Abstract
The impacts of globalisation and cultural convergence theory have been discussed in various research areas. Scholars agree that global consumer behaviour becomes increasingly homogenous. However, in tourism, there is still evidence that visitor markets are distinguished and segmented according to their nationality. This study argues that national borders have lost in significance to separate markets. By means of a correspondence analysis, this study provides evidence that European travel behaviour in Australia shows similar activity patterns. The findings highlight that country-based visitor segmentation has lost momentum.
Keywords
Introduction
In the field of international marketing, the distinction into national-border segmentation has frequently been applied with the hypothesis of cultural distinctions. However, as the effects of globalisation increased, researchers’ question whether national cultural distinctions still form a valid base to distinguish markets and whether the use of geographical market segmentation along national borders is still expedient. This article debates whether European tourist behaviour follows converging or diverging behaviour patterns. This argument is conceptualised along geographical market segmentation bases.
The purpose of market segmentation is to distinguish broad consumer groups into segments with similar preferences (Dibb and Simkin, 2010; Moscardo et al., 2008; Nachum, 1994). The aim of segmenting customers is to predict subsets of groups that will respond similarly to marketing activities (Foedermayr and Diamantopoulos, 2008; Tkaczynski et al., 2009). In terms of managerial decision-making, segmentation results lead to a decision either to standardize across countries or adapt products and services for each country (Hassan et al., 2003). To create effective results and understand tourist behaviour patterns, bases need to crystallise tourists with ‘similar tourism needs and behaviours, similar socio-demographic profiles, who are profitable, who could easily be reached with marketing communication messages…’ (Dolnicar and Grün, 2008: 63).
In the past, national cultures have often been using segmentation base (Dolnicar and Kemp, 2009; Tkaczynski et al., 2009), assuming that the distinction of national cultures forms a valid segmentation base. In tourism, the visitors’ place-of-origin, for example, is either used as a single layer of segmentation or within a multiple-layered segmentation approach; it is mostly oriented along national borders, distinguishing between location, place of residence or country of origin (Crotts and Litvin, 2003; Tkaczynski et al., 2009).
However, especially outside the area of tourism research, studies bring globalisation effects into context with bases of customer segmentation. Researchers debate if national cultures still form a valid segmentation base because studies have shown that ‘over time, the number of segments, segment sizes, and structural properties of international segments may change’ (Steenkamp and Ter Hofstede, 2002: 209). In the early 1980s, Levitt’s (1983) seminal work had shown that national cultures had become increasingly homogenous. Since then, a number of scholars advocate that cultures have become more convergent (Dann, 1993; Ko et al., 2011; Ter Hofstede et al., 1999; Usunier and Lee, 2009), rejecting the effectiveness of national borders as a segmentation base in international marketing (Foedermayr and Diamantopoulos, 2008; Ko et al., 2011).
This research examines European tourists’ activities and the extent to which they differ or converge around the geographical/national origin of each tourist group. The data were analysed with the method correspondence analysis, which provide an accessible graphical representation of association between variables: nationalities of European visitors and undertaken activities in Australia. The article identifies whether those travel activities will vary significantly or instead show a significant homogenous segment for each of the seven countries. With the results, the study contributes to the debate on cultural convergence and the understanding of European travel behaviour research. The findings add to tourism marketing practice and advocate that international tourism firms rethink approaches to European tourist markets by country of origin, concentrating instead on other affinity segments, such as age or lifestyle.
Literature review
Since Hofstede’s (1980) typology, national culture has become a variable with increasing interest to be researched upon and hypothesised. Various studies have discussed whether culture influences human behaviour. Studies have investigated whether societal culture has an effect on management and particular management styles (House et al., 2002; Jackson, 2005; Triandis, 2004; Triandis and Gelfand, 1998) and how cultural background is influencing tourist behaviour (Crouch et al., 2005; Litvin et al., 2004; Pizam, 1999; Reisinger, 2009). National cultural characteristics have been used in tourism studies to determine whether ‘nationality is one among a number of factors that account for differences in tourist behaviour’ (Pizam and Sussmann, 1995: 905). Culture on a national level stems from belonging to a specific culture and having the same nationality and geographical background (Kastenholz, 2010). In terms of the distinctiveness of both the constructs culture and nationality, Bhaskaran and Sukumaran (2007: 55) found that leading scholars in culture studies (e.g. Hofstede, 2001 and Hall, 1990) ‘tend to use the term culture and nationality interchangeably, thus implying that nation states comprise populations with a shared history and experience’. Following these arguments, and for the purpose of this study, the constructs societal culture and nationality are used interchangeably.
International marketing research and practice has utilised cultural distinctions to understand consumer behaviour, and national-border segmentation bases have been used to better target specific consumer groups (Dolnicar and Kemp, 2009; Hassan and Katsanis, 1994; Helsen et al., 1993). Initially, benefits of national border’s segmentation might seem evident: there is little or no coordination of tourism marketing activities necessary between European countries (Yip, 1995). Data are comparatively easy to generate (Scott and Parfitt, 2004), and international segmentation is considered to be more difficult for firms when not distinguishing between national borders (Yip, 1995). Consequently, researchers propose to conduct more geographically based segmentation studies (Obenour et al., 2005).
However, there is doubt arising whether those distinctions still reflect idiosyncratic customer behaviour. In the early 1980s, researchers already confirmed the trend towards gradually increasing homogenous consumer behaviour across nations (Levitt, 1983; Ohmae, 1989). The trend towards converging cultures is explained by the effects of globalisation and the theory of global cultural convergence. Supporters of this theory presuppose that there is evidence that, over the last two decades, several waves of globalisation have taken place; those globalisation tendencies have lead to changing external market drivers that have resulted in increasingly homogenous international markets (Hassan and Craft, 2005). Hence, since the beginning of the 1990s, studies started to become critical towards predetermined country bases (Dann, 1993). Choosing national cultures as a segmentation base is increasingly regarded as problematic because diversity and differences within a country are often overlooked (Ter Hofstede et al., 1999).
Various studies provide evidence that societal cultures have become more convergent (Dann, 1993; Ko et al., 2011; Ter Hofstede et al., 1999; Usunier and Lee, 2009). As a result, there is a strong argument that geographical segmentation is rather outdated. Moriarty and Duncan (1991) found evidence for convergence of cultures in the field of advertising research. The effect of advertising was that some segmentation became conducted across customer groups (e.g. age groups and lifestyle groups) rather than within national cultures.
Ko et al. (2011) conducted cross-national market segmentation within the sports brand industry. Their findings suggest that consumer behaviour is oriented on demographic and lifestyle segments rather than on societal cultural differences. Their research supports the argument for the existence of the construct of the ‘global consumer’. The notion of the ‘global consumer’ (Underhill, 2008) has been introduced as ‘global consumers whose social and cultural differences are overshadowed by their similarities in terms of psychological consumer tendencies’ (Keillor et al., 2001: 1).
In the case of Europe, the formation of the European Union further accelerated the homogenisation of cultures (Green Cowles et al., 2001; Ter Hofstede et al., 1999). Overall, there is an agreement that due to converging cultures, the value of national-border segmentation is increasingly questionable and a debate whether ‘national culture’ is still a feasible, contemporary variable to distinguish markets (Foedermayr and Diamantopoulos, 2008).
Earlier studies in the retail industry confirmed that customers demonstrate tendencies of a globally homogeneous purchasing behaviour (Segal-Horn and Davison, 1992). This view has been supported over time, and studies suggest targeting retail customers cross-nationally (Ko et al., 2011). More studies in international marketing research have discussed various approaches to segment markets along more sophisticated bases of segmentation. Unobservable bases – such as values and lifestyles – are discussed by Wedel and Kamakura (2000). They describe contemporary product-specific bases, response-based segmentation and product-specific bases. Lifestyle convergence for consumer groups across countries are findings of the latest research studies (Usunier and Lee, 2009).
However, those findings are not always recognised by marketers, and studies criticise researchers for not publishing and disseminating timely approaches to marketing segmentation. ‘Many of the guidelines offered to practitioners in the market segmentation literature are grounded more on anecdotal evidence and “common sense” rather than on a solid empirical base’ (Foedermayr and Diamantopoulos, 2008: 224). Scholars also utter that there is still not enough substantial research published on how national cultures affect consumer behaviour (Hassan and Craft, 2005; Ko et al., 2011).
In the field of tourism research, there is evidence that studies still support the argument that national cultures form a distinctive base for distinguishing travel behaviour, and researchers agree that those geographical segmentations are still frequently applied (Dolnicar and Kemp, 2009; Moscardo et al., 2000; Tkaczynski et al., 2009). According to Pearce (2005: 133), most tourism research argues for diverging and heterogeneous cultures: ‘An implicit conceptual approach underlies much of the international market segmentation research in tourism. The approach may be described as divergence and is characterized by the view that markets are culture-bound, often nationally distinct and are likely to stay that way’. Reisinger (2009: 22) also supports this argument for the distinctiveness of geographical borders of Europe; it is culturally fragmented, and countries preserve their culture: ‘European countries pay particular attention to their cultural identities, origins of artists, rituals, art works, buildings, and even whole landscapes’.
Consequently, it is unclear whether this distinctiveness continues to manifest in travel activity patterns and whether this argument is transferable to international market segmentation in tourism. The research question of this article is to investigate whether European tourist behaviour follows converging or diverging behaviour patterns. The following section introduces the research approach of this article. The method of correspondence analysis is explained, and the data sourced from the International Visitor Survey (TRA, 2012) are presented. The analysis of the activity-based segmentation is then displayed, and the results are discussed.
Method
Correspondence analysis has been used in marketing research since 1980s (Greenacre, 1993; Hoffman and Franke, 1986; Nishisato and Gaul, 1988) and recently by Torres and Greenacre in 2002. However, the method has not been applied in tourism management studies until Chen’s (2000) application for cross-cultural differences. Since then, it has been adopted more often in tourism marketing (Beldona et al., 2005; Chen, 2001; Diana and Pronello, 2010; Gursoy and Chen, 2000). In the context of tourism segmentation, Kim and Jogaratnam (2003) conducted a study on activity-based segmentation of university students.
To the authors’ knowledge, correspondence analysis is used here for the first time to analyse activities of travellers to Australia. Correspondence analysis is a technique that gives an easily accessible graphical representation of association among categorical variables. Row and column categories are plotted by extracting two dimensions that capture the maximum amount of structure of a matrix of counts, using the proximity of points to describe relationships. The closer the plotted categories, the stronger their relationship at a descriptive level – but not at an inferential level. In terms of statistical inference limitations, it should be noted that correspondence analysis is an exploratory data analysis method. As such, distances between row and column points are interpreted at an ordinal rather than ratio level. See Greenacre and Hastie (1987) for further discussion of the geometric interpretation of correspondence analysis.
The most common kind of table submitted to correspondence analysis is the two-way frequency cross-tabulation as shown in Figures 3 and 4 (Greenacre, 1993, 2007).
Data
TRA (2012) publishes International Visitor Profiles of countries that have been identified as core target markets. The data are geographically segmented according to national borders. The International Visitor Survey (IVS) data that underlie the International Visitor Profiles are accessible to the public via the Tourism Research Australia (TRA) website. Consequently, the data are available for use by Australian operators, supplying them with already segmented tourist data.
According to the IVS methodology, the survey has been conducted every year since 1981 (TRA, 2011). In 2000–2004, the sample size was approximately 20,000 international visitors who departed Australia and were aged 15 years and older. In 2005, this sample size increased to 40,000 international visitors departing Australia. The sample size provides a 95% confidence interval compared with a census (TRA, 2011). Interviews were conducted in departure lounges of the major international airports (Sydney, Melbourne, Brisbane, Cairns, Gold Coast, Perth, Adelaide and Darwin). The questionnaire contains 70 questions, and interviewers use computer-assisted personal interviewing to capture information directly into their laptop computers.
TRA (2011) states a continuous improvement in the data quality. The evaluation and adjustment of the methodology used has improved the reliability of the data. Adjusting sample size, adding a broader country range and weighting adjustments of single countries were the measures undertaken to improve data reliability. In addition, the weighting is aligned to the data of the Australian Department of Immigration and Citizenship, providing a higher level of reliability. The weighting includes ‘country of residence, main purpose of journey, airport of departure, age and sex of the visitor’ (TRA, 2011, n.p.). ‘The sample is further adjusted to match total arrivals for each target market by purpose of visit and state of arrival’ (TRA, 2011, n.p.).
The focus of this study is the seven major Australian inbound markets from Europe (six countries and Scandinavians as a region) according to Tourism Research Australia International Visitor Survey (TRA, 2012). Table 1 shows arrival numbers in declining order.
Arrivals of European travellers in Australia (2000–2010).
The activities in Table 2 have been selected from the data of the 41 activities gathered within the IVS (TRA, 2012) because they represent the 20 most selected activities for all travellers. Through sorting activities for frequency of occurrence for each country, Top 10 activities of European travellers in Australia have been determined. Figure 1 displays the relative behaviour per country for the Top 10 activities in a percentage-stacked column. Already at this point, a fairly uniform distribution of similar interests is noticeable.
Most popular activities of travellers in Australia in 2010.
Top 10 activities of European travellers in Australia in 2010 (relative count).
In a second step, the middle tier of activities is explored to see whether behaviour according to national culture can be seen outside the most popular visitor behaviour. Activities from the IVS database, ranked 11 to 20 by frequency, are shown in Figure 2.
Activities ranked 11–20 of European travellers in Australia in 2010 (relative count).
Results and discussion
The application of a correspondence analysis aims to find relationships between country and activity. The correspondence analysis was conducted in SPSS with Euclidean distance. Row and column means were removed and symmetric normalisation was applied.
In Figures 3 and 4, we see the Top 10 activities and the middle tier of activities ranked 11–20, respectively, separated by enhance visibility. The countries are symbolised by black-named points and the activities by open-numbered circles. As shown by Chen (2001), the axes of the correspondence analysis plot are dimensionless and, as such, cannot be labelled beyond their ordering (first dimension, second dimension etc.).
Plotted categories for Top 10 activities.
Plotted categories for activities ranked 11–20.
For the Top 10 activities, Figure 3 shows a rather homogenous distribution across countries. The distances between countries show only minor differences in travelling activities in Australia. One country that is shown as having a slightly distinct behaviour is the United Kingdom, with some activities more specific to that country. On the other hand, the continental European countries have a fairly similar activities pattern.
This closeness of travel behaviour could be theorised with the increased convergence of European cultures (Green Cowles et al., 2001). Previous studies in international tourism already confirmed that cultural proximity is significantly influenced by cross-cultural exchange and consumption on the tourist destination image (Kastenholz, 2010). Hence, with ongoing European unification, cultural proximity amongst its members could have increased, due to more frequent inner-European travel and exchange, but in this case, it is converging travel behaviour.
Some distinctions that can be named as specifically popular with the British are: eating out in restaurant and cafes (No. 1); going out to pubs, clubs and discos (No. 5); going to the beach (No. 3) and going to markets (No. 8). National parks (No. 6) are popular with Scandinavians and, to a lesser degree, with Germans and Swiss. Botanical gardens (No. 7) and wildlife parks (No. 9) are frequented by Scandinavians and French. Sightseeing (No. 2) and bushwalking (No. 10) are not allocated to specific countries; they are equally popular with all of them. There are slight differences that are recognisable in British travel behaviour in Australia.
This might be explained by findings of a previous study that researched differences of British, European and American backpacker behaviour. The research found slight differences and concluded that social aspects of travel – such as meeting Australians – are in the forefront of British backpackers’ travel motivation, whereas Scandinavian and German backpackers would rather concentrate on experiencing the outdoors (Buchanan and Rosetto, 1997).
While Figure 3 shows that travel activities within the Top 10 activities are similar for the seven major European countries, Figure 4 shows that, on a next tier of activity popularity, there are some country differences.
The British seem to like visiting heritage buildings (No. 13) and cinemas (No. 15); they also do charter cruising (No. 11). Swiss visitors like snorkelling (No. 16), and the Dutch like going on guided tours (No. 14). Experiencing aboriginal art (No. 17) is equal in Scandinavia, the Netherlands, Germany and Switzerland; it seems to be a feature liked by the Germanic countries. Other activities such as visiting museums or art galleries (No. 12), visiting wineries (No. 18) or the outback (No. 19) or tourist trains (No. 20) are popular activities in Australia among European visitors and feature within the 20 most-liked activities, as shown here. They cannot be attributed to a specific group of visitors, however, at least with regard to country of residence; they are of general interest to all European visitors.
Overall, the findings show the seven countries to be surprisingly similar rather than revealing differences. It can be assumed that the Top 10 activities – the ones that interest the majority of travellers – are more homogeneous in appearance than the lesser chosen activities. This is supported by the results. Even within the activities ranked 11–20, we see more similarities than differences. Again, it is the United Kingdom that has a slightly different pattern compared to other European countries. Instead of being able to attribute some activities to specific European countries, such as France and Italy, they do not seem to have specific favourites within the activities ranked 11–20 (compare Figure 4). The lack of closeness of activity plots to country plots within the correspondence analysis is then to be interpreted that these activities are of equal interest to all countries within the analysis. The findings of relatively similar activities, as displayed in Figures 1 and 2, have now been supported by the correspondence analysis (compare Figures 3 and 4), showing no specific cultural differences of European visitors with regard to the activities undertaken when travelling to Australia.
Conclusion and implications
Theoretically, the discussion in this article followed the debate on converging versus diverging cultures and the subsequent impact on international tourism marketing. The article reported on a correspondence analysis of secondary activity data from seven European tourist groups. It showed that in the case of European travellers in Australia, national cultures are becoming increasingly converging in their activity patterns during holidays. It appears that the consequences of globalisation have led to a decline in importance of using separate European national cultures as a segmentation base.
Thus, it contributes to the argument that national borders are losing their relevance as a segmentation variable. Implications towards international tourism segmentation clearly question that national borders remain an effective segmentation base, in particular, in the case of European visitors’ activities. With these results, this study offers a foundation for further research in this area, for example, on the value of lifestyle-oriented segmentation bases in international tourism marketing. Additionally, studies could investigate other geographical situations, as this article focused only on the case of the seven strongest European visitors for the target market Australia. For example, the argument of divergent cultures might hold continent distinction between European and Asian countries. While this study has a scope on activities of European visitors in Australia overall, further research should be conducted on where European visitors have travelled in Australia with regard to the specific Australian states and territories, how long they have stayed in each state, territory and overall in Australia and if visitors travelled as individuals or as part of a group. How these factors impact on the activities visitors to Australia have participated in during their stay should be explored in further research.
During the process of data collection and analysis, a further methodological issue has arisen regarding presupposed activities. When identifying and segregating activities, it is crucial to validate items first instead of making assumptions on potential activities. Some of the activities, extracted from TRA database, might differ in their significance and distinctions (e.g. go to the beach and snorkelling). Therefore, attention could be drawn to models to generate valid activities.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.