Are Image and Quality of Tourist Services Strategic Determinants of Satisfaction? Millennials’ Perspective in Emerging Destinations

Abstract

This study aims to investigate the level of satisfaction of foreign tourists, specifically millennials, based on different dimensions of destination image and the quality of the tourist services in Belgrade, the capital of the Republic of Serbia, as an emerging destination. Despite the tarnished image attached to the city after the Yugoslav Wars and the economic crisis in the 1990s, this emerging destination is becoming more popular among foreign visitors. A sample of 359 international tourists was surveyed. Structural equation modelling was used to test the relationships between the constructs. The results confirmed the positive effects of image and quality on satisfaction as well as a positive and direct effect of image on quality. Moreover, there was a partial mediation of quality between image and satisfaction. The findings have significant marketing and management implications for destination stakeholders. They provide useful insights for choosing effective strategies and an appropriate business tourism model in emerging destinations that can further help them to become more competitive.

1. Introduction

At a time when almost all places on Earth have become accessible, the competition between tourist destinations is greater than ever before. In these circumstances, the built image of tourist places, the quality of the services provided, and tourist satisfaction, as the result of tourist experience, are the basis of successful marketing and destination management strategies (Butler 1980; Iordanova and Stylidis 2019; Marques et al. 2021; Woosnam et al. 2020). Only satisfied tourists will spread a positive view by word of mouth, and these tourists will tend to return to a destination (Yoon and Uysal 2005). Past research has indicated that destination image (Assaker et al. 2011; Alcañiz et al. 2005; Bigné et al. 2001; Chen and Li 2018; Chi and Qu 2008; Lam et al. 2020; Marques et al. 2021; Wang and Hsu 2010) and service quality (Bigné et al. 2001; Castro et al. 2007; Chen and Li 2018; Osman and Sentosa 2013; Wang et al. 2017) positively influence tourist satisfaction. Nevertheless, most tourism research has focused on mass tourism destinations, while academic literature for emerging markets, although necessary, is still scarce. Moreover, of all international travel in Europe, the trips taken by millennials account for 40% (Ketter 2021). So, this is the demographic group that travels the most. Thus, it is essential to examine their travel preferences because all marketing campaigns that reach this age group can be almost completely sure of the success of their businesses (Chan 2018).

The position of emerging destinations is especially challenging as these destinations need to create successful strategies that will enable them to penetrate the volatile tourism market. As they usually lack these strategies (Marques et al. 2021), their further tourism development is questionable. On the other hand, it is known that the tourists’ perceptions differ in emerging and mature destinations (Marques et al. 2021). However, more profound academic research in this direction has yet to be developed. Therefore, it is necessary to explore how tourism demand perceives the attractiveness of emerging areas. Destination image and the quality of services are found to be the dominant factors of a destination’s attractiveness that motivate tourists to visit it (Kim 2018; Wang et al. 2017). In other words, the most important characteristics of destination image and service quality are key strategic determinants that need to be identified first. Then, their preservation or their maintenance of a high level of service needs to be ensured so that destination commercialization does not impair their quality. In line with the above, the guidelines for future destination development can be defined; these comprise one of the important prerequisites for avoiding spontaneous development, setting the basis for sustainable destination development, and achieving competitive advantages in the tourism market.

When it comes to the impact of image on satisfaction, previous studies have identified the image in several ways: (1) as a cognitive dimension in the form of a general idea of the destination image (Assaker et al. 2011; Bigné et al. 2001); (2) as a cognitive dimension consisting of selected attributes of the destination (Albaity and Melhem 2017; Castro et al. 2007; Chen and Tsai 2007; Chen and Li 2018; Chi and Qu 2008; Kim 2018; Lee and Xue 2020; Marques et al. 2021; Sanz-Blas et al. 2019; Su et al. 2020); (3) as a combination of cognitive and affective dimensions (Kim et al. 2019; Lam et al. 2020; Lee et al. 2014; Prayag 2009); (4) separately, through cognitive and affective image (Lam et al. 2020; San Martín et al. 2019; Schofield et al. 2020); and (5) as general image assessment (Sharma and Nayak 2018; Wang and Hsu 2010). It can be concluded that, although extensive research has been carried out on destination image as a predecessor of satisfaction, it has exclusively focused on two dimensions: the cognitive and the affective. The third dimension, conative image, has been previously assessed only to a very limited extent because tourists’ destination loyalty, being the intention to revisit and recommend a destination, has usually been identified with a conative image. Thus, the conative image has rarely been included in tourism studies (Woosnam et al. 2020). This study intends to implement Gartner’s (1993) three-component destination image approach and explore its effect on tourists’ satisfaction with the destination. Another prerequisite for satisfaction, which occupies an important place in tourism literature, is the quality of the tourist services. The quality has been defined in two ways, as the overall quality of the destination (Bigné et al. 2001; Castro et al. 2007) or through different destination dimensions (Chen and Li 2018; Osman and Sentosa 2013; Rahman et al. 2017; San Martín et al. 2019; Wang et al. 2017). A hypothesized destination satisfaction model also includes a direct relationship between service quality and satisfaction because service quality is another strategic determinant to be examined when developing initial marketing strategies in emerging destinations.

The emerging destination chosen for this research is Belgrade, the capital of the Republic of Serbia, located in the southeastern part of Europe (Figure 1). A study conducted by the Organization for Economic Co-operation and Development (OECD 2018) has shown that Southeast Europe, with its six emerging economies, is one of the fastest-growing regions when it comes to tourism. As a tourist destination, Belgrade is extremely attractive as it permeates east and west and north and south and has a history whose beginnings are linked to a time as early as the Neolithic and, on the other hand, the modern age. The city is located at the confluence of two rivers, the Sava and the Danube, which naturally divide its landscape into northern plains and hilly southern parts. Moreover, it has been recognized as a famous congress and city-break tourism destination in Europe (OECD 2018). A rapid and constant increase in the number of foreign tourists and nights spent has been noticed since 2009, when the global financial crisis finished. Taking into account the tourist traffic in the 21st century, there is an increase of 241% in the number of foreign tourists and 206% in nights spent when comparing the years 2000 and 2019 (Figure 2) (Statistical Yearbook of Belgrade 2020). This situation is largely expected given the ethnic conflicts and wars that took place in the former Yugoslav Republic during the 1990s, as well as the NATO bombing of Yugoslavia in 1999. All of this brought an extremely negative image to Belgrade, and the city has started recovering and slowly regaining its position on the tourist map of Europe since 2000. This could be the reason why tourists mostly come to Belgrade with modest expectations (Todorović et al. 2017). Furthermore, of the total number of tourists who stayed in the Republic of Serbia in 2019, 33% stayed in Belgrade (Statistical Office of the Republic of Serbia 2020). The data are even more impressive when it comes to international tourism because as many as 55% of foreign tourists stayed in Belgrade out of the total number of foreign visitors in the same year. Therefore, Belgrade bears the epithet of the most visited tourist place in the Republic of Serbia.

The practical contribution of the current study aims to provide useful insights for the future tourism development of Belgrade. The Tourism Development Strategy of the City of Belgrade 2020–2025 (2019) points to the lack of a business model for tourism development in the city. Moreover, there is no comprehensive research on the tourism demand in Belgrade apart from the fact that international visitors are seen as a target market (Tourism Development Strategy of the City of Belgrade 2020–2025 2019). In addition, “tourism infrastructure, accommodation, and skills” have been recognized as areas with the most room for improvement (OECD 2018), and marketing campaigns targeting millennials in Serbia should be focused on a positive image (Perčić and Spasić 2021). Therefore, the research conducted for the purposes of this study intends to provide guidelines for defining a long-term strategic goal by identifying factors that attract the most numerous group of international tourists, the millennials, who will be the backbone of tourism development in Belgrade.

Millennials, or Generation Y, are one of the four generations differentiated after the Second World War. The other three are Baby Boomers, Generation X, and Generation Z. Baby Boomers were born between 1946 and 1964 (Kaifi et al. 2012). They have already retired or will not be active workers in the next couple of years. This demographic group spends the most money on travel compared to the others and prefers smaller and quieter destinations (Hysa et al. 2021). On the other hand, Generation X comprises individuals born between 1965 and 1980 (Kaifi et al. 2012); they are interested in city-break and cultural tourism but still have a preference for peaceful destinations when traveling internationally (Hysa et al. 2021). Generations Y and Z grew up with the internet and smartphones; so, they are highly technology savvy (Ketter 2021). As there is no agreement regarding the year when Generation Y ends and Generation Z begins, in this paper millennials encompass all individuals born in the late 20th century or, more precisely, between 1981 and 2000 (Kaifi et al. 2012). Those of Generation Z are still economically dependent on their parents, while millennials have disposable income which is higher than average (Chan 2018). They are creative travelers who want to travel independently, experience the local life, and create memorable experiences. Secondly, they seek novelty, such as less-known destinations. Thirdly, they do not pay much attention to accommodation as their primary goal when traveling is gaining a unique and authentic experience, as well as getting in touch with the locals (Ketter 2021). In addition, they are the most populous generation and the one that travels the most (Expedia Group Media Solutions and Skift 2019). Thus, it is imperative to investigate their satisfaction with a destination in order to find out the factors influencing it.

Overall, there is little emphasis in academic research on emerging tourist destinations even though it is predicted that emerging economies will reach 57% of the market share by 2030, or over 1 billion international tourist arrivals (World Tourism Organization 2017). Few emerging destination studies conducted in the last five years relate to image, quality, and satisfaction. The study conducted by Schlesinger et al. (2020) examined tourists’ assessment of destination attributes, their perceived quality-of-service experience, and their loyalty in seven northern and southern Mediterranean destinations. Another study, in Sofia (Bulgaria), explored the relationship between destination image, tourism satisfaction, and intention to recommend the place and purchase its products (Marques et al. 2021). However, the current study is the first attempt to examine the impact of image (the cognitive, affective, and conative dimensions) and quality (accessibility, accommodation, and attractions) on satisfaction and the effect of image on quality, as well as the mediating role of quality between image and satisfaction in an emerging destination among millennial tourists. Moreover, the way this study views destination image and quality of tourist services differs from that of previous research. That is to say, the destination image viewed as a whole is made up of three dimensions, where the conative dimension can, to some extent, be identified with future intentions (factors: word of mouth and recommendations and revisits), while the quality of tourist services is explained via the constructs of accessibility, accommodation, and attractions. Finally, the managerial implications for both the public and the private tourism sectors will be discussed.

2. Literature Review

2.1. Destination Image

In order for the destination to be adequately positioned on the market and to stand out from the competition, it is necessary that people (consumers) have a positive awareness of it (Ekinci and Hosany 2006). The knowledge and a unique image contribute to improving its attractiveness and strengthening its competitiveness (Pike 2002). Hunt (1975) was one of the first researchers to prove that a good image of a destination can increase the number of tourists visiting it. However, it is difficult to give one definition of the destination image because “there are almost as many definitions of image as scholars devoted to its conceptualization” (Gallarza et al. 2002, p. 59). One of the most commonly used definitions is that the destination image is “the sum of beliefs, ideas, and impressions that a person has of a destination” (Crompton 1979, p. 18).

When it comes to conceptualizing the destination image, there are two main approaches: the three-dimensional continuum and the three-component approach. The first one was proposed by the authors Echtner and Ritchie (1991) and implies the existence of an attribute–holistic, functional–psychological, and general–special continuum. The latter is based on Boulding’s (1956) research, which explains that image consists of “what one knows and thinks about an object (cognitive component), how one feels about it (affective component), and how one acts using this information (conative component)” (Agapito et al. 2013, p. 472). This three-component approach, whose theoretical framework was first set by Gartner (1993), is much more prevalent in the tourism literature (Zhang et al. 2014). Thus, the concept of destination image in this research will be based on Gartner’s approach.

The cognitive component is expressed through the set of beliefs and knowledge tourists have about the attributes of a destination (Baloglu and McCleary 1999a; Beerli and Martín 2004a, 2004b; Gartner 1993), while these beliefs do not necessarily stem from a previous visit to the destination (Pike 2002). The affective component refers to the phase of evaluation and emotional reaction by reflecting on the feelings the individual associates with the visited place (Baloglu and McCleary 1999a; Beerli and Martín 2004a, 2004b; Gartner 1993). The conative component is identified with the decision to visit the destination. It derives from the previous two components and “depends on the images developed during the cognitive stage and evaluated during the affective stage” (Gartner 1993, p. 196). Unlike the cognitive and affective dimensions, which always either precede the overall image or form the same, the conative dimension is examined in three ways: (1) as an integral part of the image, (2) through the conative image or, as some authors define this term—loyalty, where this dimension is seen as a dependent variable, and (3) the division of the conative image into the idealistic future that people want for themselves, the intention to return to the destination, and word of mouth, where the first component represents the conative image that precedes the overall image; the other two components are examined separately and affected by the image (Shafiee et al. 2016).

Although the three-component approach to image prevails in tourism literature, the research on the conative image remains limited. Previous studies focus on the cognitive, or either the cognitive or the affective, dimension to address the conceptualization of the image. So, the use of Gartner’s approach is still questionable after almost three decades. Therefore, this study intends to fully implement the three-component approach to the image.

2.2. Quality of Tourist Services

The quality of the products or services is important in all business areas. In tourism, as well, visitors demand high-quality services or, in other words, services worth money. Quality in tourism has become increasingly important because the tourism market is characterized by growing competition, lack of willingness to provide services, increasing loss of individuality through product standardization, unfavorable price ratio, etc. (Kachniewska 2006, p. 40). Service quality can be defined as “the customer’s overall impression of the relative inferiority/superiority of the organization and its services” (Bitner and Hubbert 1994, p. 77).

Unlike the destination image, within which two approaches are clearly identified, the number of dimensions within the quality of tourist services is inconsistent. However, the current tourism literature recognizes two ways when examining service quality. These are the overall destination/trip quality and the perceived quality of particular products/services in the destination, such as museum exhibitions (Forgas-Coll et al. 2017; Han and Hyun 2017), accommodation (Sharpley 2000), and online tourism platforms (Priporas et al. 2017).

Regarding the overall destination quality, depending on the research objectives, a great variety of dimensions were used in previous studies, which makes it difficult to identify its essential elements. For instance, Chen and Tsai (2007) analyzed trip quality in the Kengtin region in Taiwan by using five aspects: attractions, accessibility, activities, available packages, and ancillary services. Chen and Li (2018) also used the dimension of attractions; instead of accessibility, they measured transport, and they added hotels, restaurants, and shopping to measure the perceived service quality in Switzerland. Hailu (2015) examined service quality in Aksum (Ethiopia) by measuring attractions, accessibility, and accommodation, and he added three new dimensions: amenities, activities/entertainments, and value for money. On the other hand, Wang et al. (2017) explored perceived destination quality in Danang (Vietnam) with four items related to tourist offers and experiences in the destination.

Thus, it is crucial to consider the type of destination and its main characteristics when choosing the appropriate dimensions of quality of tourist services. It can be concluded that dimensions that were mostly considered to form or affect the quality of tourist services in a destination are attractions, accessibility/transport, and accommodation.

Accessibility to tourist attractions is one of the most important factors influencing tourists’ decisions regarding the choice of attractions to visit (Boniface and Cooper 2001). It can be defined as the ability of tourists to reach their destination safely (AlKahtani et al. 2015). In tourism studies, accessibility has been explored through various aspects, such as functionality, convenience, and transportation (AlKahtani et al. 2015), geographic accessibility (Hailu 2015), functions of tourist attractions, ancillary facilities within attractions, and the quality of the road network (AlKahtani et al. 2015).

The quantity and quality of the available accommodation capacities are significant factors for tourism development (Sharpley 2000). If tourists are satisfied with the accommodation and if they can enjoy the same or greater comfort in comparison to their home, they will gladly return to the destination and recommend it. For instance, Sharpley (2000) proved that, in addition to accommodation development, the process of planning in tourism was of great importance for the entire tourism industry. Hailu (2015) showed that tourists were generally satisfied with the accommodation in the city of Aksum, particularly with the staff, who showed a desire to provide additional information, and their kindness.

Tourist attractions are defined as “basic tourist resources that attract or can attract tourists to a particular destination” (Jovičić 2008, p. 144). In his definition, Pearce (1991) emphasized the role of destination management; so, a tourist attraction is “a named site with a specific human or natural feature which is the focus of visitor and management attention” (p. 46). Kušen (2002), on the other hand, sees tourist attractions as the basis for tourism development that largely determine its overall tourism offer.

2.3. Tourist Satisfaction

In order to establish a successful business, it must be based on consumer satisfaction. This is a basic construct that needs to be analyzed at all stages, from business establishment, through development, to, ultimately, improvement. Consumer satisfaction is the subject of numerous scientific studies. Although there are different definitions of consumer satisfaction in the literature, what the authors agree on is that the assessment of consumer satisfaction includes the fulfilment/non-fulfilment of their expectations (Baker and Crompton 2000; Yu and Goulden 2006). The most commonly used definition states that satisfaction is “the consumer’s fulfilment response. It is a judgment that a product/service feature, or the product or service itself, provided (or is providing) a pleasurable level of consumption-related fulfilment, including levels of under- or over-fulfilment” (Oliver 1997, p. 8).

In tourism, the visitors’ satisfaction is crucial for the further development of the tourist destinations. By measuring it, the guidelines for improving marketing and destination management can be defined most reliably. Tourist satisfaction is defined in different ways, but definitions are mostly based on an emotional/psychological outcome or a cognitive approach. Thus, tourist satisfaction is defined as “the psychological outcome which emerges from experiencing the service” (MacKay and Crompton 1990, pp. 8–9). Laws (1998) concluded that satisfaction is a cognitive understanding of one’s own experience that is influenced by various factors. The cognitive dimension is also emphasized by Ross and Iso-Ahola (1991), for whom satisfaction is a personal understanding of one’s own experience that arises from cognitive notions of the outcome.

Based on the above discussion, it can be concluded that, in order to build and maintain a competitive position in the volatile tourism market, it is necessary to constantly compare the expectations of tourists, which are related to the image of the destination, and the experience of tourists where the quality of services dominates (Castro et al. 2007). However, in previous research, no agreement has yet been reached on a model that would most accurately assess and measure tourist satisfaction.

3. Conceptual Framework and Hypotheses

Previous research has indicated that destination image positively affects the quality of tourist services (Alcañiz et al. 2005; Bigné et al. 2001; Chen and Tsai 2007; San Martín et al. 2019), and vice versa, the quality of tourist services influences the destination image (Akroush et al. 2016; Kim et al. 2017; Moon et al. 2011; Yamaguchi et al. 2015). Because, in this research, the quality of tourist services encompasses the quality of attractions, accessibility/transport, and accommodation, and the destination image is viewed through three components (cognitive, affective, and conative), it is more likely that quality is an antecedent of image. In other words, to be able to create a perception of a destination, tourists need to experience it, especially with regard to the affective and conative dimensions. Akroush et al. (2016) showed that four dimensions of tourism service quality, assurance–responsiveness, tangible facilities–empathy, reliability, and reliability–quality of directions, positively influence destination image in the Dead Sea destination in Jordan. Moreover, a study by Moon et al. (2011) in South Korea revealed that event quality perceptions have a positive impact on the cognitive, affective, and conative dimensions of the destination image.

Based on the above-mentioned literature, the following hypothesis is proposed:

Hypothesis 1 (H1). 

The quality of tourist services has a positive and direct effect on the image of a tourist destination (cognitive, affective, and conative).

There is extensive academic research dealing with the relationship between the destination image and tourist satisfaction, which has proven that the image has a positive effect on satisfaction (Assaker et al. 2011; Alcañiz et al. 2005; Bigné et al. 2001; Chen and Tsai 2007; Chi and Qu 2008; Lam et al. 2020; Wang and Hsu 2010). For instance, research evidence from Switzerland from inbound tourists has shown that destination image, consisting of environmental quality, political security, social connectivity, and economic affordability, directly affects tourist satisfaction (Chen and Li 2018). In addition, it has been confirmed that cognitive and affective destination image positively predict satisfaction, with cognitive image having a higher influence (Lam et al. 2020).

Three studies have investigated the image and satisfaction of tourists in Belgrade (Marković 2014; Todorović et al. 2017) and Serbia (Armenski 2014). However, none of these studies applied Gartner’s (1993) approach to the destination image. Marković (2014) and Todorović et al. (2017) studied the satisfaction of foreign tourists using the HOLSAT model. The research conducted by Marković (2014) showed that tourists are dissatisfied with the offer of museums and souvenirs, the places and rural areas near the city, the prices in stores, the dirtiness on the streets, the traffic jams, and the length of the waiting times to be served. On the other hand, tourists showed the highest level of satisfaction with the green areas in the city, the safety, and the friendliness. In addition, Todorović et al. (2017) also proved that foreign tourists in Belgrade are mainly satisfied with its safety and friendliness. Their study singled out one more satisfaction attribute: good value for money. Dissatisfaction was confirmed with the public transport, higher prices in shops, and socialist architectural style. Furthermore, Armenski (2014), in her model of loyalty, showed that the cognitive and affective dimensions of the image have no effect on satisfaction.

Based on the aforementioned literature, it can be concluded that there is a research gap in examining the impact of image on satisfaction, where image consists of three dimensions (cognitive, affective, and conative). Thus, the following hypothesis is defined:

Hypothesis 2 (H2). 

The image of a tourist destination (cognitive, affective, and conative) has a positive and direct effect on the overall satisfaction of tourists with the experience in the destination.

According to some authors, satisfaction precedes service quality (Casado et al. 2004; Patterson and Johnson 1993). In this sense, a satisfactory experience can influence the client’s attitude and assessment of the perception of the service quality. However, some authors do not agree with the previous approach and claim that the quality of service precedes customer satisfaction (Oliver 1997; Tsaur and Lin 2004). In this research, the second approach will be adopted due to the fact that tourists must “come” for the service. Bowen and Clarke (2002) have shown that measuring service quality and tourist satisfaction allows destination management to fully understand how to provide the best possible quality of service to keep tourists satisfied.

Several studies have investigated the relationship between service quality and tourist satisfaction. For instance, Osman and Sentosa (2013) proved that the quality of service has a significant impact on and a positive relationship with tourist satisfaction with rural tourism in Malaysia. Chen and Li (2018) supported the hypothesis that service quality, measured by the quality of hotels, restaurants, attractions, shopping, and transport services, influences tourist satisfaction in Switzerland. In addition, Wang et al. (2017) showed that the perceived quality of the destination of Danang City in Vietnam is stronger for the first visit than for later visits. On the other hand, some authors proved that the overall quality of tourism services, measured by one item, affects the satisfaction of tourists (Bigné et al. 2001; Castro et al. 2007).

Based on the above discussion, the following hypothesis is proposed:

Hypothesis 3 (H3). 

The quality of tourist services (accessibility, accommodation, and attractions) has a positive and direct effect on the overall satisfaction of tourists with the experience in the destination.

As previous studies have shown that the quality of tourist services positively and directly influences the destination image and has a further positive and direct effect on satisfaction, the following sequence: quality—image—satisfaction has been hypothesized in this paper. So, it can be argued that image mediates the relationship between quality and satisfaction. Consequently, the following hypothesis is established:

Hypothesis 4 (H4). 

The destination image partially mediates the relationship between the quality of tourist services and the overall satisfaction of tourists.

4. Methodology

4.1. Data Collection and Sample

The pilot study preceded the main study. It was conducted on a sample of N = 44 respondents. The research was reliable because the values of the Cronbach’s alpha coefficient for each subcategory were above 0.7, except for the cognitive image, where the value was above 0.6 (which is also acceptable). The main research was conducted at the same place as the pilot study, at Kalemegdan Park and Belgrade Fortress, over a six-month period, from May to October 2018. These are the main and the most visited tourist attractions in the city. The complex is protected by the Republic of Serbia as an Immovable Cultural Heritage of Exceptional Importance. Moreover, the Belgrade Fortress is on the Tentative List of UNESCO. Thus, this location is the most suitable for tourism surveys as it allows a global insight into tourists visiting the city.

The target population consisted of international millennial tourists who visited Belgrade in the aforementioned six-month period. International visitors are “one of the priority categories” of visitors to whom direct marketing activities in Belgrade need to be directed (Tourism Development Strategy of the City of Belgrade 2020–2025 2019, p. 123). The convenience sampling method was used because the participants were selected based on their readiness to participate in the survey. The sample size was 359.

4.2. Measurement of Constructs

The questionnaire consisted of a total of 61 questions (Appendix A). The formulation of scales was based on the existing literature and personal interviews with tourists during the pilot study. A set of 22 questions was chosen to determine the cognitive image; the questions examined the expectations of tourists before arriving at the destination (Agapito et al. 2013; Baloglu and McCleary 1999a, 1999b; Beerli and Martín 2004a, 2004b; Chen and Li 2018; Gartner 1993; Lam et al. 2020; Lee and Xue 2020; Marques et al. 2021; San Martín et al. 2019; Sanz-Blas et al. 2019; Schofield et al. 2020). The following bipolar pairs were selected for the affective image (Agapito et al. 2013): sleepy–arousing, unpleasant–pleasant, gloomy–exciting, and distressing–relaxing. The conative image was examined through a total of seven items related to future intentions (Agapito et al. 2013; Gartner 1993; Woosnam et al. 2020). The quality of tourist services encompassed 12 questions within the following components: accessibility, accommodation, and attractions (Chen and Tsai 2007; Chen and Li 2018; Žabkar et al. 2010). The general satisfaction of the tourists with their stay in the destination was explored through one question where the respondents assessed their overall impression of their stay in Belgrade. A 5-point Likert scale with values from 1 to 5 was used to measure all the items. The last part of the questionnaire consisted of a total of 15 questions concerning socio-demographic characteristics and previous visits.

4.3. Data Analysis

The methods used for data analysis include descriptive analysis, confirmatory factor analysis, and partial least squares structural equations modelling (PLS-SEM). The data were processed in statistical software SPSS 25, AMOS 25, and SmartPLS 3. The statistical software SmartPLS 3 is claimed to be significantly more convenient compared to AMOS when it comes to more complicated models with several factors (Wong 2013). Considering that the model contains third-order factors, the repeated indicator approach was used.

As the missing data rate was below 10% in this study, all imputation methods could be applied (Hair et al. 1998). However, the existence of categorical variables in the model significantly narrows the choice of methods, with the most suitable method being Fully Conditional Specification—FCS (Van Buuren 2007). Moreover, this method gives statistically valid estimates when it comes to models with a large number of variables (Dong and Peng 2013). Five imputations were performed as this value is usually stated as sufficient when there is a large number of variables (Dong and Peng 2013). Confirmatory factor analysis tested all five combinations of replaced data for both the image and the quality. The first combination was chosen because it gave the best results for the fit indices.

The data distribution was examined by the assumption of normality. The coefficients of horizontal and vertical deviation, skewness and kurtosis scores, as well as the Kolmogorov–Smirnov and Shapiro–Wilk tests, were used. The skewness score was in the range between −2 and +2, while the kurtosis score was below 3 for all the variables at the univariate and multivariate levels (Appendix B). That means that the data meet the assumption of normality (Westfall and Henning 2013).

5. Results

5.1. Demographic Profile of Respondents and Travel Characteristics

Descriptive analysis was used to give an overview of the structure of the respondents according to certain socio-demographic characteristics and their previous experience in the destination (

Table 1. Socio-demographic characteristics and previous visits of respondents.

Variable	n	%	Variable	n	%
Gender			Employment
Male	201	56.0	Student	190	52.9
Female	158	44.0	Employed	146	40.7
Age			Unemployed	21	5.8
18–20	66	18.4	Housewife	2	0.6
21–25	140	38.9	Monthly family income
26–30	100	27.9	<EUR 500	43	12.0
31–35	34	9.4	EUR 501–1000	56	15.6
36–37	19	5.3	EUR 1001–1500	38	10.6
Marital status			EUR 1501–2000	54	15.0
Married/living together	100	27.9	EUR 2001–2500	32	8.9
Single	249	69.4	EUR 2501–3000	34	9.5
Divorced	8	2.2	>EUR 3000	102	28.4
Widowed	2	0.6	Repeat visit
Education			Yes	121	33.7
Primary school	18	5.0	No	238	66.3
High school	77	21.4
Bachelor	143	39.8
Master	103	28.7
Ph.D.	18	5.0

).

Regarding gender, almost 56% were males, while 44% were females. In terms of age, although all the surveyed tourists were millennials, the dominant age group was aged between 21 and 25 years (38.9%). Because of the dominant age group of the respondents, the marital status variable share is inevitably dominated by singles (69.4%). Most respondents had a bachelor’s degree (39.8%), followed by those with a master’s degree (28.7%). In terms of employment, the majority were students (52.9%) and employed (40.7%). Data related to the monthly family income show that most of them had incomes above EUR 3.000 (28.4%). Regarding the structure of the respondents by the characteristics of travel, most were visiting Belgrade for the first time (n = 238; 66.3%).

5.2. Confirmatory Factor Analysis of the Destination Image

The component of a tourist destination image is presented as a reflective model consisting of 33 parameters and 10 first-order latent factors, as well as correlations between the latent factors. The image of the tourist destination defined in this way gave poorer fit indices (CMIN = 826.449 (p = 0.000), χ²/df = 1.837, GFI = 0.879, RMSEA = 0.048, SRMR = 0.054, CFI = 0.906, NFI = 0.817, TLI = 0.889). In order to better fit the data, four revisions of the residual matrix and the modification indexes were performed. After each revision, the fit indices were recalculated until their satisfactory values were reached (CMIN = 96.312 (p = 0.049), χ²/df = 1.284, GFI = 0.966, RMSEA = 0.028, SRMR = 0.035, CFI = 0.990, NFI = 0.956, TLI = 0.986). Thus, the final model of a tourist destination image was formed, 15 parameters within six factors (Appendix C).

Convergent validity was assessed based on the statistical significance of the factor loadings by examining the composite reliability (CR) and the average variance extracted (AVE). The factor loadings of all the parameters were above 0.6, which did not confirm the proposal of the authors Hair et al. (1998) to exclude from the analysis all items that have a value of this index below 0.7. All the values of the factor loadings showed statistical significance (p < 0.01).

The validity of the tourist destination image model is presented by the correlation matrix in

Table 2. Correlation matrix of validity of tourist destination image model.

	CR	AVE	MSV	MaxR(H)	F3	F7	F4	F9	F10	F8
F3	0.739	0.587	0.351	0.743	0.766
F7	0.763	0.519	0.221	0.778	0.225 ***	0.721
F4	0.777	0.541	0.351	0.797	0.593 ***	0.262 ***	0.735
F9	0.785	0.646	0.318	0.789	0.218 **	0.190 **	0.274 ***	0.804
F10	0.781	0.642	0.337	0.796	0.389 ***	0.188 **	0.305 ***	0.564 ***	0.801
F8	0.880	0.711	0.337	0.884	0.332 ***	0.470 ***	0.410 ***	0.505 ***	0.581 ***	0.843

** p < 0.010; *** p < 0.001.

. After the fourth modification, it was found that all the factors in the model achieved convergent and discriminant validity. That is to say, the CR values of all the factors were above 0.7, the AVE values above 0.5, and the following relationship was confirmed: AVE > MSV and AVE > ASV (Fornell and Larcker 1981; Gaskin and Lim 2016). The following factors were retained: F3—Natural environment, F4—Tourist environment, F7—Affective components, F8—Word of mouth and recommendations, F9—Value for money, and F10—Revisits.

5.3. Confirmatory Factor Analysis of the Quality of Tourist Services

The construct of the quality of tourist services is presented as a reflective model consisting of 12 parameters and 3 latent factors of the first order, as well as the correlations between latent factors. The quality of the tourist services defined in this way gave poorer suitability indices (CMIN = 269.525 (p = 0.000), χ²/df = 5.285, GFI = 0.896, RMSEA = 0.109, SRMR = 0.073, CFI = 0.837, NFI = 0.808, TLI = 0.789). Three revisions of the residual matrix and modification indexes were performed until satisfactory values of the fit indices were reached (CMIN = 20.786) (p = 0.004), χ²/df = 2.969, GFI = 0.981, RMSEA = 0.074, SRMR = 0.035, CFI = 0.981, NFI = 0.972, TLI = 0.960). In this way, the final model of the quality of tourist services was established, six parameters within two factors (see Appendix D).

The validity of the model of the quality of the tourist services is presented by the correlation matrix in

Table 3. Correlation matrix of validity of tourist services quality model.

	CR	AVE	MSV	MaxR(H)	F2	F3
F2	0.824	0.542	0.295	0.843	0.736
F3	0.730	0.574	0.295	0.730	0.543 ***	0.758

*** p < 0.001.

. After the third modification, it was found that both factors in the model achieved convergent and discriminant validity. The following factors were retained: F2—Quality of accommodation and F3—Quality of accessibility and attractions.

5.4. Evaluation of the Measurement Models

Evaluation of the measurement models involves examining the reliability and validity of the measurement models of the first-order constructs: natural environment and activities (F3_Cog), tourist environment (F4_Cog), word of mouth and recommendations (F8_Con), value for money (F9_Con), and revisits (F10_Con); then, the constructs of the second order: cognitive image (Cog_Image), affective image (Aff_Image), conative image (Con_Image), quality of accommodation (Qual_Accomm), and quality of accessibility and attractions (Qual_Acc_Attr); and the third-order constructs: the image of the tourist destination (Image) and the quality of the tourist services (Quality). Figure 3 presents all the measurement models with factor loadings for each individual indicator within the corresponding construct, the R2 values, and the path coefficients for the image-satisfaction and quality-satisfaction relationships.

When it comes to the results of internal consistency and the values of average variance extracted (AVE) for the hypothetical satisfaction model, all variables had satisfactory values of Cronbach’s alpha (>0.7), ρA (>0.7) and ρC (>0.7), while the lower limit of the AVE value of 0.5 was not met by one variable (Image—0.328). By omitting items with a factor loading below 0.7 for Image, the lower limit of AVE could not be reached for any of the listed variables (Hair et al. 2017). Taking into account that these are the third-order constructs and that some authors (Borsboom et al. 2004) dispute the validity of this coefficient, all the parameters were retained. According to the recommendation by Gaskin (2017), if the correlation matrix HTMT shows satisfactory values, below 1, and preferably below 0.9 (Hair et al. 2017), the convergent and discriminant validity are confirmed.

Discriminant validity was examined using the Heterotrait–Monotrait correlation ratio (HTMT). As all the HTMT values were below the upper limit of 0.9, it can be concluded that all the constructs in the model are conceptually different (Appendix E). As this is a higher-order hierarchical model, the discriminant validity between the lower-order indicators and the corresponding higher-order factors was not taken into account. “A violation of discriminant validity between these constructs is expected because the measurement model of the higher-order component repeats the indicators of its two lower-order components” (Sarstedt et al. 2019, p. 203). Another procedure for statistically verifying these results involves bootstrapping (subsamples = 5000), which sought to test the null hypothesis (H0: HTMT ≥ 1) against the alternative hypothesis (H1: HTMT < 1) (Henseler et al. 2015). All the HTMT values were less than 1 in the 95% confidence interval, confirming discriminant validity.

5.5. Evaluation of the Structural Model

The analysis of the measurement models was followed by the analysis of the structural model. The assessment of collinearity was conducted in statistical software SPSS 25. The results of the regression analysis showed that there was no dependence between the constructs because the VIF values were between 0.2 and 5 (Hair et al. 2017) (

Table 4. Tolerance and Variance Inflation Factor (VIF) values.

	Tolerance	VIF
Image	0.802	1.247
Quality	0.802	1.247

Dependent variable: Satisfaction.

). Moreover, “if all VIFs resulting from a full collinearity test are equal to or lower than 3.3, the model can be considered free of common method bias” (Kock 2015, p. 7). The bootstrapping procedure (subsamples = 5000) was used to check the statistical significance of the path coefficients (t-values and p-values). All the relations were statistically significant (t > 1.65, p < 0.01), and there was a moderate correlation between the destination image and satisfaction (0.442) and the quality of tourist services and destination image (0.429), while there was a low correlation between the quality of tourist services and satisfaction (0.291).

The statistical significance of the coefficient of determination (R2) of the dependent variables, image and satisfaction, was significant (p < 0.1) and had values of 0.184 and 0.390, respectively. The statistical significance of the F-test confirmed the strength of the structural model (t > 1.96). Therefore, it is concluded that 18% of the endogenous image variable is caused by the exogenous quality variable. In addition, 39% of the variation of the endogenous satisfaction variable is caused by the exogenous variables, image and quality.

The result of the f2 effect size shows whether the omitted exogenous construct substantially affects the endogenous construct. The influence of the exogenous image construct on the endogenous satisfaction construct was medium (0.261), while the influence of the exogenous quality construct on the mentioned endogenous construct was small (0.114). Thus, the image construct better predicts satisfaction variations compared to the quality construct.

The coefficients of predictive relevance (Q2) and the q2 effect size were tested using the blindfolding option. All the values of the coefficient Q2 were above 0, which confirms the predictive relevance of the model in relation to the endogenous latent variable. The influence of the exogenous image construct on the endogenous satisfaction construct was small (0.109), while the influence of the exogenous quality construct on the endogenous satisfaction construct was medium (0.238).

In the hypothesized satisfaction model, one mediator connection was defined: the influence of the quality on satisfaction through the image. The direct influence of quality on satisfaction is significant (β = 0.291, p < 0.01), as is the indirect influence of quality over the destination image (β = 0.190, p < 0.01). The variance accounted for (VAF) shows the extent to which the mediation process explains the variation of the dependent variable. If this value is below 20%, mediation does not exist; values between 20% and 80% show partial mediation, while mediation is complete if the value is above 80% (Hair et al. 2017).

5.6. Hypotheses Testing

The results of the analysis of the direct effects in the hypothesized model are presented in

Table 5. Results of the analysis of direct effects in the hypothesized model.

Hypothesis	Path Coefficient	t-Value	p-Value	Inference
H1: Quality → Image	0.429 [0.349; 0.499] *	9.342	0.000	Supported
H2: Image → Satisfaction	0.442 [0367; 0.518] *	9.525	0.000	Supported
H3: Quality → Satisfaction	0.291 [0.201; 0.372] *	5.716	0.000	Supported

* 95% bias-corrected confidence intervals.

. From all the direct effects, the influence of the destination image on satisfaction had the highest value of the path coefficient (β = 0.442). This means that image directly and positively influences satisfaction, and the strength of this relationship is moderate. So, the respondents that evaluated the destination image as more favorable were more satisfied with the overall stay in the destination. Past research has also confirmed this relationship (Assaker et al. 2011; Alcañiz et al. 2005; Bigné et al. 2001; Chen and Li 2018; Chi and Qu 2008; Lam et al. 2020; Wang and Hsu 2010).

Tourism satisfaction has been the subject of intense academic research, both from the aspect of the factors that affect it and from the aspect of the further influence of satisfaction on certain factors. This research also confirmed the hypothetical relationship between the quality impact on satisfaction. The path coefficient showed a weaker relationship than the image in the overall satisfaction model (β = 0.291). The connection is positive and moderate. Other studies in tourism have also addressed the impact of quality on satisfaction, but to a much lesser extent than when it comes to image (Bigné et al. 2001; Castro et al. 2007; Osman and Sentosa 2013; Rahman et al. 2017; San Martín et al. 2019; Wang et al. 2017).

When it comes to the influence of the quality of tourist services on image, the path coefficient showed a positive and moderate relationship (β = 0.429). Hence, the tourists that were satisfied with the information about the destination accessibility and attractions, as well as the quality of accommodation, created positive image perceptions. This is consistent with previous studies (Akroush et al. 2016; Kim et al. 2017; Moon et al. 2011; Yamaguchi et al. 2015).

In order to explore the mediating effect of the destination image on the quality–satisfaction relationship (

Table 6. Results of the analysis for mediating effect in the hypothesized model.

Hypothesis	Direct Effect	Indirect Effect	Total Effect	VAF	Inference
H4: Quality → Satisfaction (via Image)	0.291 ***	0.190 ***	0.481 ***	39%	Partial mediation

*** p < 0.01.

), it is necessary to examine both the direct and the indirect effects of the quality on satisfaction. Both effects were significant (β = 0.291, p < 0.01; i.e., 0.190, p < 0.01, respectively), and this is the first prerequisite for partial mediation. Moreover, the VAF value is 0.395, which shows that almost 40% of the satisfaction is explained by the mediating influence of quality through image, which is above the lower limit of 20% (Hair et al. 2017).

6. Conclusions

This study sought to empirically prove the relationship between (a) quality and image, (b) image and satisfaction, (c) quality and satisfaction, and (d) the mediating role of image on the quality–satisfaction relationship. Moreover, the focus was on millennial international tourists in the emerging destination of Belgrade as it is located in the southeastern part of Europe where all six emerging economies have been recording growth in international tourist arrivals in the 21st century (OECD 2018). In addition, a vision of Belgrade tourism development consists in promoting its “…cosmopolitan richness of diversity and clearly preserved local peculiarities…” (Tourism Development Strategy of the City of Belgrade 2020–2025 2019, p. 105). However, to the best of the authors’ knowledge, there has been no comprehensive official research on the factors that influence tourist satisfaction, and thus, it has been difficult to establish an adequate focus of promotional activities on attractions that motivate and attract tourists to visit Belgrade. Research conducted for the purposes of drafting the Tourism Development Strategy of the City of Belgrade 2020–2025 (2019) includes only the examination of international tourists’ satisfaction with certain components of the stay. The same strategy highlights the necessity of “improving data collection system of tourism demand” (Tourism Development Strategy of the City of Belgrade 2020–2025 2019, p. 118). This supports Kim’s (2018) argument that destination management organizations (DMOs) usually conduct insufficient research on tourists’ satisfaction, making it more difficult to create appropriate marketing strategies. Therefore, it is essential to examine tourism demand in Belgrade more thoroughly as negative political circumstances have created a rather negative image of the city. Moreover, the country is still in the transition process, and the quality and quantity of accommodation capacities are far below that of European cities. In this context, destination image first, and then service quality, which are a basis of attractiveness in a destination (Kim 2018; Wang et al. 2017) and further influence its competitiveness, have to be explored. Furthermore, millennials’ preferences are crucial to examine as they travel the most in Europe (Ketter 2021). Moreover, they seek novelty and less-known destinations, such as Belgrade, where they will have authentic experiences (Ketter 2021). By examining the above-mentioned relationships and millennials’ perceptions regarding destination image and service quality, the current study intends to provide useful insights for an effective marketing strategy and an appropriate business tourism model in the emerging destination of Belgrade.

First of all, in the upcoming period, it is necessary to focus tourism development activities on forming a model that will involve the mutual cooperation of all stakeholders in tourism. Tourist destinations are complex systems, and there is no one-size-fits-all management model. Using the same management model in different destinations can give different results, and despite the negative image and modest quality of tourist services in Belgrade, the foreign tourists showed relatively high levels of satisfaction for all the examined constructs. The average values range from 2.20 to 3.99, which means that there is still plenty of room for further improvement of the overall tourist offer. Considering that both destination image and service quality positively influence tourists’ satisfaction, that service quality affects destination image, and that the indirect effect of the quality of the tourist services on tourists’ satisfaction is confirmed via the mediation of destination image, useful guidelines for future tourism development emerge. They are related to destination identity creation, marketing strategy formation, and appropriate business model selection. Tourists that experience satisfactory levels of the quality of services will create a positive overall destination image that consists of cognitive, affective, and conative components and, in turn, will be more satisfied with their stay in the destination. If the actual service exceeds their expectations, it will directly result in higher levels of overall satisfaction. In the example of Belgrade, destination identity should be built upon the natural environment (natural scenic beauty and a variety of flora and fauna), the tourist environment (kind people, safety and security, peaceful and restful atmosphere), the quality of information regarding accessibility and attractions, as well as the quality of accommodation. Marketing activities directed toward millennials are almost always successful (Chan 2018). As they are highly technology savvy (Ketter 2021), social media channels are the most effective way to present destination identity and to further attract this demographic group. These conclusions are discussed more profoundly below through theoretical and practical implications, along with the limitations and recommendations for future research.

6.1. Theoretical Implications

The results of this research can significantly contribute to the development of theories in the field of the image of tourist destinations, the quality of tourist services, and the overall satisfaction assessment. In particular, certain relationships in the applied satisfaction model were analyzed for the first time: (1) the impact of image when, in addition to the cognitive and affective dimension, it also comprises a conative dimension and as such affects quality and satisfaction and (2) the impact of quality that consists of accessibility, accommodation, and attractions, which together affect satisfaction. On the other hand, it provides useful insights for understanding millennials’ preferences regarding destination image and the quality of tourist services in emerging destinations.

Previous research is mainly based on future intentions as a component affected by the image (Chen and Tsai 2007; Wang and Hsu 2010). This research is the first to view the image as a whole made up of three dimensions, where the conative dimension can, to some extent, be identified with future intentions (word of mouth and recommendations and revisits). That is to say, the tourists’ destination loyalty is usually defined as the intention to revisit a destination and recommend it to others. However, in this study, it is a part of a conative image where one more dimension was added—value for money. It is worth mentioning that the conative image, as such, is rarely included in tourism studies (Woosnam et al. 2020) even though there is evidence that in addition to the intention to revisit a destination, the conative image should include the intention to recommend the destination and word of mouth (Agapito et al. 2013).

The empirical evidence of this study shows that the destination image is affected by the quality of the tourist services, which means that quality is an important factor in creating quality judgments during the stay in the destination by the millennial generation. Furthermore, image mediates the quality–satisfaction relationship and directly and positively impacts satisfaction. This indicates that, when deciding which direction a destination should choose for its tourism development, it is necessary to conduct comprehensive research on the image and service quality. In other words, it is essential to see which of the following factors, environmental, physical, or social, attracts tourists the most and creates the attractiveness basis of an area (Butler 1980). Considering that the quality of services provided in tourism can significantly contribute to the development of destinations and increase competitiveness, this study offers an integrated approach for a better understanding of this construct through the dimensions of accessibility, accommodation, and attractions.

6.2. Practical Implications

From the point of view of tourism management, the results of this research are valuable for all tourism stakeholders in emerging destinations targeting millennial tourists. The broad implication of the present research is that overall image, consisting of cognitive, affective, and conative dimension, as well as service quality, comprised of the accessibility, attractions, and accommodation dimensions, should be the starting point when developing a tourism marketing strategy and business model of emerging destinations. There are various emerging city destinations in the southeastern part of Europe. However, they have not gained much attention in the academic literature. The study that considered different image components in the context of destination branding in emerging city destinations is the one conducted by Marques et al. (2021) in Sofia, Bulgaria. These authors stated that it is necessary to conduct research on image and trip quality in other emerging capital cities in this part of Europe, mentioning Belgrade, among others. In this regard, the research conducted in Belgrade gives valuable insights for positioning tourism strategies in this type of destination.

When it comes to image, the conative image has the greatest influence on forming the overall destination image. In this context, the tourists who visit the destination will be a kind of representative of it after leaving it. Thus, it is imperative to take into account the overall impression the destination leaves on the individual. This impression is primarily created based on service quality. Although it has been confirmed that millennial tourists mostly choose cheaper accommodation options (Ketter 2021), the results of this study show that the quality of accommodation is of major importance for tourists. So, it is necessary that tourism managers, in cooperation with state institutions, follow the trend of tourist needs and, accordingly, invest in the construction of new, and in the adaptation of existing, accommodation facilities. The quality of the accessibility and the attractions also greatly influences the formation of attitudes about the quality of the services. In this sense, it is essential to improve the content on the institutional tourism channels and social networks, as well as to set up tourist information points on important tourist routes.

Furthermore, the cognitive image is the next significant image dimension that influences its formation. The factors such as the natural environment and the tourist environment stand out here. Thus, the prior knowledge of tourists related to natural beauty, safety, security, the kindness of people, and a relaxing atmosphere are the primary components of the cognitive image. Although the affective component has a much smaller impact compared to the previous two dimensions, it has shown that tourists prefer destinations where they feel comfortable and that abound in interesting and exciting content.

Overall, the results of this study reveal that Belgrade’s identity should be built on natural attractiveness, the tourist environment, and uniqueness (e.g., the nearby mountain Avala, the parks, the confluence of the rivers Sava and Danube, the kind people, the safety and security, the offer of local cuisine, etc.). In addition, accessibility and accommodation are other factors that require an adequate direction of marketing activities. These findings provide a potential mechanism for defining a suitable tourism business model, which Belgrade still lacks (Tourism Development Strategy of the City of Belgrade 2020–2025 2019). It should be based on the public–private partnership because the key factors influencing the tourists’ satisfaction are partly in public and partly in private ownership. Therefore, the cooperation between the public and private sectors in making important decisions regarding tourism development is a desirable way of creating a tourism business model.

6.3. Limitations and Future Research

There are several limitations associated with this study. First, all the respondents were millennials who were aged 18–37 in 2018 when the research was conducted. Choosing only one demographic group, in this case based on age, remarkably limits the study because tourists of other ages also visit the destination. Although they are not as numerous as millennials, investigating their levels of satisfaction with their stay in the destination can significantly determine future tourism development. Second, conative image is identified with future intentions. As such, it is one of the three dimensions in the model, apart from the cognitive and the affective, which forms the image that then affects satisfaction. By setting it as one of the components that precedes satisfaction, the influence of satisfaction on future intentions is excluded, although the mentioned connection is important for directing the development of tourist destinations. Third, the satisfaction construct was measured by only one parameter because the constructs of image and quality were also measured in terms of satisfaction with certain attributes, and it was estimated that, in the proposed model, it was best to assess the general satisfaction of tourists staying in the destination. Although there are previous studies that measure satisfaction in this way (Baker and Crompton 2000; Chi and Qu 2008; Yu and Goulden 2006), Hair et al. (2017) suggest that each construct should be measured by using three or more items, because it is the minimum necessary to achieve validity.

On the one hand, this study provides an important insight into the study of tourism satisfaction and the impact that image and quality have on it. On the other hand, it also offers several guidelines for future research. Firstly, future studies may focus on wider demographic groups, not only millennials. Secondly, the proposed satisfaction model can be applied to other emerging tourist destinations. This primarily refers to city-break destinations that would have less variation when it comes to items that determine constructs compared to other types of tourist destinations. This would help to reliably determine the extent to which image and quality are related to satisfaction, as well as to form reliable indicators and perform their further validation. Thirdly, the research period could cover the part of the year when the tourist season is not at its peak, e.g., during the period around the New Year holidays when the number of visits increases. In this way, the results could be compared, and the similarities and differences between the same surveys in different parts of the year could be identified. Fourthly, the proposed model saturation can be avoided by introducing additional constructs. For instance, there is an opportunity to analyze the impact of revisits on the image. On the other hand, the conative image can be seen as an idealistic future that people want for themselves, the intention to return to the destination, and word of mouth (Shafiee et al. 2016), with the first component being a conative image that precedes the overall image and the other two components affecting satisfaction. Finally, although this research tested a different conceptual model than the one that Marques et al. (2021) proposed for emerging city destinations, it provides an extremely useful basis for further research that could include unique destination characteristics.