Tourist Preferences for Revitalizing Wellness Products and Reversing Depopulation in Rural Destinations

Abstract

Wellness tourism is a growing segment that destinations can rely on to increase tourism flows and revive underdeveloped rural areas. This study elicits tourists’ preferences for wellness tourism products and the redevelopment of a wellness destination in a depopulated rural area. The research included 595 participants in Central Greece using on-site, face-to-face surveys. The survey data were analyzed using a two-step process. First, conjoint analysis was applied to determine the combination of wellness tourism product attributes that maximize visitors’ utility. In the second step, we elicited visitors’ willingness to pay (WTP) for improvements in tourism product attributes using the contingent valuation method. The findings indicate that the combination of attributes that maximizes the preference structure of respondents includes time of visit from September to October, duration of visit from 1 to 7 days, wellness tourism as the primary purpose of the visit, and wellness activities that contain both physical activities and training as well as spiritual and emotional enhancement. Using binary logistic regression, the analysis indicates that respondents are willing to pay EUR 47.33 per visit on average further to develop wellness tourism products in the study area. Empirical results confirmed that sustainability issues, quality, and entrepreneurship factors positively impact visitors’ WTP for revitalizing wellness products. In contrast, a longer duration of stay in the area and respondents’ age negatively impact their WTP. These findings suggest that structuring a customer-driven tourism product will enhance the rural destination’s efforts to attract high-yield visitors and advance the rural economy. Furthermore, practical implications reveal that Destination Management Organizations will benefit from research results to foster targeted management plans.

1. Introduction

Depopulation is a reality experienced by many countries in our modern world [1,2]. In [3], the authors conceptualize depopulation as a combination of population decline, economic downturn, loss of employment numbers, and other social issues. Similarly, Ref. [4] argues that depopulation stems from the interplay of multiple macro-processes at the local scale. The root causes of depopulation range from demographic [5] to environmental [6] and economic issues [7]. Economic factors are considered the most crucial to resolving this issue since many depopulated areas are generally economically depressed [8].

The long-term depopulation or “shrinkage” of rural areas deeply affects their development process in terms of economic, cultural, social, and environmental dimensions, resulting in a self-reinforcing vicious cycle of decreasing potential for maintaining their regional vitality [9]. Such a population loss calls for regional adaptive policies to reverse this shrinkage and keep the region’s economy alive. Tourism development remains a promising option to balance the adverse effects of economic restructuring, specifically in rural areas [10]. Tourism provides income and employment opportunities in host communities, motivating local-scale production, enhancing incomes in the local economy, and improving residents’ material standard of living [11]. From this perspective, it is crucial to investigate how tourism can help revive rural areas [12] since rural tourism is one alternative that can prevent or reverse depopulation [13]. Interestingly, in [14] the authors claim that rural tourism profoundly supports socioeconomic regeneration and diversification.

Consequently, attention should be paid to advancing the local population’s well-being and making the region a welcoming place to live [15]. In addition, researchers have identified the growing interest of rural tourism enterprises in developing their services towards wellness offerings [16].

The complexity of modern tourism markets requires destinations to consider different segments, seeing tourists as a heterogeneous pool of consumers [17]. However, the preferences of wellness tourists who visit rural destinations and the mechanisms that shape their visitation patterns have been overlooked, even though the tourism literature has emphasized the importance of wellness products in developing targeted marketing plans and destination management [18,19]. Ref. [20] argue that generating information about wellness tourism demand and the preferences of visitors who travel for wellness motives is necessary and essential.

In light of such concerns, this study investigated opportunities to redevelop depopulated rural areas by fostering wellness tourism products. We apply quantitative methods to decode visitors’ preferences for wellness tourism products that can be used to design customer-driven wellness offers. We apply relevant methodologies to understand visitors’ preferences for wellness tourism products that can be used to design customer-driven wellness offers. Customer-driven means that the customer leads the way, their voice is the primary focus, and customer satisfaction provides the constancy of purpose vital to success [21]. In this effort, the analysis aims to determine the most preferred set of attributes of a wellness tourism product and elicit consumers’ willingness to pay (WTP) for improved wellness tourism offerings, thus highlighting the potential and applicability of this growing tourism segment. To accomplish our research objectives, we rely on two stated preference methods. These methods allow the researcher to present a scenario or hypothetical market to respondents. Then, respondents state their judgment or preference for this given scenario. Furthermore, tourism stakeholders should be aware of wellness visitors’ needs since they vary extensively at different times and stages in their lifetime [22]. In this context, we rely on a two-step method to uncover visitor’s preferences for rural wellness tourism products. In the first stage, we use CA to determine the respondents’ preference structure regarding a hypothetical scenario concerning wellness tourism products. This stage reveals which combination of attributes is most preferred by visitors when experiencing wellness tourism offerings. In the second stage, we use binary logistic regression to estimate respondents’ willingness to pay as part of the CVM. In sum, our methodology estimates how much a visitor is willing to pay to experience a wellness tourism offering in the study area based on their preferences as elicited in the first stage, that of CA.

Therefore, the main research goal of this study is to identify the preference ranking of a bundle of attributes with different perceived values derived from conjoint analysis (CA) and elaborate on their monetary valuation in terms of visitors’ WTP using the contingent valuation method (CVM). This research concludes that visitors do not prefer to experience wellness offerings in the ‘high season’ of summer months with a duration of stay at 7 days. Respondents prefer to visit a wellness destination in autumn (from September to October). Also, on average, visitors are willing to pay EUR 47.33 per person per visit to receive both physical activities and training and spiritual and emotional enhancement activities. A key issue that emerges from our findings is the issue of sustainability when revitalizing wellness offerings and the quality of relevant tourism experience, which increases the willingness to pay of visitors. Furthermore, highly educated visitors are willing to pay more to receive the offered wellness services.

The paper proceeds as follows. Section 2 presents the literature review, while Section 3 develops the methodology and the econometric procedures used to analyze the survey data. Section 4 presents the empirical findings and interprets the results. Finally, Section 5 discusses the findings’ implications and offers the research endeavor’s conclusions.

2. Literature Review

Many research studies evidence that the regional economy, healthcare, and welfare status are core factors directly connected with depopulation [1].

Less populated rural areas should see tourism development as an opportunity for growth that creates wealth since they can play a dominant role in offering a unique tourism experience [13].

Many economically depressed rural areas have sought ways to embed tourism as a driver of economic development, conservation, and rehabilitation of the sociocultural and historical–artistic heritage and revitalization of local life [23]. Therefore, to advance an area as a tourism destination, especially if it has undergone a high degree of depopulation, it is important to investigate the motivations of tourists and how these motivations impact their choice of activities in relation to rural tourism [13]. Rural wellness tourism is the combination of rural resources with wellness tourism offerings. Therefore, rural destinations may match wellness services with existing tourism resources to create social value [24] and market success [17].

Consistency in the literature regarding the definition of wellness tourism is lacking, and opinions on what forms the concept of wellness diverge significantly [22]. Wellness tourism destinations provide visitors with good and welcoming natural environments and/or particular cultures to preserve or improve their health [20,25] without medical intervention during their stay at the destination [26]. Refs. [27,28] conceptualize tourism as a holistic process that directly involves the human need for physical health, beauty, longevity, consciousness-raising, spiritual sensibility, and connections with the community, the environment, or religion. Understanding the complexity and the significant number of perspectives and approaches to wellness, in the Global Spa & Wellness Summit, the concept of wellness tourism was defined as travel associated with the pursuit of maintaining or enhancing one’s personal well-being [29]. With increased economic growth, aging populations, and the advancement of human health consciousness, wellness tourism concerns a new tourism market segment promoting healthy, ecological, and green lifestyles [30]. In the literature, it is common to see the terms health, wellness, and medical tourism used interchangeably, although there is a clear distinction among these concepts [26,31].

Wellness tourism is a growing niche worldwide, providing specific opportunities to businesses offering products that promote or maintain health [32,33]. Ref. [34] investigated the dominant role of wellness tourism in developing small businesses and generating opportunities for deploying entrepreneurship by revitalizing destinations (e.g., beach resorts) and creating related work positions. As indicated by [30], the current literature is either descriptive or focuses on particular aspects of wellness tourism’s supply [19,35,36]) or demand [18,37,38]. Demand-focused trends are linked to the customer or the tourist and their wellness desires. In contrast, the supply-related trends relate to the wellness tourism providers and their response to the demand-side expectations [39]. Ref. [40] authors argue that successful wellness tourism centers are highly concerned with the issue of sustainability, offering at the same time memorable services in terms of quality and reasonable prices.

Many studies see wellness holistically [19,41,42,43], pointing to the development of tourism that is increasingly oriented toward wellness [44,45]. Specifically, as reported by [46], on the supply side, researchers have focused on the characteristics and potential of development strategies for the wellness tourism industry [35,47,48,49,50]. On the demand side, researchers have focused on the sociodemographic and behavioral characteristics [51,52,53] motivations [32,54,55], and experiences [56] of wellness tourists, along with the factors that enhance their well-being [38,57] and influence their travel decision-making [58,59]. However, the concept of visitors’ preferences for wellness tourism products has not been adequately investigated and is less visible.

Modern tourism trends demand segmentation, flexibility, integration, and consideration of the surroundings, seeing the tourist as the consumer [17]. Therefore, it is required to bring visitors’ judgments and preferences into the decision model to provide suitable offers and develop viable tourism policies in the long run.

Thus far, empirical research efforts on wellness tourism do not clearly explain or characterize visitors’ preferences in wellness destinations. Recognizing this gap in the literature, a more in-depth look at the preferences of wellness visitors is necessary to support the continuous improvement, planning, and management of this rapidly growing market. Destinations should reflect consumer preferences and satisfy demand, especially in a rapidly evolving world economy, where tourist preferences change, for instance, in response to technological developments [60].

Therefore, it is necessary to develop studies that enable providers and managers to understand the preferences of visitors who perceive and use the wellness destination differently, approaching wellness tourism as a heterogeneous market rather than a homogenous entity. In addition, preferences have been acknowledged as core elements to explain traveler behavior at the destination [61]. By investigating and understanding consumer preferences and their associated monetary value, we can develop marketing strategies that increase customer satisfaction, loyalty, and retention [62], thus supporting the economic sustainability of tourism in rural areas. Notably, it is crucial to decode and incorporate visitors’ preferences into tourism planning to enhance a destination’s economic potential [63].

Within the body of literature on wellness tourism, the majority of studies focused on issues such as motivations to travel for wellness [18,37,59,64,65], market demand analysis [20,22,35,66,67,68], and the impact of wellness tourists on residential communities [46]. The authors in Ref. [31] suggest that empirical studies should be applied in the field of wellness tourism “to glean in-depth information about who these tourists are or exactly what they are looking for, and what it is worth to them” (p. 16). In sum, the previous research has not sought to study preferences for wellness tourism, and many of the questions remain unanswered about visitors’ preferences at wellness destinations. To address this literature gap, we first employ CA to determine the combination of characteristics that maximizes visitors’ utility. In the context of tourism, CA has been used extensively to study visitors’ preferences [69,70,71,72,73,74,75,76].

This study enriches the existing literature since it includes a new set of variables to elicit visitors’ preferences under the same methodological approach, offering meaningful and interpretable results. Although these variables are key determinants when planning a tourism product, it is untested or unobserved in similar research efforts. Furthermore, the study combines the inference power of two stated preference methods: CA and CVM. This methodological approach remains unexplored in relevant literature despite its applicability. Consequently, this study covers a wide range of development opportunities that act as driving forces in the context of tourism. The scope of the study includes the time of visit and seasonality, as well as environmental protection and sustainability, which are crucial issues for many countries, especially as a function of natural resource use and rural development. Such a comprehensive and holistic approach is rare in empirical research on tourism development.

We go one step further and estimate how visitors value these preferred attributes of the proposed wellness product in monetary terms to develop a research-based approach that can be used for effective destination marketing. The significance of visitors’ willingness to pay in tourism studies has been widely highlighted [77,78,79,80,81,82,83,84,85,86,87,88]. Nevertheless, to the best of our knowledge, no empirical research studies have shown the link between preferences elicited with CA and visitors’ WTP for these preferred attributes.

The concept of WTP should be considered not merely as a monetary variable in the tourism category but as a critical factor that can have multiplicative effects on a destination’s performance. In other words, visitors’ WTP should be seen as a determining factor that may generate positive externalities and multiplicative effects in destination economies [89]. Our study also highlights the potential for wellness-tourism-driven economic revitalization to scale back depopulation effects on rural places, a critical sustainability challenge of rural landscapes. This is a paramount contemporary issue in developed and developing worlds alike, resulting from the global trend of urbanization [90].

The key that can unlock the potential of wellness tourism to revitalize depressed rural areas is creating and establishing customer-driven tourism products that meet visitors’ needs and fulfill expectations. Doing so requires a better understanding of visitors’ preference structure or the preferred attributes of the wellness offerings. Our framework considers the different attributes that could be part of a wellness tourist’s preference structure, such as duration of visit, whether wellness is the primary motivation for a trip, the season in which the trip is to take place, the type and focus of wellness activities, and the added benefits of wellness tourism, such as the development of human resources and environmental sustainability (Figure 1). The final step is determining the willingness to pay for these preferred attributes.

When developing tourism destinations, a key research topic that consistently emerges is that of seasonality issues, which in turn interrelates with time of visit and duration of stay. Tourism seasonality calls for offering tourism products beyond the conventional ‘high-season’ period of summer months. Hence, extending the period during which tourists visit the destination remains an issue with research interest to grow and sustain tourism demand and relevant visitation rates. These issues are widely investigated with empirical research models that consider when a potential visitor might prefer to visit a destination based on a set of alternatives and how many days the visitor is willing to spend at a destination. Arguably, the season of visit to the destination remains a core issue in reducing seasonality effects [91,92]. For instance, Ref. [93] mentions that a wide range of studies use seasonality variables such as the number of tourists per month, the number of overnight stays per month, or the average length of stay per month. Similarly, Ref. [94] examined the time of visit at a destination, indicating three types of seasonality patterns: off-peak (e.g., tourism activity occurs throughout the year), single-peak (e.g., tourism activity that occurs in certain months such as summer months), and two-peak (e.g., tourism activity with two main peaks, for instance, the summer and winter seasons).

Consequently, these fluctuations as drivers of seasonality issues are addressed in the proposed model to understand relevant effects at rural wellness destinations better. Not surprisingly, seasonal fluctuations are related to the type of tourism and kind of activities in rural areas [91], which in our case are wellness tourism offerings to revitalize rural areas. Thus, the purpose of the visit is deeply interrelated to the visitors’ travel motivation factors at the interface of destination attributes. Notably, the purpose of a visit widely concerns what a destination offers (e.g., wellness offerings) so that it can motivate a visitor to make the decision to visit the destination. Consequently, based on these destination attributes, an emerging issue of research appears regarding what a visitor prefers to do at the wellness destination. For instance, a visitor prefers to experience physical, spiritual, or both types of wellness activities. These activities form a holistic wellness product to maintain or enhance visitors’ personal well-being, which is the core of wellness tourism [95]. Importantly, in terms of destination attributes, the proposed model considers the motive of travel, namely if the visit to the study area is primarily motivated by wellness offerings, or if wellness plays a secondary role in the visit. Ref. [96] claims that motivation factors help to identify attributes that should be promoted and determine markets in which tourist motives can be associated with destination attributes, features, and resources. Also, destination choice emanates from tourists’ assessment of attributes and their perceived utility value, which can be examined using the joint CA–CVM model. Moreover, destination attributes can be recognized as the benefits that visitors seek or expect to receive and experience at a particular destination [97]. Therefore, preference heterogeneity might be a useful research instrument to formulate policies to confront tourism seasonality based on diversified visitors’ judgments of a tourism offering. In addition, economic variables related to the seasonal distribution of tourism are highly important in empirical studies [98]. On the same wavelength, socioenvironmental variables affect the growth process of a destination [99]. Such a research perspective supports the formulation of customer-driven tourism products tailored to visitors’ preferences, thereby developing a unique wellness tourism experience that provides ‘value for money’ and ‘value for time’ spent for the delighted visitor. Supportively, the visitor’s changing preferences and needs are partial driving forces for the rise and fall of a destination [99].

3. Methodology

3.1. Study Area

The empirical component of this study is centered around the existing but derelict thermal lagoons in the town of Kamena Vourla, in the Municipality of Molos–Agios Konstantinos, in the Prefecture of Fthiotida, in Central Greece. Figure 2 shows the location of Kamena Vourla in Greece and provides an overall glimpse of the wider region. Only a few hot thermal spring lagoons exist in the study area, known locally as ‘Koniavitis’ thermal springs (Figure 3). This area was selected as a study site because it serves as an archetype of similar settings where the landscape’s dominant feature is an interface between the rural environment and naturally occurring or preexisting wellness resources.

However, these lagoons are underdeveloped, meaning that no supportive facilities and relevant infrastructure are present in the area. For instance, buildings, treatment installations, and bathing facilities around and very close to the location of the thermal spring lagoons are out of use, derelict, or abandoned. The current capacity for visitors to bathe in the thermal water is very limited due to the size and shape of these lagoons and the derelict state of support infrastructure. In practice, it is an open-access resource. The face-to-face survey was conducted at the Koniavitis springs to capture visitors’ judgments for further wellness development of the existing pools and redevelopment of the existing infrastructure, as shown in Figure 3. Tourists were intercepted at a point very close to the existing underdeveloped lagoons, between the site of the lagoons and the center of the town of Kamena Vourla. The distance between the site and the town center is approximately 1 km. The interviews were conducted in the Greek language since most visitors were domestic.

3.2. Data Collection

We used a structured questionnaire to elicit respondents’ preference rankings and willingness to pay for an ideal wellness tourism product. The questionnaire was composed of an introductory section and four main sections. The introductory sections contained information about the research and its purpose. The first section included questions related to the travel behavior of respondents and their prior experience with wellness tourism products. Respondents stated if they had visited the research area in the past (Yes or No), the current duration of stay in the study area (days), and if they had received wellness tourism services in the past in any area (Yes or No). Furthermore, respondents were asked to state if they believe that the development of wellness tourism in the area can transmit to the visitor a message of: (i) the protection of the study area through their responsible behavior (Yes or No) and (ii) a relationship between sustainability and the use of the area’s natural resources, such as the thermal spring water (Yes or No). Respondents were also asked to state if they believe that the development of wellness tourism can contribute to: (i) developing new skills in terms of human resources (Yes or No); (ii) improving the quality of the overall tourism experience in the area (Yes or No); (iii) enriching the provided tourism product in regard to other forms of nature-based tourism such as fishery tourism (Yes or No); (iv) developing new tourism facilities such as marinas for cruise boats and yachts (Yes or No); and (v) developing entrepreneurship in fields such as the food and beverage sector or gastronomy (Yes or No).

The second section of the survey included the CA and the combinations of attributes of the proposed tourism product. The third section of the survey included the CVM to elicit respondents’ WTP to access or enjoy the preferred tourism product. Finally, in the fourth section of the survey, respondents were invited to answer questions concerning their socioeconomic and demographic characteristics. The final questionnaire was the output of a comprehensive literature review focusing on wellness tourism product attributes and formulating questions based on local peculiarities, specific needs, and the authors’ expert judgment. Data collection took place in the summer of 2021, from 1 June 2021 to 31 October 2021.

The cluster sampling technique was used as a sampling method since no preexisting sample of tourists to the study region was available [100]. Cluster sampling is a random sampling technique in which a cluster “constitutes the sampling unit, instead of a single element of the population” [101] (p. 2). The population of interest is divided into nonoverlapping populations called clusters [102,103]. The clusters (N) denote the sampling frame. After dividing the population into N clusters, a simple random sample of these clusters (n) is selected [101,104]. In our case, the N clusters denote the days between 1 June 2021 and 31 October 2021, namely 153 days. Then, from these N clusters (N = 153 days), the researcher randomly selects n clusters (days) to conduct the on-site interviews (simple random sample). In our case, with the help of a random number table, 45 days (n clusters) were selected to deploy the survey. Each day of the 45 days of the research constitutes the sampling unit of the research [105]. The term cluster denotes each particular day on which the research took place. According to the cluster sampling technique, the respondents are the elements or items of each day of the 45 days of data collection. The cluster sampling technique is more cost-effective and time-saving than other options [100,105,106]. During the sampling period, 595 questionnaires appropriate for statistical analyses were collected out of a total of 688 responses. Ninety-three responses were deemed unsuitable for statistical analysis because of incomplete answers.

3.3. Conjoint Analysis (CA)

Conjoint Analysis (CA) incorporates the preferences ranked by the targeted users of a product and/or a service and is used to “decompose the holistic information about respondents’ reactions (e.g., statements or choices) to a set of stimuli into the relative importance of each level of each factor (or attribute) according to a prespecified utility model” [107] (p. 4). In other words, CA is based on the premise that a consumer’s utility from the consumption of a good can be decomposed into part-worth utilities or values regarding the benefits or attributes of this good offer [108,109].

The first step in applying the method was determining the attributes of the proposed tourism product and the respective levels for each attribute. For the present study, four attributes were used, as indicated in

Table 1. Attributes and levels of the wellness tourism product.

Attributes	Levels
Time of visit	November–March April–June July–August September–October
Duration of visit (days)	1–7 8–14 15+
Purpose of visit	The primary purpose is wellness tourism The secondary purpose is wellness tourism
Wellness tourism services (Types of wellness activities)	Physical activities and training Spiritual and emotional enhancement Physical activities and training and spiritual/emotional enhancement

. The first attribute was the ‘Time of the visit’ (levels: November–March, April–June, July–August, September–October), and the second attribute was the ‘Duration of the visit’ in days (levels: 1–7; 8–15; 15+). The third attribute was the ‘Purpose of the visit’ (levels: The primary purpose is wellness tourism; The secondary purpose is wellness tourism), and the last attribute was the ‘Types of wellness activities’ (levels: Physical activities and training; Spiritual and emotional enhancement; Physical activities and training, and spiritual/emotional enhancement). The ‘Time of the visit’ was chosen to investigate if potential wellness visitors prefer the high or peak tourism season to enjoy wellness activities. The attribute ‘Duration of the visit’ was selected to understand how many days future visitors prefer to stay in the area, which can be integrated into potential wellness tourism packages. Finally, the ‘Purpose of the visit’ was included in terms of searching for the specific reason for the visit. The ‘Purpose of the visit’ is inherently tied to the prioritization of destination marketing and management efforts for placing wellness tourism at the core of the redevelopment of the area as a wellness destination or considering wellness tourism as a secondary development giving priority to other forms of tourism. The ‘Types of wellness activities’ were selected to understand if visitors prefer holistic physical and spiritual options to improve their wellness (Table 1).

The first attribute is composed of four levels; the second is composed of three levels; the third one includes two levels; while the fourth encompasses three levels, resulting in 4 × 3 × 2 × 3 = 72 possible combinations of product attributes. In the context of CA, each combination of product attributes represents a profile of the proposed tourism product that the survey participants must rank. In practice, ranking 72 combinations is not feasible due to respondents’ potential confusion and fatigue. Therefore, to avoid dysfunctions and low questionnaire completion rates, a fractional factorial design was produced using the orthogonal main-effect plan resulting in sixteen combinations of attributes. The orthogonal design (orthogonal main-effect plan) ensures that the attributes’ levels are uncorrelated and that the estimate of one factor is unaffected by the estimates of the others [110,111]. Respondents were presented with all these 16 combinations of characteristics representing wellness tourism visits in the form of a profile. The profile presented to the respondents can be seen in Appendix B. All survey respondents were visitors to the study area. Respondents were asked to rank these combinations of characteristics. The results were coded by assigning a value between 1 and 16: the most preferred combination took the value one (1), and the least preferred the value sixteen (16). This method can define the preference structure concerning the proposed tourism product based on respondents’ judgments. The preference structure provides evidence of how important each attribute is in the overall decision and how the various levels within an attribute influence the overall preference, which is measured by a set of parameters called part-worth utilities [112]. Part-worth utilities represent the relative value respondents place on each product attribute’s levels [113]. CA data are analyzed with Ordinary Least Square Regression (OLS) to estimate part-worth utilities [114,115]. The conjoint action of two or more attributes represents the independent variables, and the respondents’ preference rankings represent the dependent variables [52]. Based on the same logic, all utility values for each of the sixteen profiles (combinations of attributes) can be calculated, determining the preference structure based on respondents’ judgments.

3.4. Contingent Valuation Method

The Contingent Valuation Method (CVM) is a stated preference method that considers both use and non-use values [116]. In much empirical research, CVM defines what a natural resource is worth [117]. A CVM survey asks respondents to state their preference in a hypothetical or contingent market, allowing researchers to estimate demand for goods or services not traded in existing markets. Respondents undergo a survey-based choice experiment that elicits whether or not they are willing to pay (WTP) for a change in the provision of goods or services for a specified price [118]. Common payment vehicles in CVM studies may be specific amounts of additional taxes or entrance fees. Analyzing the data using regression methods that recognize the binary nature of the dependent variable, researchers can estimate the amount of money respondents are willing to pay for the hypothetical good or service.

In the present study, the hypothetical scenario presented to survey respondents involved a proposed wellness tourism product, the attributes of which were ranked by the CA according to respondents’ preferences. After the ranking process, respondents were presented with a single-bound dichotomous choice (i.e., yes or no) question. In the single-bound dichotomous choice framework, the amount of money respondents choose whether or not they are willing to pay (bid amount) is randomly prespecified in the questionnaires. The payment vehicle selected was a hypothetical entrance fee to enhance and further develop the existing thermal springs, lagoons, and wellness infrastructure from the derelict status quo outlined in the study area section. The bid amount ranged from EUR 5 to EUR 50 with a step increase of EUR 5. The single-bound format more accurately imitates real-life market situations in which consumers decide if they want to pay a specific price for a commodity or not. Such an approach is preferred to the open-ended elicitation format, as it is considered to be more incentive-compatible [119]. The single-bound dichotomous choice WTP question simulates to a great extent real market situations “in which the decision is a take-it or leave-it situation, and because it has better properties in terms of incentive compatibility (see NOAA panel report in [120] Arrow et al. 1993), it is more widely used than the open-ended format” [121] (p. 1827).

The willingness to pay question was introduced by telling respondents that by developing a tourism product based on wellness tourism, we might be able to improve the offered tourism experience in terms of quality and sustainability. Respondents were further told that the financial support of such an endeavor would require the payment of an amount in the form of an entrance fee (ticket) per visitor per day. Then, respondents were asked if they were willing to pay X (EUR) to enhance wellness tourism in the study area. Respondents could answer “Yes” or “No” to this question in a take-it-or-leave-it decision. Thus, the respondent’s acceptance of the bid amount reflects their willingness to pay that bid value.

An essential decision point in the analysis is the treatment of protest votes. For example, respondents may object (protest) to the principle of placing a monetary value on the commodity provided, or they may feel strongly that the responsibility for provision falls on another actor, such as the Government [121]. Additionally, respondents may protest against some components of the CVM survey [122]. For example, protest bidders often object to the imposed market structure in a CVM study [123] by disagreeing in principle with the payment vehicle or the mechanism to provide the good or service. Therefore, protest votes should be discriminated from genuine “No” answers because the bid level is too high. Differentiating between genuine or true “No” answers and protest “No” responses is based on responses to a follow-up question requesting the selection of reason(s) for refusing the bid from a menu of options [121]. However, there is no common agreement in the literature on the criteria for categorizing responses as a protest or true zero value [123]. Debriefing questions to define the reasons behind refusing the bid is essential to applying the CVM [124]. Therefore, following CVM practice, protest votes were excluded from the calculation of WTP [122] if individuals indicated in their debriefing answers that they disagreed in principle with the payment vehicle or were not interested in wellness tourism.

3.5. Binary Logistic Regression (BLR)

Logistic regression models the probability of presence and absence given the observed values of the predictor variables [125,126]. In logistic regression models, the dependent variable can be either binary or categorical, whereas the independent variables can be a mixture of continuous and categorical or binary variables. In BLR models, the dependent variable has a binary value, namely “Yes” (accept bid amount) or “No” (reject bid amount). In the BLR model used in the present study, we coded with 1 the answers of respondents who accepted the bid amount and with 0 the responses related to No answers. The variables used in the BLR model are described in

Table 2. Variables included in the BLR model.

Variable Name in the BLR Model	Interpretation of Variables	Unit of Measure—Variable Coding
bid	The amount of money offered	(EUR)
edu	Education level	Years
income	Monthly income level	(EUR)
age	Respondents’ age	Years
Believes that wellness tourism can be a mechanism for:
envirovar1	The protection of the study area through their own responsible behavior (in terms of pollution and hygiene)	Binary variable coded with yes (1) and No (0)
envirovar2	The importance of sustainability through natural resource exploitation (e.g., thermal spring water, seawater)	Binary
skills	Development of new skills in terms of human resources	Binary
quality	Improvement of the quality of the offered tourism experience	Binary
tourproduct	Enrichment of the provided tourism product (e.g., in regard to other sustainable forms of tourism such as fishery tourism)	Binary
facilities	Development of new tourism facilities (such as marinas for cruise boats and yachts)	Binary
entrepreneurship	Development of entrepreneurship (e.g., in the field of the food and beverage sector—gastronomy)	Binary
loyal	If respondents have visited the research area in the past	Binary
welltour	If respondents have visited and received wellness tourism services, in the past, in any area	Binary
stay	Duration of stay in the study area	Days

.

4. Results

4.1. Sociodemographic Characteristics

In the survey, 595 participants offered complete responses for further analysis. Descriptive analysis showed that there were 345 male visitors and 250 female visitors. A total of 311 respondents were in the age group of 31–40 years old. The mean age of the respondents was 39.7 years. Most survey participants (212 respondents) had a monthly income of EUR 1501–2000, while 182 respondents were found to be self-employed. Concerning their level of education, the largest proportion of the respondents were well educated since they possess a university degree (183 respondents).

4.2. Results of Conjoint Analysis

The utility scores (estimates) or part-worth utilities and the respective standard errors for each level of each attribute are presented in

Table 3. Part-worth utilities of level attribute included in CA. The bold letters indicate that this specific combination of levels of the attributes is the most preferred by respondents.

Attributes	Levels	Utility Estimate	Std. Error
Time of visit	November–March	−1.574	1.034
	2. April–June	−1.602	1.034
	3. July–August	0.897	1.034
	4. September–October	2.280	1.034
Duration of visit (days)	1–7	0.840	0.796
	2. 8–14	0.493	0.934
	3. 15+	−1.335	0.934
Purpose of visit	The primary purpose is wellness tourism	1.262	0.597
	2. The secondary purpose is wellness tourism	−1.262	0.597
Types of wellness activities	Physical activities and training	−0.326	0.796
	2. Spiritual and emotional enhancement	−2.153	0.934
	3. Physical activities and training AND spiritual/emotional enhancement	2.479	0.934
Constant		8.371	0.660

. Figure 4 provides the graphical representation of the part-worth utilities, which measure the unique subjective judgment of preference. Part-worth utilities are used to determine the importance level of each attribute for that product and/or service [112]. A low part-worth utility indicates less value; a high part-worth utility indicates more value for the specific level of the attribute. Negative values do not mean denial or rejection. The higher the utility values, the higher the preference, but higher negative values lead to lower utility. The total utility of the most preferred combination can be derived by adding the most preferred levels of each attribute in Table 3. The constant value is also included in the calculations. As a result, the most preferred combination of the tourism product makes a total utility value of 15.232.

The importance values reflect the contribution of each factor to the overall preference of the good or service consisting of a bundle of attributes.

Table 4. Relative importance of attributes included in CA.

Attributes	Importance Values
Time of visit	25.7
Duration of visit (days)	29.4
Purpose of visit	14.6
Types of wellness activities	30.3

presents the range of the importance values, while Figure 5 graphically illustrates the importance summary regarding the attributes that compose the proposed wellness tourism product. Attributes with greater utility contribute more compared to those with lower utility. The relative importance of each attribute shows its importance relative to other attributes. These importance values sum up to 100%. For example, at 30.3%, the “Types of wellness activities” attribute is the most crucial attribute, with “Purpose of the visit” being the least important attribute. The Pearson’s R (0.881) and Kendall’s tau (0.717) statistics reflect the correlations concerning the observed and estimated preferences. Both statistics are high and significant.

The findings suggest that the combination of attributes that maximizes the preference structure of respondents is: time of visit from September to October, duration of the visit from 1 to 7 days, the primary purpose of the visit is wellness tourism, and the type of wellness activities include both physical activities and training and spiritual/emotional enhancement. Moreover, the attribute ‘types of wellness tourism’ is more important than the other three attributes. The second most important attribute is ‘duration of stay.’ The third important attribute is the ‘time of visit,’ whereas the ‘purpose of visit’ is the least important.

4.3. Results of the Contingent Valuation Method

In the survey, 344 participants stated a positive WTP answer, while 150 respondents expressed a ‘clear no’ answer (not protest votes). The classification results (

Table 5. Classification table of WTP responses using the BLR.

Observed
	Predicted		Percentage Correct
	NO	YES
NO	301	89	77.2
YES	50	883	94.6
Overall percentage (%)			89.5

) reveal that the binary logistic regression correctly classified 89.5% of the responses. Overall, 261 respondents provided a No answer to the WTP question. Of these No answers, 101 were identified as protest votes and, consequently, were excluded from the analysis. As a result, the total number of answers included in the binary logistic model was 494. Protest votes were categorized based on choosing the following options as justification for ‘No’ votes: ‘although the development of a wellness tourism product is important to me, I am not willing to pay any amount’ (47 responses), ‘I disagree with the payment method for developing wellness tourism in the study area’ (23 responses), ‘I believe that my financial support will not be managed appropriately’ (19 responses), ‘It is an obligation of the state to provide the needed financial support’ (9 responses), ‘I do not have adequate information for the study area in terms of developing wellness tourism’ (3 responses).

The binary logistic regression results confirm that all coefficient signs had the expected and statistically significant signs (

Table 6. Results of binary logistic regression.

Variables	B	SE.	Wald	df	Sig.	EXP(B)	95% CI for EXP(B)
							Lower	Upper
bid	−0.199	0.019	114,649	1	0.000	0.820	0.790	0.850
edu	0.114	0.065	3.102	1	0.078	1.120	0.987	1.272
income	0.001	0.0001	3.116	1	0.078	1.001	1.000	1.001
age	−0.024	0.013	3.304	1	0.069	0.976	0.951	1.002
envirovar1	1.002	0.378	7.038	1	0.008	2.723	1.299	5.707
envirovar2	0.521	0.298	3.061	1	0.080	1.683	0.939	3.016
skills	0.592	0.345	2.949	1	0.086	1.807	0.920	3.550
quality	0.747	0.447	2.796	1	0.095	2.110	0.879	5.063
tourproduct	0.668	0.368	3.293	1	0.070	1.951	0.948	4.016
facilities	0.968	0.397	5.955	1	0.015	2.632	1.210	5.725
entrepreneurship	0.619	0.357	3.002	1	0.083	1.857	0.922	3.739
loyal	0.649	0.311	4.368	1	0.037	1.914	1.041	3.517
welltour	0.491	0.296	2.747	1	0.097	1.634	0.914	2.920
stay	−0.055	0.033	2.865	1	0.091	0.946	0.887	1.009
Constant	3.126	1.300	5.781	1	0.016	22.779
Nagelkerke R2 0.602 Hosmer and Lemeshow 0.232

). The bid amount coefficient had a negative sign, indicating that respondents were less willing to pay for higher bid amounts. This means that the higher the bid amount, the lower the likelihood (probability) of the respondent paying the offered amount. Respondents’ education level (‘edu’) significantly and positively impacted the probability of being willing to pay for the proposed wellness tourism product. The more educated the respondent, the more willing the respondent was to pay for the improvement in wellness tourism products in the study area. In the case of ‘income’ level, the higher the income, the more willing the respondent was to pay. The negative sign of the variable ‘age’ indicates that the older the respondent, the less they are willing to pay for the proposed wellness tourism product. The same is true in the case of the duration of stay (‘stay’). The longer the duration, the lower the willingness to pay.

The positive sign of the coefficient associated with the ‘envirovar1′ variable indicates that respondents are more willing to pay for a wellness tourism product that transfers the environmental message to visitors. The authors in Ref. [127] argue the importance of achieving the desired equilibrium concerning the socioeconomic system and nature’s dynamics. Likewise, the positive sign of ‘envirovar2′ suggests that respondents are more willing to pay if they recognize the relationship between sustainability and natural resource use (e.g., thermal spring water, seawater). Furthermore, results reveal that respondents’ willingness to pay increases if they consider that the offered tourism product can contribute to developing new skills in terms of human resources (‘skills’). In the same vein, respondents’ willingness to pay increases based on the quality of the offered tourism experience (‘quality’), the enrichment of the provided tourism product in regard to other nature-based forms of tourism (‘tour product), developing new tourism facilities such as marinas for cruise boats and yachts (‘facilities’), and developing entrepreneurship in the field of the food and beverage sector–gastronomy (‘entrepreneurship’). Additionally, findings suggest that respondents who had visited the area before (‘loyal’) are more willing to pay. Moreover, respondents who had previously visited a wellness tourism destination and received relevant services were more willing to pay (‘well tour’).

The mean WTP was calculated by adding all regression coefficients except the bid coefficient. Then, the sum was divided by the bid regression coefficient. We concluded that the mean willingness to pay equals EUR 47.33. To ensure the validity of the outcome, we performed a Monte Carlo simulation with 1000 reiterations (Figure 6) using the Krinsky–Robb procedure [128]. The result of the Monte Carlo simulations complies with our point estimate of WTP calculated with the results of the logistic regression model.

5. Discussion

Depopulated rural areas need targeted plans to revive their economies and communities based on high-leverage tourism economic activities. On the other hand, in the aftermath of the COVID-19 pandemic, many questions have been raised about revitalizing tourism activities and the possibility of the tourism model evolving or transforming [129]. The revitalization of wellness tourism attractions to develop customer-driven products constitutes a promising but also challenging path to improve and promote currently depopulated and underdeveloped rural areas.

To this end, it is vital to understand wellness tourists’ preferences so that services and products can be designed to match customers’ expectations [31]. Customers form their outcome satisfaction when making comparisons of product attributes and performance to their prior expectations [130]. The set of attributes in this CA study included when to visit (time of visit), how many days (duration of visit), why (purpose of visit), and what to do (type of activity). Initially, this might be seen as a generic approach to the proposed attributes that might compose a wellness offer. However, all these attributes provide practical information about questions (or concerns) that potential visitors ask themselves (or consider) when selecting a destination (e.g., wellness destination).

We investigated if visitors prefer physical activities, spiritual activities, or both. The findings suggest that visitors prefer destinations that satisfy their need for wellness in various diverse arenas, from the physical to the psychological and spiritual to the cultural [131]. The precise type of physical and spiritual activities to be implemented can be the subject of an additional research effort. Previous literature has documented wellness destinations that have developed products such as spa centers, massage, yoga retreats, new age events, scenic hikes, local fresh foods and flavors, full-body wellness treatments, and experiences built upon healing mind–body practices, such as meditation, or organized programs based on escapism and relaxation [32,41,44]. However, including specific types or sets of activities in the context of CA when defining the term physical and/or spiritual activities would create many attributes and levels, potentially creating confusion among respondents of a rank-and-choice survey. Five or six attributes may be considered adequate and manageable, providing dependable results [132]. Therefore, it may be necessary to examine preferences in a piecemeal manner if CA is to be used in such research endeavors.

We applied the CA and CVM to analyze the economic feasibility of revitalizing a depopulated rural destination by developing wellness value propositions based on tourists’ wants and needs for diversified and integrated wellness products.

In this study, we selected the time of visit as an attribute of the proposed wellness tourism product to identify if potential visitors prefer the high or low season. This attribute can provide evidence for the importance of developing year-round destinations. Such an attribute can be used in relevant studies and shape a picture of when visitors prefer to enjoy the offered tourism experiences. The same seems reasonable concerning the duration of the visit regarding carrying capacity and sustainable natural resource use. However, these sets of CA attributes have rarely been used in empirical studies, even though they are essential pieces of information to formulate offers or develop enhancements of existing tourism attractions and destination attributes.

Research findings allow us to consider concrete and specific practical implications. Specifically, emphasis should be put on determinants that make tourists willing to pay more for experiential products during their visit to rural wellness destinations. For instance, destination marketing professionals should target well-educated and environmentally sensitive potential tourists who not only prefer but also pay for quality and sustainability in the offered experience. Consequently, campaigns and communication plans should target this specific market segment. Additionally, all interested parties should attract investment and improve facilities to increase relevant tourism demand. One important issue is that of tourism seasonality, which concerns tourism demand on a yearly basis. Research shows a clear disposition of respondents to avoid overcrowded, overpopulated, and ‘overconsumed’ time periods, with a preference for the autumn months.

On a local scale, research findings indicate that it is vital to offer a set of physical training activities, a wide range of options and alternatives for stress discharge, and spiritual enhancement, all of which are recognized as new customer-driven trends in the existing wellness market. Moreover, the mean age of visitors was 39.6 years old, meaning that the destination should focus on activities, facilities, and wellness attractions that are close to the age of visitors regarding the appeal, use, and practical issues. These findings are crucial in expanding the study area’s wellness appeal and redeveloping it with a reinvigorated potential and new focus based on the preferred attributes and visitors’ willingness to pay.

The study area is at the center of recent efforts to regain its popularity in the context of wellness tourism. Planning and redevelopment efforts of the rural area will be undertaken as part of an investment plan (i.e., master development plan). These abandoned facilities and thermal lagoons will be rebuilt, reconstructed, restored, and repurposed, providing the bedrock on which rural development in the study area can flourish with good long-term prospects. These new and preferred wellness offerings and the new facilities are expected to return the lost tourism demand and increase relevant visitation rates beyond the traditional high season in summer.

At a broader level, Destination Management Organizations (DMOs) should bring together a wide range of tourism stakeholders, to structure effective management plans for sustainable tourism development [133]. Therefore, one alternative would be to formulate and advertise relevant tourism packages based on visitors’ preferences by considering all the preferred attributes derived from the application of CA and CVM. Another option could be to take these results into relevant decision-making processes in the broader context of effective destination marketing and management. This would help state or regional administration policies to become adaptive since these policies should reduce regional inequalities by funding tourist infrastructures and/or retaining capital for tourism-related investments [134].

Regarding the CVM, research findings indicate that visitors are more willing to pay for tourism wellness products that advance the concept of sustainability. Therefore, tourism products should provide wellness services that transfer the sustainability message to visitors as an integrated component. Nowadays, visitors seem to prefer destinations that respect the environment and where the tourism experience serves to raise awareness about sustainability. The question that lies in the background is: Do we need more tourism, or do we need better tourism? This is a question that each tourism stakeholder should attempt to answer in the process of managing and marketing destinations. In our opinion, better tourism means an inclusive wellness tourism product that respects the environment, advances the offered tourism experience in terms of customer-driven preferences, and provides opportunities for destinations to experience self-sustaining (endogenous) development. For instance, redeveloping a wellness tourism product can help secondary entrepreneurship develop (e.g., gastronomy tourism with local products, outdoor recreation alternatives, and relevant facilities) and advance the quality levels of the offered tourism experience. This is based on the fact that visitors are willing to pay more for wellness destinations that have the potential to meet customer demands and fulfill their expectations.

Furthermore, one would expect that a robust wellness tourism economy leaves margins for innovation and investments resulting in increased visitation rates. Increased willingness to pay will increase tourism consumption, translating into higher tourism receipts and expenditures in wellness destinations. In turn, it is expected that the contribution of wellness tourism to the overall economy will be positive and significant. Notably, this is the case for tourism-dependent economies, which constantly seek ways to increase tourism’s direct and total contribution to the region’s GDP.

Stated preference methods such as conjoint analysis and contingent valuation or willingness to pay are widely used to make inferences about products that do not yet exist, but that could be reasonably implemented or developed. We mention that an investment plan budgeted at 10 million Euros is about to be implemented in this area to redevelop the wellness tourism offerings. This investment will benefit the local community in terms of socioeconomic factors such as employment and income. As it is planned and expected, decent, long-lasting work positions with a wide range of roles and responsibilities for the workforce will be available. Furthermore, this endeavor can be expected to increase the local income levels considerably. For instance, tourism flows in the area based on visitors’ preferences and willingness to pay are expected to support the anticipated demand for the recognized and empirically investigated new wellness tourism offerings. Moreover, these two determinants of growth, namely, employment rates and income levels, will minimize leakages and embed sustainability in the local population, which will find stimulus towards a promising future in their native region. Research findings can provide investors with supporting information to enrich relevant decision-making processes. For instance, the findings of the study offer evidence of the economic potential of extending the tourism period and thoroughly considering what visitors prefer to do when visiting the wellness destination in terms of physical and spiritual wellness activities. Similarly, willingness to pay offers a monetary perspective of how much visitors are likely to spend to experience the new offerings, whereas they provide inputs and insights for the fiscal policy the destination might follow.

Research findings indicate that the wellness enhancement will offer opportunities to develop new skills and human resources expertise, increasing visitors’ willingness to pay. Additionally, research results suggest that wellness enhancements and developments enrich the provided tourism product (e.g., in regard to other forms of nature-based tourism), provide the basis to develop new tourism facilities (e.g., marinas for cruise boats and yachts), as well as help build entrepreneurship (e.g., in the field of the food and beverage sector–gastronomy). These findings provide concrete evidence that wellness tourism can become a stimulus to develop the destination in a wide range of attributes since all these attributes increase visitors’ willingness to pay. These issues can be an example or the basis for further use by other wellness destinations worldwide since the attributes of this study’s proposed product may differ based on the destination’s current situation or other specificities within the area of interest.

Translating customer preferences into destination attributes demands key measures that “put a strategy into language that everyone can understand” [135] and help remain competitive by creating new high-value products. The process is continuous, with needs being triggered and satisfied by products that will stimulate future demand [125]. In addition, academic research concludes that the pandemic will positively influence consumer behavior in favor of wellness tourism in the medium term [136,137]. In this regard, destination readiness is crucial to alleviate the negative impacts of the pandemic and remain open to receiving visitors [138].

This study provides future research opportunities to enrich the methodological approach and glean novel empirical insights about rural wellness tourists and their preferences. First, researchers may use different product attributes or enrich those used in this study. Similarly, researchers can identify a new preference structure given the unique characteristics or special features of other study areas or based on the individual interests and judgment of the investigators who implement the analysis. Second, one can elaborate in depth on reasons for expressing a protest vote in the CVM. Specifically, multivariate analysis can be used to identify the profiles of the respondents who provided a protest vote. This may include factor–cluster analysis to segment participants who protest against paying money to receive wellness tourism offerings. Third, this study potentially can be used for other types or forms of alternative or special interest tourism that use natural resources and wish to increase visitation rates: for instance, agri-tourism, wine tourism, or coffee tourism. These tourism types are mostly offered or developed in rural areas, as is the case in the present study.

6. Conclusions

The present study aimed to investigate development opportunities to revive depopulated rural areas in the context of wellness tourism. We applied relevant methodologies to this effort and defined relevant customer-driven tourism products. First, we applied CA to determine the preference structure of respondents and identify the combination of the most preferred attributes. To provide a holistic research perspective in the context of stated preference methods, we used CVM to calculate the mean WTP of participants. Research findings suggest that the combination of attributes that maximizes the preference structure of respondents is: time of visit from September to October, duration of a visit from 1 to 7 days, the primary purpose of the visit is wellness tourism, and the type of wellness activities experienced contain both physical activities and training as well as spiritual/emotional enhancement. The respondents’ WTP for these preferred attributes to further support the wellness developments in the study area, examined as a potential entrance fee to access the wellness offering, equals EUR 47.33 per visitor. Additionally, respondents’ answers provided evidence concerning a wide range of variables that impact their WTP.

Respondents’ preferences seem to be a sustainable foundation for implementing wellness tourism products. This destination management and marketing approach supports customer satisfaction and enhances the tourism system’s performance status. Increased visitation rates, tourism season expansion, and host community improvement can be achieved through realistic goals and investment opportunities to meet visitors’ preferences. Although the variables chosen for this research were derived from the literature and specific features in the study area, a new set for a different destination could be selected using field checks and background information from various reports.

7. Limitations

As is the case with other empirical research conducted in recent times, this study’s limitations concern the COVID-19 pandemic and the time at which the field research took place. The data collection for this study began in June 2021 and ended in October 2021. The data collection process had been scheduled before the pandemic and could not be postponed or transferred due to time constraints and obligations of the authors. As a result, restrictions taken at that time from the Hellenic Government might have impacted the visitation rates in the study area and, consequently, potential participants in the on-site interviews. However, the authors did manage to collect a considerable number of completed questionnaires and further proceed with the analysis.

A second limitation concerns the number of attributes included in the conjoint analysis of this study. We included four attributes split into relevant levels. Although this number is adequate and recommended for this type of analysis, one could have used additional attributes with their levels. However, this approach would generate a large number of combinations to be ranked by the participants, creating confusion and potential refusal or incomplete questionnaires.