1 Introduction

Social media websites (i.e. TripAdvisor, Yelp) support tourists to search for information and form an image about a particular destination. As a result, these websites are very popular among tourists and evolved into primary online information sources for them. The spontaneous content creation and sharing in forms of blogs, reviews and posts illustrates the active role of tourists (Camprubí et al. 2013). Subsequently, tourists have more power over how a destination image is projected and communicated (Tapscott and Williams 2006). The traditional branding strategies and image formation are enhanced by user generated content (UGC). This also implies that the actual costs for the destination occur when unsatisfied tourists share their experiences in Web 2.0 platforms, and potential visitors become discouraged by the unsatisfactory comments (Camprubí et al. 2013). Considering the fact that successful branding relies partly on a positive relationship between the tourist and the destination (Ekinci 2003), Destination Management Organizations (DMOs) need to start treating the Web 2.0 as a new opportunity to learn about tourists’ opinions about the destination (Boulin 2008). Thus, DMOs need to start incorporating their consumers into the branding process to assure a significant level of destination brand efficiency. Furthermore, DMOs need to understand the evolution of destination image in social media spaces (Blain et al. 2005; Garcia et al. 2012; Veasna et al. 2012; Çakmak and Isaac 2012; Reinhold et al. 2015). The understanding of the technological dynamics as well as the development of pro-active strategies will support DMOs to capture a strong position in the highly competitive tourism market. One effective way of doing so is through monitoring and analyzing the different kinds of destination images to achieve a deeper understanding of tourists’ experiences (Jalilvand et al. 2012; Çakmak and Isaac 2012). According to Munar (2010) the transformation of massive amounts of UGC into strategic knowledge is crucial for DMOs to develop and/or maintain their competitive advantages. Subsequently, the capacity of firms to extract and integrate valuable knowledge from consumers exposed in social media spaces can enhance their marketing strategies (Munar 2012).

Furthermore, DMOs have to be aware of is the refocus of functional attributes to emotional attributes in branding campaigns. The functional attributes of tourist destinations alone no longer help destinations to attract travellers, mainly because of the high product similarity and growing substitutability of destinations (Lee and Allen 1999; Pike and Ryan 2004; Martin and Del Bosque 2008; Usakli and Baloglu 2010). Nowadays, positioning a destination around the feelings it generates, and the ability to offer visitors unique experiences, relationships, meanings and self-expressions is a strong competitive advantage (Papadimitriou et al. 2013; Hosany and Prayag 2013). Hence it is a necessity for DMOs to understand tourists’ emotional links to a destination and to develop a distinct relationship with their visitors (Hultman et al. 2015; Veasna et al. 2012). Consumers can positively evaluate the brand and become emotionally attach to the brand (Aaker 2009). Aaker (1997) refers to the concept of brand personality, which can be explained as human characteristics associated with a brand. This also implies that a brand personality enables the creation of symbolic effects for the consumer: the effective match of brand personality and consumers’ self-congruity creates a holiday status symbol, and, an expression of a lifestyle (Aaker 1996). Usakli and Baloglu (2010) posit that the match will lead to favourable destination attitudes (i.e., positive word-of-mouth, intentions to return and/or to recommend). Subsequently, DMOs need to understand which connotations consumers positively evaluate, and how they attach themselves with the destination (Keller 2003a, b, 2009; Ferns and Walls 2012). In other words, DMOs need to strive to develop a distinctive destination personality that meets travellers’ actual and symbolic needs.

Previous research by Dickinger and Költringer (2011) investigating the perceived image of Vienna of tourists and non-tourists, revealed different dimensions and connections tourists make with the city of Vienna. However, as Költringer (2012) argues, there is still a lack of understanding in which way tourists emotionally connect themselves with Vienna, subsequently referring to brand personality. This requires an integrated approach understanding tourist emotional experience (Garcia et al. 2012; Blain et al. 2005). In addition, recent work of Tussyadiah and Zach (2013) demonstrates how companies can benefit from strategically integrating knowledge from social media spaces. The large amount of available user-generated knowledge forces companies to go beyond listening and observing their consumers. Munar (2012) posits that social media has an enormous potential to help DMOs to understand their visitors’ image formation.

Hence, this study aims to integrate Aaker’s brand personality theory (1997) to explore the usage of UGC to effectively understand how tourists experience products and services. Furthermore, the study aims to provide marketers insights into consumers’ emotional links with specific service settings. Thus, besides Aakers’ five dimensions, consumers affective feelings are analyzed. Therefore, this paper considers the basic emotions tourists experience rather than purely the destination brand personality concept, as analyzed by Dickinger and Lalicic (2015). By integrating these two concepts a better understanding of emotional-based experiences can be achieved. Furthermore, research is limited in comparing offline and online data generated methods for concepts such as brand personality (Pitt et al. 2007). In addition, research applying the concept of brand personality and emotions dominantly surveyed consumers. Therefore, the study also aims to demonstrate the effectiveness of strategic knowledge revealed in UGC compared to retrieving it from the conventional survey method. Thus, overall this study aims to demonstrate DMOs’ opportunities to effectively shift their approaches of branding as well product development strategies to using user-driven reviews. For DMOs’ with emotional-based strategies, this study can provide insights into where there is a need to improve and/or enhance tourists’ interaction with the particular service setting. The study insights can subsequently support DMOs to structurally analyze UGC and work towards innovations in their organizational processes assisting a profitable enhancing growth.

2 Literature review

2.1 Destination branding

A destination can be multi-functional and co-consumed by multiple consumers (Hankinson 2010). The selection by the consumers can be across different neighbouring areas and the tourist experience may be not promoted originally like that by the DMO (Ashworth and Kavaratzis 2010). Hence, the DMO has little control over the tourist experience. However, in order to overcome this, the topic of branding has been introduced. In fact, the introduction of branding has been discussed intensively among scholars and managers. According to Aaker (1991) the purpose of branding is to differentiate its product from those of the competitors. Generally branding is a process that attempts to influence how consumers interpret and develop their own sense of what a brand means for them. As Ashworth and Kavaratzis (2010) argue, destination branding attempts to transfer those meanings to the operational environment of place management and it centres on the conceptualization of a specific destination as a brand. This means that destination branding is meant to develop a memorable bond or an emotional link between the target marketer while respecting the broader values and goals of the community that is keeping the sense of the place (Kozak and Tasci 2006). Ritchie and Ritchie (1998) have defined a destination brand as: ‘a name, symbol, logo, word mark that both identifies and differentiates the destination, it conveys the promise of a memorable experience that is uniquely associated with the destination, it also serves to consolidate and reinforce the recollection of pleasurable memories of the destination experience.’ (Ritchie and Ritchie 1998, p. 103).

Therefore, destination branding is also perceived as a hands-on marketing tool for DMOs to coordinate the different stakeholders in one theme and support the values that destinations have to offer (Ritchie and Ritchie 1998). Hankinson (2010) state that destination branding supports DMOs to communicate the various service organisations together under a shared set of associations. Subsequently, this will enable them to manage the intangible part of the tourism experience in more efficient ways (Pike 2008; Tasci and Kozak 2006; Gnoth 2002). Taglines, slogans and logos can help DMOs to communicate the formal elements of the brand (Munar 2010). Munar (2012) state that the brand can, thus, on the one hand help the DMO to deal with their lack of ownership of the destination elements and, on the other hand, enable tourists to make the associations between the different attractions, services and agglomeration services. However, destination branding is more than creating a catchy advertisement, slogan or logo (Ekinci et al. 2007).

A strong destination is recognized instantly and establishes deeper connections with travellers’ values and self-concept (Blain et al. 2005; Ekinci et al. 2007; Pike 2008; Qu et al. 2011). Furthermore, destination branding is a way to communicate a destination unique identity by differentiating a destination from its competitors (Qu et al. 2011; Ferns and Walls 2012). Hence, the selection and association of attributes that represent the main values of the destination brand makes the branding process a rather deliberated practice (Knox and Bickerton 2003). The combinations of products and services need to be a unique mixture of functional attributes and symbolic values supporting the positioning of a destination (Hankinson 2004, 2010). Hankinson (2010) perceives a destination brand as (1) a combination of perceptual entity, (2) a tool for relationships, (3) a way to communicate, and (4) a value enhancer. This implies that the multidimensionality of the destination brand construct consists of functional, emotional, relational and strategic elements (Ashworth and Kavaratzis 2010). Hankinson (2010) states that the development and management of the destination brand is perceived as a process depending on the effectiveness of the DMO’s leadership. Therefore, if destination branding is done in an effective manner, it will (1) give tourists an assurance of quality of the tourist experience, (2) reduces tourist search costs and, (3) will help the DMO to establish unique selling points (Blain et al. 2005; Ferns and Walls 2012). Thus, destination branding is vital in the current destination management practice, where there is almost infinite tourist opportunities and travel locations. The development of destination brands has become a strategic tool for many tourist destinations in countries, regions and cities (Garcia et al. 2012; Hanna and Rowley 2008; Caldwell and Freire 2004).

2.2 Destination brand personality

As discussed in the previous section, the process of branding starts with carefully choosing one or more brand elements to serve as trade-markable devices (e.g., logos) (Murphy et al. 2007). The trade-markable devices need to distinctly identify the destination, and begin the formation of strong and consistent brand associations reflecting the attribute, affective, and attitude components of an image (Murphy et al. 2006). Attributes are defined as perceptual tangible and intangible features characterising the destination. The affective components are representing tourists’ personal values and meanings, deriving from the attributes. Attitudes are the overall evaluations based upon attributes and affective feelings, acting as a basis for actions and future behavior. Hence, the topic of destination personality has been used in different studies to explicitly illustrate tourists’ attachment to a destination (Morgan and Pritchard 2004; Baloglu and Brinberg 1997; Hankinson 2004). Murphy et al. (2009) argue that brand personality is commonly used in organization studies, defining personality as enduring traits that differentiate individuals. Aaker (1997) defined brand personality as the set of human characteristics associated with a brand. She further developed the Brand Personality Scale (BPS) represented by competence, excitement, ruggedness, sincerity and sophistication.

Brand personality concept is rather an anthropomorphic metaphor, where human qualities are attributed to non-human objects, such as brands. Consumers have the tendency to view brands having human characteristics using words such as ‘cool’ and ‘young’ to describe brands such as Coca Cola and Nike (Usakli and Baloglu 2010). A well-established brand personality influences consumers’ preferences and patronage (Sirgy 1982; Sirgy and Su 2000; Selby 2004). As Thomson et al. (2005) posit, an individual who is satisfied with a brand might have an emotional attachment to it. Moreover, consumers tend to select brands that are congruent with their needs but also with their personality characteristics (Geuens et al. 2009). Thus, the self-image of consumers and brand personality is an important implication for related attitudes and future behaviors (Aaker 1996). Therefore, Aaker (1996) postulates that a brand personality has the possibility to create symbolic effects to consumers.

Also in a tourism setting there is a match found between brand personality and consumers’ self-concept, having a positive effect on tourists’ behavioral intentions (Murphy et al. 2009). Ekinci and Hosany (2006) define destination personality as: ‘The set of personality traits associated with a destination’ (p. 127). As Ekinci et al. (2007) argue, destination personality traits can be associated with a destination in a direct way, through people in a community, citizens in a city, hotel employees, restaurants and tourism attractions, or through tourist imagery, defined by a set of human characteristics associated with the typical visitor of a destination. Hence, a destination can also pose a personality that consumers use as an avenue for self-expression or to experience the emotional benefits that differentiates the destination from competitors. The higher the match, the more likely tourists will have a favourable attitude towards the destination, which spill-over in word-of-mouth and intention to re-visit (Usakli and Baloglu 2010). Destination brand personality has, therefore, positive indirect effects on intention to return and intention to recommend through the self-congruity concept (Usakli and Baloglu 2010). In an indirect manner, personality traits can be attributed to a destination through marketing programs, such as cooperative advertising, value pricing, celebrities of the country, and media construction of a destination (Ekinci et al. 2007). Morgan and Pritchard (2004) claim that building a powerful destination brand is about developing a rich, appropriate brand personality. Branded destinations are able to establish an instant emotional link with the customers, which can lead to greater loyalty (Palmatier et al. 2006; Hsu and Cai 2009).

Different studies have investigated the concept of destination brand personality by the use of Aaker’s Brand Personality Scale (BPS) (see Table 1 for an overview). All the studies show how the brand personality concept is related to specific attitudes and feelings towards a destination. Furthermore, the existing studies have found three, four and/or five dimensions representing the tourism destination brand personality (TDBP). For example, Hosany et al. (2006) illustrate that tourists ascribe personality characteristics to destinations based upon three salient dimensions: sincerity, excitement and conviviality. D’Astous and Boujbel (2007) found six dimensions (agreeableness, wickedness, snobbism, assiduousness, conformity and obtrusiveness) through unstructured interviews. Ekinci et al. (2007) found three dimensions (sincerity, excitement and conviviality). Murphy et al. (2007) study provides four main dimensions (sophistication and competence, sincerity, excitement and ruggedness). Pitt et al. (2007) used a corresponding analysis for website brand communication. They, thus, focused on how branding strategies include brand personality dimensions. The study illustrates that either DMOs have a strong brand positioning integrating the brand personality dimensions or DMOs hardly integrate of Aaker’s dimensions. Sahin and Baloglu (2009) analyzed travel brochures and internet sites and found five dimensions representative: competence and modernity, originality and vibrancy, sincerity cool and trendy, and conviviality. De Moya and Jain (2013) performed a correspondence analysis for travel brochures and found four dimensions of popularity, sincerity, excitement and sophistication. Papadimitriou, Apostolopoulou and Kaplanidou (2013) used 16 items from Hosany et al. (2006) to test urban tourism and the brand personality concept. They found two main dimensions of sincerity and excitement, postulating that destination personality is an antecedent of destination image.

Table 1 Overview of studies in tourism and BPD
Full size table

The previous studies analyzing Aaker’s scale in a tourism setting demonstrate the applicability of the scale. However, it also demonstrates that it depends on the content, as a result that there is no agreement on the ‘brand personality destination-scale’. Interestingly, the various studies analyzing Aakers’ theory from a consumer perspective were able to demonstrate how brand personality closely relates to concepts like destination satisfaction, destination image and self-congruity. Furthermore, feelings of destination personality demonstrates to positively impact tourists’ experienced value (Seljeseth and Korneliussen 2013). Studies analyzing the concept from a marketers’ perspective and its integration into branding strategies, dominantly use text mining techniques and correspondence analysis. These studies demonstrate how DMOs have a variety of brand personality dimensions reflected in their online and printed communication. However, none of the consumer-oriented studies analyzed user-driven assessments to indicate Aakers’ brand personality dimensions. Furthermore, the studies fail to indicate to which extent brand personality dimensions are triggered by specific services settings as well as to which extent the dimensions occur together. Given the rise of UGC, Aakers’ theory needs to be tested in this context too.

2.3 Strategies to deal with UGC

The intense development of social media spaces used by tourists also raises issues for practitioners regarding how to deal with it. Firstly, the content created can indirectly serve as brand management tool for many brands, and subsequently also for destinations (Seraj 2012). Secondly, consumers’ power is hereby increased and often marketers do not know how to react to this social phenomenon (Labrecque et al. 2013). Often marketers ignore the rise of social media because they do not understand what it is, the various forms it can take, and how to engage with it and learn from it (Pitt and Berthon 2011; Berthon et al. 2012).

According to Munar (2010) DMOs have various options to profit from these developments, she indicated three main strategies: mimetic, advertising and strategic. The mimetic strategies imply that DMOs can copy the style and e-culture of social network sites to create their own web site. This type is a rather conservative strategy, which is characterized by the organization keeping the main locus of control of web content on the organization (Munar 2010; Marchiori et al. 2012). The mimetic strategy is a rather easy and inexpensive way to participate in Web 2.0. It allows DMOs to keep control of UGC, DMOs can remove unwanted and/or inappropriate content (Munar 2010). Morgan et al. (2011) state that DMOs can also re-direct ads and follow a rather static approach of online content management. Munar (2010) refers it as advertising strategies, illustrating how these strategies support DMOs to benefit from the pool of information provided by tourists. Munar’s (2012) study demonstrates how the DMOs in her study dominantly integrate these approaches. Nevertheless, the lack of users’ participation and dynamism makes the advertising approach problematic and conservative (Munar 2012). The lack of cultural integration between online community and traditional corporate portal makes the mimetic strategy problematic (Munar 2012). Hence, the third strategy indicated by Munar (2010), as the strategic approach. This strategy is based upon monitoring and trend analysis and can act as a valuable tool in forecasting destinations (Munar 2010; Marchiori et al. 2012; Költringer and Dickinger 2015). DMOs can transform a large amount of UGC into strategic knowledge by examining, selecting, classifying, monitoring and evaluating the content (Marchiori et al. 2012; Költringer and Dickinger 2015). This can support a DMOs’ understanding of tourists’ image formation process (Munar 2012; Marchiori et al. 2012; Morgan et al. 2011; Çakmak and Isaac 2012). However, this demands the enhancement of skills and competences of those responsible for destination branding. In fact, Munar’s (2012) study demonstrates how DMOs rarely integrate the strategic approach, and if they do, it is rarely converted into specific initiatives. Lately, DMOs started to use a new a strategy called ‘immersion strategy’. In this case, DMOs can take the initiative to develop a social network or community based on users’ contribution (Munar 2012). As a result, synergies between the corporate and social media platforms can be enhanced (Munar 2012).

The aforementioned strategies are various creative pro-active strategies marketers can develop to manage this new social phenomenon and integrate into DMOs’ marketing strategies. Therefore, research as well as practitioners need to start analyzing UGC from a different perspective and use innovative tools to reveal hidden knowledge useful for innovating products and services. DMOs have to start working towards a systematic approach to deal with UGC. Hence, this study will provide the first insights into the use of a structural approach, to deal with UGC as strategic input for developing innovative brands, products and services compared to a survey-based method.

3 Method

3.1 Data collection and sample selection

The analysis of this study is based on two sources applied to one destination; Vienna, Austria. In order to understand how tourists would like to connect themselves with Vienna and how they would describe Vienna in terms of affective feelings, a questionnaire was created to gauge information regarding visitors’ image and connections to the city of Vienna (Költringer 2012). The questionnaire was distributed among Vienna Tourist Boards’ social media platforms targeting international visitors, subsequently yielded 599 responses.

Then, online consumer reviews were collected and compared with the outcomes from the survey. As Dickinger et al. (2011) state, users report about their experiences in a destination may differ when provided in an anonymous social media platform as compared to an online survey. Pitt et al. (2007) furthermore posit that the dominant use of consumer surveys to solicit respondents’ perceptions of a brand, calls for comparisons with other methods. User-driven reviews were collected from TripAdvisor. TripAdvisor is a popular third-party review website for tourists to reviews as well as to find relevant information for a holiday (Kaufer 2014). TripAdvisor allows consumers to complete reviews on several elements of a holiday; restaurants, accommodation and sights. Given the aim of this study to focus on the most dominant elements of the tourist experience, reviews were collected from these three service settings. Furthermore, in order to aim for a candid reflection, the same numbers of reviews in the range of negative, average and positive were systematically collected by various research assistants in November 2012. Thus, in line with the TripAdvisor scale, negative reviews had ‘1’ score, average reviews ‘3’ score and positive reviews ‘5’ score. Only reviews written in English were collected. A final set of 1104 TripAdvisor reviews was used for the analysis of this study, which average around 360 reviews per service setting.

3.2 Operationalization: dictionary design

The reviews as well as the open ended questions of the survey were subject to a text mining approach; the dictionary-approach. Various authors demonstrate how data mining methods are suitable approaches to automatically extract and analyze free-text customer feedback from online reviews of travel platforms (Schmunk et al. 2014; Stepchenkova et al. 2009). This study uses the computer-aided content analysis program ‘WordStat’ to guide the dictionary-approach. WordStat compares a list of words selected by the researcher (dictionary) against the text loaded into the software and returns the frequencies with which these words occur in the text (Pollach 2011). The dictionary design is defined by theory-driven variables and data-driven keywords. In order to build the dictionary, the first basic words were collected. Then, the algorithm grew by searching for synonyms and antonyms until no new words are found. Lastly, a manual inspection was done by two researchers to finalize the list. Two dictionaries were developed in this study. First, Aaker’s brand personality scale (BPS) was used to operationalize the phenomenon of brand personality and used for the content analysis. The BPS consists of five dimensions with a list of linked main sub dimensions developed by Aaker (1997), see Table 2. In total 555 words were collected, with an average of 20 % represented by every dimension of the BPS.

Table 2 Brand personality and related keywords (based upon Aaker 1997)
Full size table

In addition to the brand personality scale, the study analyzes the emotions used in reviews. Given the lack of agreement on basic emotions in general research as well as in tourism studies, this study decided to use the key identified emotions as guidance (Lazarus 1991). Many authors agree on six emotions which are joy, surprise, fear, anger, sadness and disgust (Lazarus 1991; James 1884; Watson 1913). Laros and Steenkamp (2005) argue that basic emotions allow the understanding of consumer’s feeling effectively. According to Han and Jeong (2013) researchers who examine emotional aspects of consumer behavior can take a categorical dimension approach. The categorical dimension uses several independent mono-polar categories of emotional responses. Laros and Steenkamp (2005) additionally proposes to introduce a hierarchy of consumer emotions, since emotions can be considered at different levels of abstraction (most general, basic emotion level, and subordinate level). Hence, this study applied a categorical approach based upon six basic categories of emotions. Table 3 provides an overview of the main dimensions and linked key-words based upon Lazarus categorization (1991). In total 776 words are collected that equally spread across the six categories of emotions.

Table 3 Basic emotions and related keywords (based upon Lazarus 1991)
Full size table

Lastly, both dictionaries were pre-processed by stemming algorithms. Wordtstat assist analysis with issues related to hyphens and punctuation marks and indicating possible stems of words (Lui 2009). In order to deal with grammar mistakes, abbreviations and sentiment, the corpus-approach was used to overcome the shortcoming of the dictionary-approach. By the development of linguistic rules, more adjective sentiment words and their orientations from the corpus were able to be identified (Lui 2009). For example, the rule about conjunction ‘and’ implying that conjoined adjectives usually have the same orientation (i.e. this city is beautiful and spacious). Similar rules were developed to deal with sentiment changes such as ‘but’ and ‘however’. The corpus-approach helped to find orientation and sentiment consistency in the text. Furthermore, the program allows analyzing and carrying out text in several formats, reduces words in canonical form, univariate frequency analysis and bivariate comparison between a textual field and any kind of variable. Thus, in this study, the three service settings were categorized into three categories, which subsequently allowed the comparisons of frequency of the dictionary dimensions. Lastly, WordStat assisted the study in performing correspondence analysis and Chi-test analysis among the different categories and service settings. Along with the correspondence analysis, dendrograms (tree-structured graph) were visualized to indicate the hierarchical clusters of the brand personality and emotion dimensions. According to Pitt et al. (2007) the dendrogram created through correspondence analysis is useful in uncovering structural relationships between variables. Thus, for this study, the dendrogram were especially useful for illustrative purposes to indicate the co-occurrence of the different dimensions.

4 Results

4.1 Analysis brand personality dimensions in social media

First, the three service settings are combined and the frequency analysis has been performed based upon the five dimensions of Aaker’s brand personality scale. Overall, sincerity is the most mentioned dimension, followed up by sophistication and excitement. The dimensions of competence and ruggedness are the dimensions the least mentioned. Furthermore, the dendrogram support the understanding of how the dimensions occur together in the reviews. Figure 1 illustrates how excitement and sincerity occur together, whereas ruggedness is often mentioned separately from other dimensions.

Fig. 1
figure 1

Brand personality dimensions in social media

Full size image

Then, the different BPD are analyzed among different service settings, Fig. 2 provides an overview, how the different services are represented and as compared to the overall reviews representation.

Fig. 2
figure 2

Brand personality dimensions among the three service settings-social media

Full size image

For the reviews related to restaurants, the most dominant dimension is sincerity and is hereby the most expressed and mentioned as important element of the experience in Viennese restaurants. Furthermore, sophistication, excitement and competence are mentioned on an average level. Ruggedness is mentioned the least of all dimensions. For sights, the dimensions of sincerity, sophistication and excitement are equally presented in reviews, whereas competence and ruggedness are close with percentage and contain the lowest references. Lastly, in reviews of accommodations, also the dimension of sincerity occurs the most frequently. Sophistication, excitement and competence are closely together in terms of percentage mentioned in the reviews. Ruggedness is with 6.9 % the dimension that is mentioned the least. Specific examples are shown in Table 4.

Table 4 Examples per service category and brand personality dimensions
Full size table

4.2 Analysis of emotions in social media

The representation of emotions in all reviews shows that surprise is the most prominent emotion. Joy is the second prominent emotion reflected in the reviews with 30.2 % Smaller percentages are divided among sadness, anger, disgust, and fear. The dendrogram (Fig. 3) illustrates how anger and joy are often mentioned in the same document; similar pattern can be found between surprise and joy. Fear, in contrast, does not show to be related to any other emotion.

Fig. 3
figure 3

Emotions in social media

Full size image

When analyzing the separate service settings according to their emotions differences are detected (see Fig. 4). For restaurants, surprise is the most dominant emotion, followed-up by joy. The emotions of sadness, anger and fear are marginally represented, whereas fear is hardly mentioned. For sights similar patterns of emotions can be indicated, fear is presented by 1.3 %. Lastly, for accommodation, emotions similar patterns can be found. Table 5 provides concrete examples for TripAdvisor reviews.

Fig. 4
figure 4

Emotions among the three service settings-social media

Full size image
Table 5 Examples per service category and emotions
Full size table

4.3 Conventional study

4.3.1 Comparison: brand personality

As discussed in the section about image, tourists’ perceptions about Vienna and information people gather prior to their trip contribute to the formation of an overall evaluation of the city. However, when tourists actually visit the city this image can change either positively or negatively through some elements of the experience travelers make in the destination. Therefore, this section looks closer into the differences between tourists describing their experiences when asked directly (as in survey) and when using social media. Table 6 provides an overview of the results based on the reviews reflecting the brand personality dimensions represented in social media compared to the results from the open-ended survey questions. The first dimension, sincerity has been mentioned more often in the questionnaire than online (51.1 and 39.9 % respectively). Ruggedness does nearly not show up in the questionnaire. It seems that a different language is used in a formal setting of data collection like a survey than when travelers express themselves freely in a review. All other dimensions, such as sophistication, excitement and competence are significantly different in the two settings.

Table 6 Comparisons of brand personality dimensions in reviews and questionnaires
Full size table

Table 7 displays the presence of the respective brand personality dimensions for accommodations, sights and restaurants as identified on TripAdvisor. The results are significantly different for all dimensions apart for ruggedness. Ruggedness is present for all three service products which imply that some travelers use critical or even negative language when talking about those three sectors. Sincerity is most present for hotels with 47.4 % whereas sights only receive 27.3 % and restaurants 40.7 %. Sophistication and excitement are most present for sights with 26.8 and 25.6 % respectively. For these two dimensions restaurants are following on the second place and hotels on the third place. Finally, competence is the highest for hotels, followed by restaurants and sights. The presence of the various brand personality dimensions provide an understanding of which service setting support the presence of a specific personality dimension for the overall destination brand.

Table 7 Comparison per service setting of brand personality dimensions in reviews and questionnaire
Full size table

4.3.2 Comparison: emotions

In this section the emotions expressed in social media are compared to the overall perceptions of Vienna as voiced in a conventional survey. Table 8 provides an overview of the differences. In fact surprise (53.1 %) is lower than in the questionnaires (77.1 %). Joy is more prominent in social media with 30.8 % compared to 21.4 % in the survey. Though the emotion of sadness was with 1 % indicated in the questionnaire and is experienced with 6 % according to social media. Hence, it seems that negative emotions are much more often present in social media than in conventional studies. This should alert practitioners as surveys may not provide all details and aspects of a service evaluation. Accordingly social media can be considered a valuable additional source to help improve service provision.

Table 8 Comparisons of emotions in reviews and questionnaire
Full size table

Table 9 provides an overview of the emotions per service category to provide deeper insights into possible differences. The emotion surprise is experienced for 49.9 %. Looking into the different service settings, surprise are mentioned the most at reviews about restaurants and sights but less in accommodations. Reviews about accommodations are mentioning joy the most often, whereas in restaurants and sights reviews, joy is present in similar levels. The category of hotels does show 4.4 % of sadness, 7 % in sights as well as in restaurant reviews. Anger is most often motioned for accommodations with 6.9 %, followed sights (4.8 %) and restaurants (4.7 %). Accommodations are also in the lead when it comes to ratings of disgust (6 %) followed by sights (5 %) and restaurants (4.7 %). Fear is also only significantly mentioned slightly more in reviews of accommodations (1.3 %).

Table 9 Comparison per service setting of emotions in reviews and questionnaire
Full size table

5 Discussion

The importance of brand personality being an integral part of a DMO’s branding strategy has been indicated by research linked with destination image studies and destination branding efficiency. Moreover, the importance of consumer attachment to a brand and the emotional links to a product and/or service has been proven to successfully differentiate a brand among competitors. Also, in the field of tourism this trend has been recognized and DMOs have started to respond to that. The various studies in the field of tourism applying Aaker’s brand personality scale demonstrate the effect on tourists’ positive destination image, satisfaction and attachment (i.e., Hultman et al. 2015; Chen and Phou 2013). Pitt et al. (2007) illustrate how DMOs have started to integrate brand personality dimensions in their marketing communication strategies. However, the dominant usage of survey-based methods explaining this concept limits the understanding of consumers’ emotional attachment to a destination.

Research has acknowledged that there is a need to integrate UGC as strategic knowledge to understand consumers’ experience as well as to innovate marketers’ business models (Tussyadiah and Zach 2013; Berthon et al. 2012; Reinhold et al. 2015). This study aimed to fill the gap in brand personality research in the field of tourism using user-driven reviews. Then, as a response to Pitt et al.’s (2007) suggestion, offline and online data-generating methods were compared to assist marketers in exploring the brand personality topic from a consumer perspective. In fact, this study demonstrates how survey-based data and review-based data are distinctly different when analyzing emotion-based experiences. In particular, the dimensions of excitement, sophistication and competence are significantly more represented in social media than in the conventional survey. When looking at results for emotions, questionnaires do not exhibit too much negative content. Feelings of anger, disgust and sadness are hardly expressed while they seem to be present in social media. These outcomes are in line with previous studies aiming to compare offline and online methods. For example, Dickinger et al. (2011) show how offline and online destination images are similar. However, given the richness of UGC, various additional topics were revealed (Dickinger et al. 2011). In this case, negative emotions were significantly more represented in reviews. Hence, the social media environment proves to be a more appropriate environment to indicate sentiment-based topics rather than surveying consumers.

The second theoretical implication that can be derived from this study lies in the exploration of Aaker’s theory and emotion-based theory in the various service settings. Previous studies have analyzed brand personality on a destination level (Murphy et al. 2007; Papadimitriou et al. 2013; Seljeseth and Korneliussen 2013). This study demonstrates the difference between the service settings based upon emotions experienced and thus indirectly the applicability of Aakers’ scale in these settings too. Interestingly, sincerity and sophistication are the two dimensions that are consistently presented among all service settings. Hosany et al. (2014) analyzed destination emotions that are purely focused on positive emotions (joy, love and positive surprise). This study also focused on negative emotions (disgust, fear and anger). In this case, this study was able to demonstrate the different emotions that tourists experience. In particular, negative emotions were significantly present in reviews related to accommodations and restaurants. Thus, when attempting to improve consumers’ experiences with the brand and overall re-visitation, analyzes of negative emotions have to be included. Chen and Phou (2013) illustrate how brand personality facilitates a higher level of destination satisfaction. Hence, also for service providers, consumers’ satisfaction can be influenced if marketers are aware of the brand personality as well as the emotions (negative and positive) associated with it.

Thus, with regard to managerial implications, this study illustrates how UGC can help marketers to adjust their branding strategies according to user-driven evaluations. Sahin and Baloglu (2009) illustrate that DMOs have started to integrate brand personality dimensions into their communication. This study shows how consumers emotionally perceive the brand, which might not be in line with the destination identity. Furthermore, specific service providers can benefit from this study. Accommodation, tourist attractions and restaurant managers can start to integrate Aaker’s theory into their branding strategies. For example, accommodation reviews have a high number of ruggedness–related feelings. Marketers have to decide which dimensions need to be experienced, and which approach they need to take in order for consumers to feel these dimensions. For example, negative experiences represented in reviews highlight the need for improvement (i.e., avoiding situations where consumers start to feel sad or disgusted). Furthermore, marketers can also aim to have a more balanced representation of different dimensions. This study shows how, for example, accommodation reviews illustrate higher levels of competence compared to the other two service settings. However, managers of restaurants might aim for this dimension too. Thus, understanding which elements in the service design trigger feelings of competence can be translated into an innovative restaurant experience. For example, managers can train their employees to demonstrate levels of devotion, self-confidence and trustworthiness. Or in the case of creating higher levels of excitement, managers should provide experiences that are irreplaceable. They can train employees and/or design a business model that enables the firm to provide personalized experiences. This shows that managers can have the power to trigger brand personality-feelings in consumers, subsequently steering a positive image and consumer attachment. Furthermore, the emotion-based analyzes help managers to understand consumers’ levels of satisfaction more than brand personality dimensions do. Given that, for example, consumers express negative emotions in accommodation reviews, managers can implement specific service recovery actions to steer these affective feelings. Naturally, the two theories are closely linked and, thus, managers can steer them simultaneously and aim for positive and unique experiences.

This leads to the second implication of this study. The use of innovative methods supports managers to retrieve strategic knowledge from and about consumers. Tourism research emphatically shows how UGC with specialized data mining techniques helps marketers to drive businesses model innovation. The dictionary-approach shows to be very supportive when searching for specific dimensions. On top of that, this study shows how another topic, affective-based experiences, can be used when analyzing the large amount of UGC. This helps marketers to successfully shift their branding and marketing communication strategies to an emotional-based focus. Given that this stems from consumer-driven evaluations, the success of innovative branding strategies as well as tourists’ interaction with the destination and/or specific service provider can hereby be guaranteed.

Future studies need to analyze the long term effect of branding strategies steered by consumer evaluations. Given the recent development of this topic, longitudinal studies could help to analyze the success of this user-driven approach. Furthermore, a comparison between various destinations could yield insights into how branding strategies steered by affective feelings generate competitive advantages. Additionally, future studies should analyze specific elements of the experiences that are closely linked to brand personality dimensions and emotions. A qualitative approach would allow for a richer understanding of this topic. Another interesting avenue for future studies would be to compare DMOs’ marketing communication with user-driven evaluation content. This study did not consider possible brands (hotel chains) interfering with the analysis. Thus, future studies could also analyze the accommodation sector in particular. Again, a comparison would help to generate insights into possible competitive advantages stemming from user-driven branding campaigns and product/service development.