Customer loyalty is a recurring topic in academic research. Since 1956, more than three thousand articles have been written on “customer loyalty” in journals with an impact factor in the area of social sciences, according to the results offered in the Web of Science (WOS). However, there are only ninety-five articles on “destination loyalty”, the first of them dating from 2001. Most of these studies try to find out the factors which antecede or explain the tourist's loyalty toward a destination (Bigné et al., 2001; Chen & Tsai, 2007; Faullant et al., 2008; Olorunniwo et al., 2006; Poria et al., 2011; Um, Chon, & Ro, 2006; Wang et al., 2010; Yuksel et al., 2010).

According to the literature, the benefits of loyalty are related with the current and future value of the benefits and with the organization's continuity (Moliner, Gil, & Ruíz, 2009). In this vein, the tourist's loyalty implies a stable source of incomes, as well as an improvement of the destination's benefits as customer retention means less costs than those required to attract new customers.

Most works define destination loyalty in relation to the intention to revisit and the intention to recommend (Chen & Gursoy, 2001; Chi & Qu, 2008; Chi, 2011; Oppermann, 2000; Pike, 2010; Prayag & Ryan, 2012; Yoon & Uysal, 2005). Loyalty is multidimensional, behavioral and attitudinal components having the greatest acceptation. From the first perspective, loyalty is reflected in repeat purchases, while attitudinal loyalty considers recommendations to friends and family. If both aspects are considered the construct is predicted better (Dimitriades, 2006) and a more exact representation of the tourist's loyalty is provided (Chen & Gursoy, 2001). When exclusively contemplating a tourist's behavior through repeated visits to a destination, this may not really reflect loyalty in the area which we are dealing with. When tourists travel with the motive of having a new experience, it is unlikely for them to repeat the visit to a destination, so destination loyalty may not involve repetition (Iso-Ahola, 1982).

Destination loyalty is a key element in marketing strategies in so far as it is the best predictor of post-visit behavior (Chen & Chen, 2010). This loyalty is achieved by surpassing the tourists’ expectations, providing unique experiences and assuming a commitment with these tourists. At the same time as valuing their opinions and reducing the effort to deal with the problems which arise during the relationship, they must be considered as unique customers (McGarry, 1995). The response of these tourists is seen in their favorable intention to revisit the destination and to give positive references about it in their close environment (Mohamad, Ali, & Ghani, 2011).

The intention to return to the destination is considered to be a key component of loyalty and many authors have gone thoroughly into this concept, trying to determine the factors which enable attaining high values in the probability of repetition (Backman & Crompton, 1991; Cronin, Brady, & Hult, 2000; Petrick, Morais, & Norman, 2001). In the field of tourism there are various works which have used this indicator as a measure of loyalty (Kim, Eves, & Scarles, 2009; Petrick, 2004; Qu, Kim, & Im, 2011).

The likelihood of recommending the destination to other people is considered these days as a clear indicator of loyalty. The intention to recommend refers to “the informal person-to-person communication between a perceived noncommercial communicator and a receiver regarding a brand, a product, an organization, or a service” (Harrison-Walker, 2001, p. 63). This intention to recommend the destination is of great relevance for its success as the high perceived risk in the decision of choosing the tourist destination brings about a strong need for qualified information and trust in the tourist. In this way, the references received from third parties are going to affect the choice of the destination (Kivela & Crotts, 2006; Nam, Kim, & Hwang, 2016; Qu et al., 2011; Tussyadiah, Park, & Fesenmaier, 2011).

The literature has tackled the study of a broad set of antecedents to loyalty, among others: image, perceived quality, satisfaction (Bigné et al., 2001), the safety of the destination, cultural differences, previous experiences (Chen & Gursoy, 2001), the motive for traveling and satisfaction with the trip (Yoon & Uysal, 2005). Most of the previous research in this field points out that the tourist's satisfaction, the destination's image and the quality of service are predictors of loyalty. A recent study has shown that satisfaction, perceived value and the destination's image, among others, are antecedents to the destination loyalty of domestic tourists in China (Sun et al., 2013).

Expectations: expectations play a central role in the discussion of customer satisfaction and service quality. Parasuraman, Berry, and Zeithaml (1988) define expectations as predictions made by consumers about what is likely to occur during an imminent transaction, or as beliefs about attributes which firms in general should have. Numerous studies centered on the tourist sector have found a direct relation between the customer's expectations and satisfaction (Oliver & Burke, 1999; Omachonu, Johnson, & Onyeaso, 2008; Voss, Parasuraman, & Grewal, 1998), and between this and loyalty (Prayag, Hosany, & Odeh, 2013). Although there are no studies which directly relate expectation with loyalty, it can be considered that the latter will be affected, at least indirectly, by the tourist's expectations regarding the destination.

Service quality: In the broad sense, service quality is “a global judgment, or attitude, relating to the superiority of the service” (Parasuraman, Zeithaml, & Berry, 1988, p. 16). In tourism, many works analyze the role of service quality in forming behavioral intentions (Baker & Crompton, 2000). Service quality has been related with satisfaction (Chi & Qu, 2008; Fornell, Johnson, Anderson, Cha, & Bryant, 1996; Mason & Nassivera, 2013; Petrick, 2004), at the same time as being a variable which favors word-to-mouth comments and contributes to the repetition of the visit (Cole & Illum, 2006).

Tourist service quality involves the valuing of the aspects which are under the control of the supplier, inasmuch as from a global perspective, the quality of the tourist experience is associated with psychological benefits and affective satisfaction (Cole & Crompton, 2003). In the tourist sector, there are many studies which find a positive relation between perceived quality and satisfaction with the accommodation (Alexandris, Dimitriadis, & Markata, 2002; Clemes, Gan, & Ren, 2011; Ingram & Daskalakis, 1999).

Satisfaction: Satisfaction is one of the most important concepts in the marketing literature due to it being a fundamental element for the survival of any business (Anderson & Sullivan 1993; Chiou, Droge, & Hanvanich, 2002; Kozak, 2001; Sun et al., 2013; Zeithaml, Berry, & Parasuraman, 1996). Diverse works in the area of tourism have dealt with the study of the tourist's satisfaction (Song, Veen, Li, & Chen, 2012; Sun et al., 2013). According to these works, satisfaction is a requirement for the success of a tourist destination due to this satisfaction being one of the most important predictors of a tourist's loyalty (Ozdemir, Aksu, Ehtiyar, Cizel, Cizel, & Icigen, 2012). The tourist's global satisfaction is considered to be the most important antecedent of tourist destination loyalty (Chi & Qu, 2008; Chi, 2012; Mohamad, Ali & Ghani, 2011; Neal & Gursoy, 2008).

In the tourist area, satisfaction comes from tourists comparing their expectations with the results of the experience of the trip, the experience surpassing the expectations (Kim & Perdue, 2011).

The consequences of satisfaction with a tourist experience are basically an increase in the intention to return to the destination as well as recommending it to third parties (Back & Parks, 2003; Back, 2005; Chi & Qu, 2008; Chi, 2012; Clemes et al., 2011; Kozak & Rimmington, 2000; Kozak, 2001; Petrick et al., 2001; Prayag & Ryan, 2012; Yoon & Uysal, 2005). Both the repetition of the visit and the recommendation are considered two key indicators of customer loyalty, a consequence of the tourist's satisfaction (Prayag, Hosany, & Odeh, 2013). Satisfaction has also been related with the tourist's perception of a destination's quality and image (Chi, 2012; Song et al., 2012).

Destination image: Image management is one of the most important responsibilities of any destination due to the visitors’ image of a destination having a strong impact on their evaluation and the behavior of selecting a destination (Chi, 2012; Nunkoo, Smith, & Ramkissoon, 2013; Sun et al., 2013). Through promoting a tourist destination it is meant to project an image which makes this destination attractive and desirable for potential tourists (Beerly & Martín, 2004).

One of the first definitions suggests that the destination image can be defined as the expression of all the knowledge, impressions, prejudices and emotional beliefs that an individual or group has concerning a specific place or object (Lawson & Baud-Bovy, 1977). Many conceptualizations of the concept of destination image have been contributed over time based on the research approaches used. Nevertheless, three inter-related components can be identified (Martín, Beerli, & Nazareno, 2016): 1) perceptual/cognitive, related to tourists’ beliefs about a destination's attributes; 2) affective, related to the emotional response of the tourists about the destination; and 3) global, which refers to the global impressions that the tourist has of the destination. The destination image can be defined as “the perceptions about the place as reflected by the associations held in tourist memory” (Cai, 2002, p.723). In the same line, Assaker, Vinzi, and O’Connor (2011) affirmed that they consider the destination image to be a holistic perception.

The destination image influences the tourists’ future intentions, be it their intention to revisit the destination or recommend it (Bigné et al., 2001; Chen & Tsai, 2007; Mahasuweerachai & Qu, 2011). It is therefore an antecedent of the tourist's destination loyalty (Ashworth & Goodall, 1988; Chen & Tsai, 2007; Cooper, Fletcher, Gilbert, & Wanhill, 1993; Lee, Lee, & Lee, 2005; Mansfeld, 1992), and a key element to consider in measuring this loyalty.

An index is a composite measure which summarizes various indicators or specific observations about the same phenomenon. Its design responds to the need to synthetically and simply express the variation in a set of values which in their primary form are difficult to understand. A loyalty index must therefore be a synthetic value, established in a scale, which enables the quantification of loyalty and that includes all the relative elements which determine it.

The majority of loyalty indices work on the perceptions which customers have about the topics included in surveys used to measure the observable variables, and tend to include three variables: global satisfaction, repurchase likelihood, and the probability of recommending the product to other people (Best, 2007; Geng-Qing & Qu, 2008). However, given the differential characteristics of destinations and fruit of the theoretical review carried out, we propose the following expression for the tourist loyalty index (TLI):

• TLI=EXPT×PQ×GID×TSI×IRECOM×IREP

• where

• TLI=tourist loyalty index

• EXPT=expectations

• PQ=perceived quality

• GID=global image of the destination

• TSI=tourist's satisfaction index

• IRECOM=intention to recommend

• IREP=intention to repeat the visit

The indices of loyalty are quantified on a scale between zero (minimum value) and a hundred (maximum value). In them, the customer satisfaction index (CSI) tends to be expressed based on 100, the rest of the variables referring to a unit basis. Following this model, all the variables were appropriately recoded, in such a way that the value five of the initial scale became the value one on the final scale (a hundred in the case of the satisfaction variable) and the value one of the initial scale was recoded to the value zero of the final scale. Having recoded all the variables, the tourist loyalty index (TLI) was calculated applying the expression proposed (TLI=EXPT*PQ*GID*TSI*IRECOM*IREP). Table 3 shows the descriptive statistics for the TLI and its components, considering the number of valid questionnaires, the average value of each of the indicators, the mode, and also the minimum and maximum values that the indicators attain.

The loyalty-overnight stays matrix is an adaptation for our study of the matrix proposed by Best (2007) to manage customer loyalty and profitability. This instrument enables the classifying of the customers of any organization into four types, according to their loyalty and profitability, using the averages of both variables as cut-off values. In this case, we have used as a variable that summarizes loyalty the TLI calculated in the previous phase and as a variable summarizing profitability the number of overnight stays of the tourists in their visit to the destination. Table 4 shows the results obtained and the descriptive profile of the tourists which are included in each group.

Based on both variables, TLI and overnight stays, and setting out from the loyalty-profitability matrix (Best, 2007), we establish the following categories of tourists:

• •

  Unprofitable tourists: these are the ones who spend less nights in the destination than the sample's average value and who also show a loyalty below the average. Normally, they are new to the destination and the majority do not have the intention of returning. They are 24.7% of the sample analyzed and have an average TLI of 28.60 and an average number of overnight stays of 1.6 nights. Referring to their sociodemographic profile, we note that most are from Spain, are between 26 and 44 years old and have secondary school studies. Regarding the characteristics related with their behavior when traveling, we point out that only 13% travel more than three times per year, that 41% are on their first visit to Seville, that they came by plane (42%) or by car (41%), and that 54% organized their own trip, without using an agency or intermediary. Lastly, and related to the TLI (Table 5), these are the tourists who have the least probabilities of repeating their visit and of recommending the destination to their close relatives. We also note reduced values in the indicators of global image, of expectations, and of perceived quality. All this is reflected in the level of satisfaction with the destination, the least of the four groups of tourists analyzed.

  Table 5.

  TLI structure.

  	Unprofitable	High potential	High performance	Underutilized
  GID	0.881	0.822	0.975	0.985
  EXPT	0.787	0.666	0.987	0.980
  PQ	0.791	0.843	0.992	0.988
  TSI	85	88.6	97.8	99
  IRECOM	0.791	0.942	1	1
  IREP	0.652	0.852	1	1
  TLI	28.6	33	93.4	94.5
• •

  Underutilized tourists: these tourists are 47.9% of the sample and are characterized as having a higher than average loyalty (TLI=94.50) and a figure of overnight stays below the average (1.6 nights). Their sociodemographic profile shows that they are between 26 and 44 years old, have had secondary education and reside in Spain. Concerning how they travel, 21.1% do so more than three times per year, 44% are visiting Seville for the first time, and have come by plane (50%) or by car (30%), and 48% have organized their own trip, without the intervention of an agency or intermediary. In these aspects, they are differentiated from the unprofitable tourists, as they travel more, tend to use planes more and have more trust in agencies and other intermediaries to plan their trip. If we analyze the TLI structure for this group of tourists, we note that they are the most loyal of the sample, having maximum values in recommendation and repetition. The indicators for global image, satisfaction, expectations and perceived quality are also very high, which means that we have a group of people who will be intrinsically loyal to the destination and who will probably repeat the experience in the future, if they have options which interest them.

• •

  High potential tourists: this is a group which stays more than five nights in the city but has scant destination loyalty (TLI=33). They are 16.9% of the sample analyzed and are characterized by their youth (the under 26 predominate), though it is the group with the greatest percentage of people over 44 years old (25%). It is also the segment with the greatest proportion of university studies – 66.3% – and they are tourists who come mainly from countries belonging to the European Union (43.8%). Their behavior when traveling indicates a greater frequency than the rest of the groups analyzed as 39.7% travel more than three times a year. The percentage visiting Seville for the first time is also significant (60.2%), as is their tendency to organize their own trip without intermediaries – 81% do so in this way. The TLI data indicate a poor valuation of the destination image in comparison with the other groups of tourists and they are also the ones who have a more reduced service expectation.

• •

  High performance tourists: the last group of tourists is characterized by an above average TLI (93.4) and a number of overnight stays which doubles the sample's average (6). They are 16.9% of the total of tourists and their sociodemographic profile indicates that they are from Spain (48%) or the European Union (31%), although this is the group with the greatest percentage of tourists from other continents (20% of the total). They are mostly under 44 years old and the majority of them have university studies. The tourist profile shows that 37% travel more than three times a year, that they mostly use planes (64%) and that they are very independent when organizing their trip (66% do not use intermediaries). Lastly, the TLI study shows that they are tourists with the greatest expectations and perceived service quality, have a high intention of recommendation and of visiting the destination again. The indicators which reflect more reduced values for this group are the global image and satisfaction.

Table 5 synthetically gathers the average values of each of the indicators which make up the TLI index, based on the four groups identified, in such a way that the differences between them can be noted.

This work has two main managerial implications. Firstly, the indicator proposed can help the agents involved in the management of tourist destinations to quantify the loyalty of visitors very easily. The measurement of the loyalty of tourists is a fundamental step to design and implement practices of improving loyalty. To foster tourist loyalty it is necessary for the agents involved to act on each of the components of the TLI; that is, on the expectations of the tourists, the level of perceived service quality, the destination's global image, the visitor's satisfaction, and also their intention to recommend the destination and repeat their visit.

On the other hand, the segmentation of the visitors according to their level of loyalty and profitability enables the development of specific strategies of relational marketing with the customers (Best, 2007). Personalized programs can be recommended for the high performance tourists, so that stable long-term relations are established based on contributing a greater value to the tourists in order to foster their satisfaction. These are special customers with whom it is necessary to develop relations at an individual level. For the high potential tourists, strategies would have to be designed which allow constructing loyalty-based relations; they are visitors prone to be easily lost. The underutilized tourists, although they are very loyal to the destination have a low profitability. In this case, communication tools can develop a fundamental role with the aim of increasing the number of these visitors’ overnight stays. Finally, with the unprofitable tourists it is recommended to go somewhat deeper into an analysis of them with a view to knowing if, through the appropriate strategies, it is possible to transform them into loyal and profitable tourists. In any case, great marketing efforts should not be made with this segment.

From a theoretical perspective, this work is relevant as it proposes a specific index to measure the loyalty of tourists. The index proposed is of great usefulness for the agents involved in the planning and management of destinations. Thus, we have estimated loyalty in a novel manner, contemplating all those determinant factors which, according to the literature, influence the visitor's loyalty. The index proposed is characterized by its operational simplicity, as each of the indicators which intervene can be measured easily and quickly. In the very competitive context in which any tourist destination is developed these days, to attain the visitors’ loyalty confers a competitive advantage and, therefore, works which go thoroughly into this idea must be considered valuable.

Moreover, the study offers a typology of tourists based on their loyalty and number of overnight stays. The identification of this group of tourists is considered indispensable for the development of different marketing strategies which foster destination loyalty, at the same time as contributing to increasing the profitability generated by these tourists.

One of this work's limitations is the design of the very TLI proposed. In this indicator, all the components have the same relative weight. Nonetheless, it would be necessary to go deeper into its analysis with a view to identifying if all the variables have the same relative importance in the indicator, or if it is necessary to introduce coefficients which enable differentiating the importance of each of the aspects contemplated.

Likewise, it would be desirable to use other indicators apart from profitability; in particular, those directly related with the incomes generated by the tourists in the destinations.

Future works could include more characteristics of the tourists, such as their personality, their degree of familiarity with the destination, their motives for visiting the destination, and so forth, with the aim of analyzing the possible differences in the degree of loyalty of the tourists based on these variables. Furthermore, in this work loyalty is investigated in the context of an urban destination. Its appropriateness could be studied in the framework of other, different destinations.