific customer profiles. Yet, the predominant focus has been dedicated to examining customer-observed heterogeneity rather than unobserved heterogeneity, thus potentially obscuring valuable insights for both theoretical and practical fields. This overlooks various behavioural factors as latent variables, which could provide significant contributions to a better understanding of customer segmentation. In this line, in comparison to other studies on marketing segmentation, this study contributes to the literature on unobserved heterogeneity segmentation by utilizing FMM segmentation, which has recently been detected as an emerging method to be employed in tourism marketing research. In this way, our findings expand the existing body of literature on FMM implementation in tourism segmentation [7,9].

Furthermore, this paper contributes to the academic literature on marketing segmentation in hospitality by expanding the answer to the question of the most effective criteria for market segmentation in tourism [3,6,7]. Recent marketing literature research has identified motivation to consult eWOM, perception of ICT, and relational innovation between other behavioural variables as highly important for the current purchasing journey of consumers, especially in the hospitality field, which has significantly evolved during the last few years. In this line, our findings confirm that customer perceptions of ICT, relational innovation, and motivations to consult eWOM are shaping distinct tourist segments. Therefore, the latent segmentation methodology has identified three segments with different influences of the named variables corroborating the utility of the FMM in segmentation analysis in marketing [2,5,7].

Thus, the first and second segments of consumers were shown to perceive a significant impact from all three variables, namely, ICT, relational innovation, and motivations to consult eWOM. However, these variables impact negatively in the first segment, showing that respondents reported relatively low evaluation for current variables. In contrast, ICT, relational innovation, and motivations to consult eWOM positively impact the second segment, where the tourist that belongs to this profile provided the highest evaluation scores, especially on the motivations to consult eWOM. Guests’ profile descriptions for both segments do not differ drastically, pointing to the fact that most of the guests from the second segment prefer to stay in higher-rated hotels compared to the first segment guests. In this sense, it might explain why the guests from the second segment positively perceived ICT and relational innovation and reported the highest motivations to consult eWOM. As the current hospitality literature suggests, the level of service, communication, and ICT facilities, together with expenses and risks, rise simultaneously as the hotel star category increases [2].

Finally, the third segment reports the significant, however negative, impact only by the motivations to consult eWOM, presenting in this way the lowest motivation to consult eWOM than other segments. This specific group showed the oldest age segment, represented mainly by men and choosing upscale class hotels, which might explain the reason that other variables lose relevance for this type of guest.

Overall, the formation of these three segments illustrates the discriminatory capacity of motivations to consult eWOM, relational innovation, and ICT implementation. According to the current studies [2,3,7], the motivations to consult eWOM are revealed as a variable that significantly affects the formation of the three segments. However, relational innovation and the implementation of ICT are variables that have a significant impact only on the first two segments, with a negative impact and a positive impact, respectively. Based on the findings obtained, the first two groups are the most vehemently opposed, as the former comprises consumers with low levels in these three criteria, while the latter is made up of consumers with high levels in all the bases. In contrast, the third segment is distinguished by the lowest levels of motivation to consult eWOM compared to the previous groups.

Therefore, the ability of these variables to identify statistically different segments of guests allows us to advance research on segmentation in hospitality, following the call of research proposed by Donicar [1]. Furthermore, both the perception of relational innovation and variables associated with technologies should be considered in segmentation processes in this context in line with previous studies [5,6].

Moreover, our study enhances the conceptualization of ICT, relational innovation, and motivations to consult eWOM, responding in this way to the research calls of several review studies, e.g., [27,29,33,36]. The findings provide insights into how these variables impact different consumer segments within the hospitality industry, extending the knowledge proposed by previous studies, i.e., [7,48].

Regarding the practical contribution of this study’s findings, hotel managers might utilize discovered guest segments in several ways. Since the importance of segmentation bases differs for each segment, tourism companies should design or adapt strategies according to their characteristics to focus marketing efforts more effectively.

Firstly, in terms of motivations to consult eWOM, the practitioners might want to focus on the second segment that showed the highest motivations. For example, it will be important for a hotel to interact and provide responses to guests who have already left their review notes on the booking platforms or sites. In addition, it will create a positive image of a company that cares about its clients’ opinions. In the case of the first and third segments, which showed low and negative motivation to consult eWOM, a different way of communication might be considered to reach them. For instance, the company might want to use direct communication or email marketing to help build a personal target message, especially for consumers from the third sector.

Secondly, considering customers’ relational innovation and ICT perception, results suggest enhancing the degree of technological and nontechnological innovation as it is, indeed, relevant for the first and second segments. The first segment showed a negative perception of named variables that might point to poor hotel information and communication technologies and low-quality interaction with the company. The fact that clients care about such things should draw the attention of the hoteliers. They might want to increase their presence in social networks, which helps build relationships with clients through frequent interaction. Finally, suppose the perception of innovation and ICT implementation is not relevant for the third segment. In that case, the hotel will have to develop actions aimed at changing guests’ attitudes towards these variables, for example, by highlighting the importance of innovation in services and technologies to increase customer value and differentiate from its competitors.

Despite all of those opportunities based on our research findings, some research limitations must be noted. First, to generalize the findings to other tourism sectors and cultures, future research might want to outline customer segments within other business types and countries. Since most of the sample concentrated on guests in four- or five-star hotels, the discriminatory capacity of the segmentation bases could be evaluated for other types of accommodation with different service levels, such as lower- or higher-category hotels. The type or level of service of the hotel could help to better differentiate the profile of the segments in order to better adapt marketing strategies. Second, the results showed relative similarity between the composition of the first and second segments. Thus, more specific demographic, geographic, psychographic, and behavioural traits such as level of income, frequency of travel, beliefs, and others should be gathered to better segment description. Considering consumer attitudes towards tourism and hospitality, as well as their preferences and interests, would help identify different lifestyles that could also improve the profile of the segments. The motivations to write a review could also be studied to capture not only the eWOM consultation phase before the purchase decision but also the diffusion phase of the experience. Finally, after the pandemic consequences, it might be a great opportunity to apply segmentation analysis based on those variables such as, for example, perceived risk, trust, control, and safeness, which assess consumer attitudes toward service companies in post-pandemic times.