International Brand Strategy: Relation
between Market Size and Brand
Adaptation
Program: MSc Business Administration: International Management
Thesis Supervisor: Dr. Vittoria Scalera
Subject: Master Thesis
Student: Florian Ritter van de Kamp – 11372702
Statement of originality
This document is written by Student Florian Ritter van de Kamp who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
International marketing strategy has become a relevant topic of international business literature, and this thesis will focus on one of the subtopics within the international marketing strategy: the international brand strategy. The decision of whether to use a standardized brand or an adapted brand can be a key decision, ultimately determining success or failure in a new market. On the one hand, a standardized approach can benefit from cost savings, deriving from economies of scale and scope, but, on the other hand, a standardized approach fails to fully take into account local consumers’ tastes and preferences.
Based on a literature review of the topic, the most important factors in the decision whether to use an adapted or standardized brand approach identified are culture and market characteristics, such as competitors and consumers, among others. We decided to test whether market size also plays a role in this decision, since this factor is mentioned in the literature on international marketing strategy but not in the literature on international brand strategy. Our specific research question is: how does the market size of the host country affect the brand strategy?
For the cultural factor, we decided to include a newer cultural difference parameter, the psychic distance approach; more specifically, we use the linguistics and religious psychic distance ratios as moderators, since these factors have proved to have a big influence on the branding strategy. After gathering data on the beer industry in Asian markets, we test our hypotheses. The testing of our hypotheses show that market size has no influence on the degree of brand adaption, contrary to what we had predicted. However, the psychic distance values certainly influence the degree of brand adaptation aside from other factors included in the research.
Key words: International Brand Strategy; psychic distance; market size; culture;
Table of Contents
1. INTRODUCTION ... 5
2. LITERATURE REVIEW ... 9
2.1. Marketing Strategy: Adaptation or Standardization ... 9
2.2. International Brand Strategy ... 10
2.3. Brand Strategy and Culture ... 11
2.3.1. Cultural Differences ... 11
2.3.2. Linguistic Differences ... 11
2.3.3. Perception Differences ... 12
2.3.4. Psychic Distance ... 12
2.4. Brand Strategy and Market Structure and Size ... 13
2.5. Brand Adaptation ... 14 2.6. Brand Standardization ... 15 2.7. Impact on Performance ... 15 2.8. Research Question ... 15 3. HYPOTHESIS DEVELOPMENT ... 17 3.1. Market Size ... 17 3.2. Culture ... 19 3.2.1. Linguistic Differences ... 19 3.2.2. Religious Differences ... 20
3.3. Overview of Hypotheses to be Tested ... 21
4. METHODOLOGY ... 22 4.1. Sample ... 22 4.2. Variables ... 23 4.2.1. Dependent Variable ... 23 4.2.2. Independent Variable ... 24 4.2.3. Moderator Variables ... 24 4.2.4. Control Variables ... 25 4.3. Empirical Strategy ... 28 5. RESULTS ... 29 5.1. Descriptive Statistics ... 29 5.2. Correlations ... 30 5.3. Regression Analysis ... 31
6. DISCUSSION AND CONCLUSIONS ... 36
6.1. Discussion ... 36
6.2. Conclusions ... 39
6.2.1. Theoretical Contributions ... 41
6.2.2. Managerial implications ... 41
6.3. Limitations and Direction for Further Research ... 42
1.
I
NTRODUCTIONIn today’s globalized world, it is a challenge for companies to develop a successful marketing strategy for every market. The crucial decision of whether to use a global, standardized, approach or different, customized strategies adapted to every single market can be the key factor in determining success or failure in a specific market. IBM is an example of a company that successfully applies a global, standardized approach. It uses the same marketing strategy in every market they operate in, and even create global advertising campaigns, thereby benefiting from economies of scale (Aaker and Joachimstaler, 1999). On the other hand, companies such as Procter & Gamble use both a global and a local strategy, motivated by the differences identified in different markets (Aaker & Joachimstaler, 1999). This can be seen in their consumer goods products, which consists of a wide variety of brands.
Several factors influence the decision of whether to use a standardized or individually tailored marketing strategy approach. Theodosiou and Leonidou (2003) reviewed all the studies on this subject conducted over the last years and summarized their findings. Despite the many studies done already, their review revealed that research on this topic is still in its early stages. More recent research on the topic still lack attention for this specific issue, especially in the emerging markets (Erdogmus, Bodur and Yilmaz, 2010). On the one hand, those in favour of a standardized approach point out that it allows economies of scale and scope to be achieved, consistent messages to be delivered, and cross-national markets to be attracted (Yip, 1992). On the other side, those who favour an individually tailored marketing strategy highlight the fact that significant differences exist between consumers, cultural and socioeconomic conditions, and market conditions in different countries, making a customized approach worthwhile (Douglas and Wind, 1987). In this context, Özsomer and Altaras (2008) add that the national market characteristics and cultural differences should be taken into consideration (Özsomer & Altaras, 2008).
Although companies generally do not use either extreme approach, adopting strategies that lie somewhere between the extremes outlined above, some argue that a global approach to brand marketing should be used but duly taking into account local insights—to be able to deliver a standardized message—while promotion should be customized (Barron and Hollingshead, 2004). Unfortunately, little empirical evidence is available on the best option in terms of
profitability (Cavusgil and Zou, 1994), and how each option affects brand performance (Jain 1989; Szymanski, Bharadwaj & Varadarajan, 1993).
Some authors argue that opting for a standardized rather than a tailored strategy may be motivated by the desire to achieve higher profitability through economies of scale and scope (Steenkamp, Batra & Alden, 2003). Moreover, using a global approach may also speed up market entry by reducing the need for time-consuming modifications (Steenkamp, Batra & Alden, 2003).
This study focuses on one specific element of the entire marketing mix that has to be addressed by companies and has not yet received enough attention—the international brand strategy (Hofer, 2015). Although this topic is not yet a major concern, it does constitute a key issue within a company’s overall marketing strategy. Strong brands help companies establish an identity and develop a solid customer base, as well as provide the basis for future brand extensions (Douglas, Craig and Nijssen, 2001; Whitelock and Fastoso, 2007). The decision to use a standardized global brand in a market or adapt the brand to a specific market defines an important element of the marketing strategy.
Among the factors that influence the brand strategy decision are the host country’s cultural characteristics (Roth, 1995)—for example, many companies change their standard brand names in foreign markets to avoid linguistic problems (Francis, Lam and Walls, 2002). Companies can also decide to develop new brands that are deemed more appealing in the new market (Usunier and Shaner, 2002). Specific market structure factors may also induce companies to adapt their brands (Alashban, Hayes, Zinkhan and Balazs, 2002), among others, the intensity of competitors and buyers, because they shape the market in which the company will operate. Together with cultural factors, these market structure elements are external factors that cannot be influenced by the company itself (Hofer, 2015). Organizational factors—the firm’s competitive strategies, the coordination of its marketing activities and specific product characteristics, and the like—also influence the brand strategy adopted. These elements can be controlled by a company (Hofer, 2015).
A relevant external factor that cannot be controlled by a company is market size. Not much research has addressed the question of how the market size of a country could influence a company’s decision of whether to use a global, standardized, brand or an adapted one. Only
limited research has been done in this area by Townsend, Yeniyurt & Talay (2009), based on a hypothesis on the likelihood of a brand being introduced in a market influenced by market size. When companies decide on whether to enter a new market, they first evaluate the market attractiveness, considering that larger markets generally have a higher potential to generate sales and revenues (Townsend et al., 2009).
In view of the above, this study investigates the influence market size has on the brand strategy decision, more specifically, how market size affects the decision to use a standardized, global, brand or a customized brand. The question we thus try to answer is: How does the market size of the host country affect the brand strategy? This question is particularly interesting because it will give managers another aspect to take into consideration when developing their international brand strategy. Considering the very limited amount of research done to date on this specific subject, this study will make a valuable contribution to the knowledge on international brand strategy. More in general, this study will contribute to the international marketing literature by providing new insights into the factors that play a role in determining the optimum international marketing strategy.
As mentioned above, the role of culture in this context has been widely investigated (Roth, 1995). Culture is a very important factor to be considered when defining a marketing strategy, because cultural norms and beliefs are powerful forces that shape people’s perceptions, dispositions, and behaviours (Steenkamp, 2001). Therefore, the moderating effect of culture when deciding on a brand strategy will be considered; more specifically, the Douglas Dow framework will be used to differentiate between the country of origin of the company and the host country.
The findings from the research already done on brand strategy only focus on a few specific market products with particular characteristics in every case, so it’s sometimes difficult to aggregate the results found for one sector to other industries. In addition, the research conducted mainly involves United States companies. This thesis will specifically use data about the beer, spirits and wine industry in Asian countries.
This thesis is structured as follows. Chapter 2 presents the relevant literature on brand strategy and the different aspects that influence brand strategy across markets, among relevant factors. Chapter 3 introduces the theoretical framework, based on three hypotheses
and certain propositions. Chapter 4 addresses the methodology used for this study—how the data were collected, and what variables and what method were used for this research study. Chapter 5 presents the results of the statistical analysis Chapter 6 discusses the results, conclusions, and limitations of the study. It also suggests some areas for further research. A list of references is provided after the main text.
2.
L
ITERATURE REVIEW2.1. Marketing Strategy: Adaptation or Standardization
The issue of standardization or adaptation in marketing strategy has been widely discussed since the publication of an article by Levitt about the globalization of markets (1983). Levitt (1983) concluded that markets across the world had grown more similar, thanks to improvements in communications, travel, and the emergence of similar demographic segments across the world. He argued that because of these factors, firms could increasingly serve the world using large-scale production methods, thereby benefiting from economies of scale. Moreover, a consistent message could be delivered and cross-national markets attracted. More recent research on the topic also suggests that the homogenization of consumer tastes and need patterns can facilitate the use of a standardized approach, its ultimate goal being to save costs thanks to uniformity (O'Donnell & Jeong, 2000).
However, some authors oppose the standardization, for example, Douglas & Wind (1987), arguing that consumers, cultural and socioeconomic conditions and market conditions still show significant differences across different markets, so a customized, adapted approach should be used for every distinct market. Given these differences across markets, it is difficult for firms to completely standardize their marketing strategy. In fact, O'Donnell & Jeong (2000) argue it is preferable for the firm to localize, in other words, tailor the marketing strategy to the specific conditions of each and every market.
These arguments present two divergent options: (i) the use of a standardized marketing strategy or (ii) the use of an adapted marketing strategy. In practice, however, companies rarely use either of the two extreme approaches when defining their marketing strategy (Quelch & Hoff 1986). Neither of these extreme approaches is conceivable for a firm, and the ultimate decision on what approach to use will ultimately always depend on their relative profitability, but there is little evidence on which approach will ensure the highest profitability and performance (Jain 1989; Szymanski et al. 1993; Cavusgil & Zou 1994). The main factors that favour an adaptation of the marketing strategy are the following: external or environmental constraints, the nature of the marketing infrastructure, competition, and internal constraints (Douglas & Wind, 1987). Later studies added culture as an important
factor to be considered when developing and adapting the standard marketing strategy (Dawar & Parker, 1994). Steenkamp (2001) also approaches this issue from a cultural angle—he developed a new cultural approach that combines Hofstede’s dimensions and Schwartz, resulting in a new framework for defining a marketing strategy.
The determining factor when deciding whether to use a standardized or adapted approach should be profitability, in other words, the standardized approach is appropriate when it has a positive impact on performance (O'Donnell & Jeong, 2000).
Following this brief review of the main arguments for using a standardized or tailored marketing strategy, the concept of brand strategy will be discussed in more detail. First a definition of brand strategy is given, then the influence of cultural factors on brand strategy is reviewed, and finally external market factors are considered.
2.2. International Brand Strategy
The present study focuses on international brand strategy, and we therefore first look at the definitions of brand strategy found in recent literature. Whitelock and Fastoso (2007, p. 266) define international branding as “a field within international marketing concerned with the challenges that companies face when their brands cross national borders. These challenges relate to the essence of the brand in terms of brand name, brand visual (e.g., logo, colours) and sound elements (e.g., jingles, music), and brand personality”. Cheng et al. (2005, p. 505), provide a broader definition of international brand strategy, referring to it as “the process of developing a firm’s brand equity that appeals to overseas target customers’ positive attitudes about the brand”.
Another relevant element in international brand strategy is the brand promotion, which refers to the execution of advertising in every market and the techniques in sales promotion of the brand (Hofer, 2015). The brand promotion includes aspects that help to differentiate and identify the brand while improving the brand’s performance (Hofer, 2015). This study will deal with the essence of the brand elements as defined by Whitelock & Fastoso (2007): brand name.
The research on international brand strategy has primarily addressed the factors that play a role in the decision on whether to use a standardized or adapted approach when dealing with a particular brand (Alahsban et al. 2002). The following section discusses various research studies that have been done on international marketing literature, specifically on international brand strategy.
2.3. Brand Strategy and Culture
A couple of studies have been conducted to identify the factors that affect the brand
adaptation strategy chosen. The following sections discuss the studies that identified culture as one of the main factors.
2.3.1. Cultural Differences
Roth (1995) identified some cultural factors that lead to brand adaptation or standardization. The study classified brands into different categories; social, functional, and sensory, then tested the market share of different types of brands by category and the Hofstede’s cultural dimensions. Regional socioeconomic factors were also considered to measure the performance of the brand in terms of market share. The conclusion of Roth’s article is that the environmental characteristics of a host country have two effects: they are important indicators of customer segment market potential and provide insights into the emphasis managers should place on the different types of brands (functional, social, and sensory).
2.3.2. Linguistic Differences
Other research studies have identified other cultural aspects that influence the adaptation of brands. Francis, Lam and Walls (2002) researched the impact of linguistic differences on international brand name standardization. The study used a sample of Fortune 500 companies and their brands in the Chinese market. The results of this study suggest that (i) firms can capitalize on the localization of their brand names by adopting names that are more likable in the host country than their original name, and that (ii) firms generally adapt their brand name when expanding into markets that do not allow easy phonetic and semantic transfer of the brand in question.
Usunier and Shaner (2002), in their study on how to introduce a brand into a new market, focus mainly on linguistics. Many cultural and language aspects should be considered when building a new brand, which firms can use to their advantage—by drawing on the more favourable aspects of the new market’s language— when adapting a brand to a new market.
2.3.3. Perception Differences
Foscht, Maloles, Swoboda, Morschett and Sinha (2008) conducted an interesting study about the perception of brands in different cultural environments. The study considered six different markets with their respective cultural environments. The study concluded that the same brand was perceived differently in different cultures, despite its identical positioning. This means that if a firm wishes to achieve the same brand perception in different countries, it needs to develop brand positioning strategies that emphasize the characteristics that enable consumers to perceive the product in a similar way. This study also indicated that even when one and the same brand name is used in different markets, the brand will be perceived differently.
As we can see, various cultural aspects definitely have an influence on the branding strategy selected. In many cases a brand necessarily has to be adapted in order to be appropriate for a given market. All the studies mentioned here point out that the larger the cultural difference between the host and the home country, the stronger the brand adaptation needed.
2.3.4. Psychic Distance
Some authors prefer to use a broader concept when talking about differences between countries—psychic distance, one of the most commonly cited but vaguely defined terms (Dow and Karunaratna, 2006). It can generally be defined as the sum of factors that affect the decision to do business between two different markets, and has been developed in the management-oriented international business literature. The construct was initially developed at the decision-maker level (Dow and Karunaratna, 2006).
Dow & Karunaratna (2006) contributed to the concept of psychic distance by developing and testing several indicators that are commonly used to measure this attribute, such as differences in language, religion, and political systems. This broader concept of differences between countries gives us a better understanding of the factors that influence decisions in
international business, because it not only considers cultural distance, but other factors as well. In recent years, psychic distance has increasingly been mentioned in international business literature (Dow & Karunaratna, 2006).
2.4. Brand Strategy and Market Structure and Size
In addition to identifying the cultural factors that affect the necessary degree of brand adaptation, research has aimed to identify other factors that affect the introduction of a brand in a given country. The key question in this context is whether the standard brand needs to be adapted to some extent to be successful in the host country or the standardized global brand will do. Alashban et al. (2002) researched the background and implications of brand name strategy, specifically the international strategy issue of brand name standardization versus customization.
The study identified some market structure factors that firms should consider before entering a new market. For example, they discovered that using a modified brand name should be considered when there is a large number of buyers and sellers in the market, and when distribution channels are highly coordinated. Another conclusion drawn by the study is that it is more likely the brand will have to be adapted when a company enters a rather distant country. In any case, firms should seek brand standardization whenever possible because it brings cost savings, and increases sales volume, which in turn leads to increased revenue. Only US companies were considered in this particular study.
Townsend et al. (2009) developed a framework incorporating aspects of environmental uncertainty, mimetic behaviour and experiential learning and their relation to the globalization of brands. Some of the hypotheses tested had to do with the relation between market size, cultural distance, and economic distance among other environmental effects on the one hand and the propensity to introducing a brand in a new country on the other hand. The study does not really address the use of a global or tailored brand when entering a new market, but rather whether a company decides to enter a new market. The study suffers from some limitations—among others, the fact that it only considers a single industry, the automotive industry—which may hinder the generalization of its results. If other concepts and specific factors had also been considered, the results would have been much more reliable.
2.5. Brand Adaptation
An important point to consider is what exactly is meant by adaptation and the decisions that have to be taken when deciding on the brand strategy. According to Erdogmus, Bodur and Yilmaz (2010) the decision on branding includes the following elements:
• Brand positioning: unique and distinct values that are emphasized in the way a certain brand is positioned;
• Visible brand elements: name, logo, package, label and product design;
• Brand peripherals: core branding issues related to general marketing strategy such as warranties and after-sales service.
The authors argue that firms generally tend to decide on standard or tailored branding on three different levels. Brand positioning is more directly related to marketing communication activities, the visible brand elements are the tangible brand components such as product design and features, and the brand peripherals concern infrastructural investment and long-term strategies. Most of the time, brand positioning and visible brand elements tend to be kept standardized, while brand peripherals are more often adapted to local markets (Erdogmus et al., 2010).
When developing a broader brand portfolio with different, customized brands for every market, it is crucial not to forget the importance of developing a clear, global brand architecture as well (Douglas, Craig and Nijssen 2001). The purpose of the Douglas et al. (2001) research is to examine the issues involved in building the comprehensive brand architecture. The central aim of branding is establishing the firm’s identity and building its position in the global marketplace among customers, retailers and other market participants. To be successful at this requires firms to develop a clear international branding strategy. Townsend et al. (2009) developed a framework classifying the different levels of brand in terms of their scope and use. They identify four different types of brands: (i) domestic brands, those that are used only in one (home) market; (ii) regional brands, those that are used in more than one market but all within the same geographical area; (iii) multiregional brands, those that are used in several countries in more than one geographical region but lack a standardized marketing program; and (iv) global brands, those that are sold in several geographical regions, including all major continents (North America, Europe, Asia); they also have a standardized marketing program.
2.6. Brand Standardization
Those favouring a standardized brand strategy argue that it allows economies of scale and scope to be achieved in marketing activities. Using a global brand may also speed up market entry by reducing the need for time-consuming modifications (Steenkamp, Batra & Alden, 2003). Another, newer argument is that consumers prefer brands with a “global image” over those of local competitors (Steenkamp, Batra & Alden, 2003). A standardized approach should only be used if maximizing profit is the ultimate aim (Samiee & Roth, 1992).
An intermediate solution for a successful brand strategy is proposed by Barron and Hollingshead (2004). They propose that a global brand approach be used to deliver a standard message in all markets, but instead tailor promotion to every individual market (Barron & Hollingshead, 2004).
2.7. Impact on Performance
The impact on performance has been discussed in general terms in section 2.1. When the aim of branding is to maximize performance, firms should favour a standardized approach (Alashban et al., 2002). However, as can be concluded from the relevant literature, several factors should be taken into account when deciding on the optimum brand strategy and the standardized approach will not be the best one under all circumstances.
It should be pointed out that some studies argue that there is little evidence to support one option over the other, if a company aims to maximize profits (Cavusgil & Zou, 1994). Samiee & Roth (1992) contend that the only valid reason for choosing the standardized approach is if it is indeed associated with higher profits.
2.8. Research Question
After reviewing some of the relevant literature, some general conclusions can be drawn. Different environmental and cultural factors affect the decision whether to use a standardized or adapted brand when entering new markets. Market structure is one important factor to consider in this context (Alashban et al., 2002). Cultural aspects should also be considered, and can be identified by using the Hofstede dimensions (Roth, 1995). The introduction of a
brand in a new market is also influenced by market size, and cultural and economic distance among other environmental factors (Townsend et al., 2009). Having a clear, global brand strategy and a comprehensive, global brand architecture, is also important in the different markets that a firm is present in (Douglas & Wind, 1987).
Now that the literature review has been completed, the research question can be formulated. Both internal and external relevant factors have been identified. Internal factors are linked to the company itself, while external factors lie outside a company’s control when entering a new market. Among the external factors, culture is very important, as are market conditions such as number of competitors and market size. While the market size of a given country is another important factor in deciding whether to use a standardized or adapted brand, it is one that has so far not received sufficient attention. Adapting a brand entails additional costs for a company (Alashban et al. 2002) and the bigger the market, the higher the costs of tailoring the brand to local conditions. A standardized approach, on the other hand, brings economies of scale and scope, resulting in reduced costs. Thus, the main research question that arises is: How does the market size of the host country affect the brand strategy?
The literature review revealed that the factor mentioned most often as explaining the decision to adapt a brand is culture. More specifically, the more distant a culture and environment, the more likely it is a company will opt for a customized brand (Alashban et al. 2002). Consequently, in this research the moderator effect of culture will be incorporated into this research study. As the main cultural aspect that influences the decision to adapt a brand is linguistics (Usunier & Shaner, 2002), linguistics will be incorporated as a moderator.
The present study will focus on beer, wine and spirits brands, since the consumption of alcohol differs between cultures, specifically because of different religions. Religion will therefore also be included as a moderator (Cochran, Beeghley and Bock, 1988). The Douglas Dow psychic distance framework will be used because it identifies religious and linguistic differences between countries.
3.
H
YPOTHESISD
EVELOPMENTFollowing the literature review of international brand strategy, next the hypotheses to be tested in the present study will be developed. As can be deduced from the research question, the market size of the host country will be the independent variable. The dependent variable will be the international brand strategy of the different companies, represented by the degree of brand adaptation ratio. The international brand strategy includes the decision on the type of brand to use in a given market—a domestic (local) brand, used only in one country, a regional brand, used in more than one market but all within one geographical region, or a global brand, used in several regions (Townsend et al., 2009). Companies thus face the struggle of adapting their own brand to a new market or using a standardized, global brand. In this context, companies must decide, among other things, on the visible brand elements, which include name, logo, package, label, and product design (Erdogmus et al., 2010). This study will specifically look at the brand name and whether to adapt it or use a standard, global name.
In the remainder of this chapter, the hypotheses to be tested and the theoretical arguments underlying them are presented.
3.1. Market Size
When a company decides to enter a new market, it looks at that market’s potential to generate sales and revenue, in other words, it evaluates the market’s attractiveness. One way to evaluate the market potential and attractiveness is by gauging the market size (Townsend et al. 2009). The bigger the market, the higher the potential demand, and thus the higher the potential to generate sales and revenue. Essentially, the market attractiveness as determined by market size is a basis for estimating demand conditions and market potential. It is expected that larger markets offer more favourable demand conditions. For companies, it is therefore key to develop a successful brand strategy in large markets because of their potential to generate sales and revenue (Townsend et al. 2009).
Large markets are usually also marked by a relatively high number of competitors (Alashban et al., 2002), meaning larger markets tend to be more competitive. In more competitive markets, standardization is harder to achieve because of the high number of competitors
(Alashban et al., 2002). Therefore, it is worthwhile for companies operating in larger markets to develop a marketing mix adapted to the local environment in order to be competitive and establish a strong position in relation to the other actors in the market (Alashban et al., 2002). The larger the market, the more adaptation is required, which implies a higher cost, but the higher sales derived from a large market cover the added costs of adaptation (Theodosiou & Leonidou 2003).
The type of product also plays an important role in determining the best approach to use. In a consumer goods market, purchasing decisions are based on a more “emotional” criteria as opposed to the “rational” approach seen in industrial goods (Theodosiou & Leonidou 2003). And a tailored brand works better in an “emotional” setting because it can appeal directly to the national tastes and habits that largely determine the purchasing decision in a consumer goods market (O'Donnell & Jeong, 2000). This difference between the industrial and consumer goods market is not only explained by the different criteria underlying the purchasing decision, as explained above, but also by the larger number of buyers typical of a consumer goods market. This leads to suggest an adapted approach in consumer goods markets (Roth, 1995).
The only argument that favours the use of a standardized approach is the higher profitability that can be achieved thanks to economies of scale and cost savings (Samiee & Roth, 1992). When entering larger markets with a high potential for sales and revenue, the higher sales that can be achieved with an adapted brand approach compensate for the added costs of brand strategy adaptation, resulting in an overall positive result (Theodosiou & Leonidou 2003). Thus, although a tailored approach entails higher costs (O'Donnell & Jeong, 2000), these higher costs are compensated by the potential for increased sales. This relation can particularly be observed in consumer goods markets, where brand adaptation is the preferred strategy (O'Donnell & Jeong, 2000).
Therefore, we hypothesize the following:
3.2. Culture
The literature review of international brand strategy and more in general the international marketing strategy revealed that one of the most important elements determining the strategy to follow is culture. The differences between the home and host country of the companies, in terms of cultural elements, are a key element that cannot be ignored by companies planning to go abroad. These differences in national market characteristics and cultural differences should be taken into consideration when developing the brand strategy (Özsomer & Altaras, 2008). Cultural differences are also important because cultural beliefs and norms are powerful forces that shape people’s perceptions, dispositions, and behaviours (Steenkamp, 2001), and this can lead to a different interpretation of the messages companies are trying to deliver.
For the reasons outlined above, culture is chosen as the moderator for our study. We consider two specific elements of culture that play the largest role in defining international brand strategy, which are discussed in the following sections.
3.2.1. Linguistic Differences
In the definition of the international brand strategy, linguistics plays an important role. Brands have visible elements such as name, logo, package, label, and product design, and have to be positioned in the consumer’s mind. Although different brands can have similar visible elements, the positioning can still be perceived in a different way because of different linguistic and cultural elements (Foscht et al., 2008). In fact, companies should take advantage of the phonetic and semantic differences between languages and adopt names that are more likable in the host country than their original name (Francis et al., 2002).
The greater the linguistic differences, the greater the need for adaptation of the brand’s visible elements. Usunier and Shaner (2002) also argue that firms should consider language aspects when entering a new country with a brand, and that this can be taken as an opportunity for companies to develop a likeable name in the host country. Integrating this cultural aspect in our initial hypothesis as a moderator, we hypothesize that linguistics has a positive impact on the relationship between market size and brand adaption, this effect being higher when the linguistic differences between the home and cost country are higher. The
literature reviewed also assumed that the higher the cultural difference, the higher the need for adaptation (Alashban et al., 2002).
Our second hypothesis thus is the following:
H2: Market size has a positive impact on the degree of brand adaptation, and linguistic difference has a positive moderating effect on this relationship.
3.2.2. Religious Differences
Religion still plays an important role in influencing consumer and social behaviour (Fam, Waller and Erdogan, 2004). Different religions tend to influence the way people live, the choices they make, what they eat and drink, and whom they associate with. Religion also has an impact on consumption patterns. These patterns are usually related to restriction of consumption of certain foods and beverages. For example, Jews and Muslims do not eat pork, Hindus do not eat beef, and drinking alcohol is forbidden or severely limited for Muslims and strict Protestants. Religion also has an influence on gender roles in some religions (Fam et al., 2004).
Some studies in the marketing literature argue that religion greatly influences purchasing decisions in two ways—first, by directly influencing religious codes of conduct, these codes can prohibit or suggest some purchasing decisions affecting the own personal choice, and second, by influencing attitudes and values toward consumption (Essoo and Dibb, 2004). Alashban et al. (2002) also add that religion can make specific products taboo in some societies. Therefore, we choose religion as the second moderator, and the effect of this moderator is deemed positive because as cultural differences (for instance involving religion) increase, so does the need for adaptation (Townsend et al., 2009).
We thus hypothesize that religious differences between the home and host country will positively influence the relationship between market size and brand adaptation. The larger the religious differences, the larger the positive influence (Alashban et al., 2002).
H3: Market size has a positive impact on the degree of brand adaptation, religious differences have a positive moderating effect on this relationship.
3.3. Overview of Hypotheses to be Tested
Our hypotheses are the following:
H1: Market size has a positive impact on the degree of brand adaptation.
H2: Market size has a positive impact on the degree of brand adaptation, and linguistic differences have a positive moderating effect on this relationship.
H3: Market size has a positive impact on the degree of brand adaptation, and religious differences have a positive moderating effect on this relationship.
As explained above, the moderators in our main hypotheses are cultural characteristics of the host countries. One of the most important cultural aspects that forces brand adaption is linguistics (Usunier & Shaner, 2002). Therefore, linguistic differences will be one of the moderators. Another important cultural aspect in our research is religion. Since our database is on alcohol products, religion is a major factor in product consumption across countries (Fam et al., 2004). The conceptual model presented in figure 1 shows the hypotheses and moderators in a graphic way.
4.
M
ETHODOLOGYThis thesis uses a quantitative research design. This research identifies the relation between the brand adaptation degree used by different companies and the size of the host market. Chapter 3 presented the hypotheses to be tested. This thesis takes a deductive approach—it starts from a theoretical basis, building on theories found in the relevant literature on international brand strategy.
The following sections present the sample that is used for this study, followed by the variables to be used to test the hypotheses.
4.1. Sample
For the research of this thesis, secondary data are used. The data used include information on the beer, wine, and spirits industry in Asian countries between 2006 and 2015. These data include market share and brands for all companies operating in the mentioned industries. On a country and industry level, the market size is included. The primary source of these data is Passport from Euromonitor International. Passport is a global market research database that provides data on industries, economies, and consumers worldwide, helping analyse the market context of different industries. Euromonitor International was established in 1972 as an independent company that provides unmatched detail and unbiased content for every region, country, and industry. Thanks to its extensive international network, it can provide detailed global, national, and local business information (Euromonitor International, 2017). For this study we selected the Asian countries, because they provide us a proper mix of developed and emerging countries. The sample comprises developed countries (with developed country economic indicators) and emerging countries (with lower economic indicators). The industry selected is alcoholic beverages, which consists of three subcategories: beer, wine, and spirits. This industry has seen an important growth in the last year. The most significant growth has been recorded in the beer category of the emerging markets—explained by the sustained economic growth and the rising of disposable income levels (Asia's Food Drivers: Rice, Oil and Meat, 2016). In the spirits category, a growth has been observed with new opportunities for multinational companies that are not yet present in the markets. The wine category has been marked by slower growth, especially influenced by
the Western lifestyle and health benefits that have been adopted by some sections of the Asian population (Asia's Food Drivers: Rice, Oil and Meat, 2016).
To have a comprehensive set of data, we selected the last 10 years with available data; more specifically, our data cover the period from 2006 to 2015. In this way, we obtain panel data with the same cross-sectional data observed during several years. The data selected were market size by category and country, market share by brand, and volumes by brand and by country.
In addition to the data we gathered from Passport, we collected data from the ORBIS database to get additional information on the companies included in the sample. From the World Bank database, we obtained information at the country level. Finally, we included the Douglas Dow psychic distance framework for the differences between countries in the psychic distance elements.
4.2. Variables
In the following section, the different variables that are used to conduct the study are presented. The relation with the hypotheses presented in the previous chapter is also mentioned.
4.2.1. Dependent Variable
The dependent variable of this study is the degree of brand adaptation. To measure this, the different brands are organized according to the framework developed by Townsend et al. (2009). As we only have information on the brand name and the countries or regions where each brand is used, a simplified version of the framework developed by Townsend et al. (2009) is used. To properly classify the brands, additional data were obtained from Passport, including the brand presence in the rest of the regions of the world. As a result, the brands could be classified based on the following criteria:
• Local brands: brands present in only one country;
• Regional brands: brands present in more than one country but within the same region (a continent, in this context).
To develop the ratio that is used, each brand is classified as either adapted (that is, using a local strategy) or standardized (that is, using a global strategy). Local brands utilize a local strategy, while global and regional brands utilize a standardized strategy.
The ratio is calculated as follows. First, we looked at the brand portfolio of each company in the different markets/countries they are active in. The total number of brands using a local/adapted strategy is divided by the total number of brands of the company, to generate the adapted brand ratio per company. In short, the following equation was used:
Brand Adaptation Degree = Number of local brands per company per country / Total number of brands per company.
4.2.2. Independent Variable
The independent variable of this study is market size. The data on market size are obtained from the Passport dataset and are classified by category and then for each country for the relevant years. Market size is expressed in United States Dollars (USD). This variable is present in all hypotheses that are tested. Market size is an important factor in defining an international marketing strategy—the larger the market, the more brand adaptation is required (Theodosiou & Leonidou 2003). Market size is also used as a market attractiveness measure, providing a basis for market potential and abundant demand conditions (Townsend et al., 2009).
4.2.3. Moderator Variables
The moderator variables are the measures presented by Dow & Karunaratna (2006) as psychic distance stimuli. Specifically, our moderators are the differences in psychic distance stimuli between host and home country for every company in our dataset. Religious and linguistics differences are considered, as indicated in hypotheses 2 and 3.
For the language measurement, Dow & Karunaratna (2006) applied the following approach. Three different measures were established. First, the distance between the major languages for each pair of countries is measured on a 1–5 scale, 1 representing the same language and 5 representing different language families. Indicators 2 and 3 represent the proportion of the population of country A than can speak the language of country B, and the other way around. These indicators also range from 1 to 5, 1 representing more than 90% and 5 less than 1%.
Finally, the indicator that is actually used is built by reducing the above three indicators to a single factor using confirmatory factor analysis.
A similar procedure is used for religion. First, the distance between the major religions for each pair of countries is measured on a 1–5 scale, 1 representing the same denomination and 5 representing different religion families. Indicators 2 and 3 represent the proportion of the population, also on a scale from 1 to 5, and 1 representing more than 90% and 5 less than 1%. Finally, the indicator that is actually used is built by reducing the above three indicators to a single factor using confirmatory factor analysis.
The ratios for each country pair are the moderator variables that are used in the study, every country pair involving the home country of the company and the host country.
4.2.4. Control Variables
Control variables are included to eliminate possible biases. The control variables that are used are company variables and country variables.
For company variables, the ORBIS database is used to get the relevant data. The following are the control variables on company level:
• Company size, measured in assets per year, in USD. This control variable is expected to have a positive effect on the dependent variable. Larger companies have more resources at their disposal, which are needed to support the additional adaptation costs (O'Donnell & Jeong, 2000).
• Company experience, measured in company age. A positive effect is expected from this variable on the dependent variable, as more experience leads companies to consider the different tastes and preferences of the markets and develop a tailored international brand strategy (O'Donnell & Jeong, 2000).
• Nationality of company. This variable is included as a control dummy variable with all the countries of origin.
For country-level control variables, the database of the World Bank is used. The following are the country-level control variables:
• Population of the host country, measured in number of persons. This variable is expected to have a positive relation, because a bigger market favours adaptation (Theodosiou & Leonidou, 2003).
• GDP of the host country, measured in USD per year. This variable is expected to have a positive effect on the dependent variable, since a bigger market requires more adaptation (Theodosiou & Leonidou, 2003).
• GPD per capita of the host country, likewise measure in USD per year. This variable is expected to have a positive effect on the dependent variable, as higher GDP per capita means more purchasing power, and therefore a higher need to adapt to local tastes (Alashban et al., 2002).
Additionally, Douglas Dow psychic distance measurements are added. As these variables indicate differences between countries, the coefficient is expected to be positive, as the higher the difference between countries, the higher the degree of adaptation needed (Roth, 1995). The ratios added are the following:
• Educational difference: this ratio is measured with three ratios: literacy, secondary education, and tertiary education;
• Industry development difference: this ratio is measured with nine different development ratios, as among others, GDP, energy consumption, and number of cars, among others.
• Political difference: this ratio is measured by the degree of democracy and the political ideology of the party in power.
Finally, the year is also included as a control variable. It is included as a control dummy variable with the 10 years that are considered for the research.
Table 1 summarizes the variables presented.
Table 1. Definition and Source of Variables
Variable Code Definition Source
Adaptation Ratio Adapt_Ratio Ratio between adapted and total brands per company
Passport (2017)
Market Size Market_size Market size per country in USD
Passport (2017)
Language Difference Dow_lang Language difference ratio Douglas Dow (2006)
Interaction 1 Interact_1 Adaptation Ratio x Language difference
Religious Difference Dow_rel Religious difference ratio Douglas Dow (2006)
Interaction 2 Interact_2 Adaptation Ratio x Religious Difference
Educational difference
Dow_edu Educational difference
ratio Douglas Dow (2006) Industry development difference
Dow_indu Industry development
difference ratio
Douglas Dow (2006)
Political difference Dow_pol Political difference ratio
Douglas Dow (2006)
Company Size Company_asset Assets ORBIS (2017)
Company
Experience Company_age Company Age ORBIS (2017)
Nationality of Company
Home_country Home country of company
ORBIS (2017)
GDP per capita GDP_cap GDP per capita in host country
The World Bank (2017)
GDP GDP_host GDP host country The World Bank
(2017)
Population Pop_host Population in host country
The World Bank (2017)
Year year Year Passport (2017)
4.3. Empirical Strategy
To answer the research question presented earlier, the three hypotheses presented above are tested. The research method used—the IBM SPSS Statistical program—is one that is able to test hypotheses.
As mentioned before, data were first gathered from different sources, and next uploaded into SPSS in order to perform the different statistical tests. First, descriptive statistics and correlation tests were performed to detect excessive correlation between the independent variables. Some categorical variables were reshaped as dummy variables (year, nationality). Finally, a regression analysis is performed. A linear regression is used to test the different hypotheses. This particular approach was selected because of the continuous characteristics of the dependent variable.
The model is the following:
!"#$%#%&'( *#%&' = ,-+ ,/0#*12% 3&42 + ,56&(78&3%&93 :&;;2*2(92 + ,<=2>&7&'83 :&;;2*2(92 +
,?@"89#%&'(#> :&;;2*2(92 + ,AB("83%*&#> :2C2>'$D2(% :&;;2*2(92 + ,EF'>&%&9#> :&;;2*2(92 +
,GF'$8>#%&'( + ,HI:F + ,JI:F $2* 9#$ + ,/-K'D$#(L !72 + ,//K'D$#(L 3&42 + ,/5M2#* +
5.
R
ESULTSThis chapter presents the results of the statistical test performed in SPSS on the data. First some descriptive statistics are presented, including the dependent, independent, and control variables. Then a correlation matrix is presented, which presents the correlation between the independent variables and the dependent variable. The dummy control variables of year and company nationality are not included in this first stage, but are included in the final regression analysis. Then the regression analysis results are presented.
5.1. Descriptive Statistics
The beer category is selected for further analysis because a beer market exists in all the countries included in the data sample. Of the total of 2,410 observations we have of companies operating in one of the Asian markets per year, only 1,297 are valid observations. This is because we only have information on company age for 1,780 companies and on company assets for 1,311 companies.
The adaptation ratio goes from companies that customize all their brands (1) to companies that only use standardized brands (0). The smallest market in the sample is the Pakistani market and the biggest is the Chinese market. Population ranges from about 4 million in Singapore, the smallest country, to more than 1 billion in China, the biggest country. of the country with the lowest GDP per capita (US$654) is Vietnam and the country with the highest GDP per capita (US$56,000) is Singapore. With this indicator, we can observe the differences in development between the countries included in the sample.
The company age goes up to 111, the oldest company being Bitburger Braugruppe GmbH from Germany. The biggest company in terms of assets is Anheuser-Busch InBev N.V. from Belgium (with more than US$142,550,000 in assets). The smallest company in this context is Mount Shivalik Industries Ltd. from India.
Table 2: Descriptive Statistics
Variable N Min Max Mean Std. Dev.
1. Adaptation Ratio 1297 0,00 1,00 0,74 0,33 2. Language difference 1297 -2,43 0,53 0,02 0,81 3. Religious difference 1297 -1,30 1,53 0,41 0,62 4. Educational difference 1297 0,00 2,07 0,53 0,50 5. Industrial development difference 1297 0,00 2,48 0,84 0,76 6. Political difference 1297 0,00 2,07 0,76 0,69 7. Market Size 1297 34,50 81.487,00 13.307,00 21.596,00 8. Host country population 1297 4.401,00 1.371.220,00 418.439,00 571.480,00 9. Host country GDP 1297 10,23 13,04 11,83 0,72 10. Host country GDP per capita 1297 654,28 56.000,29 14.984,80 16.486,99 11. Company age 1297 0,00 111,00 75,00 40.879,00 12. Company assets 1297 6.973,00 142.550.000,00 25.229.068,00 31.921.713,00 5.2. Correlations
Table 3 presents the Pearson correlation coefficients between the dependent, independent, and control variables of this study. The adaptation ratio has a positive correlation with all the variables included, showing the highest correlation with the religious difference index (r = 0.323, p = 0.01) and with the language difference index (r = 0.053, p = 0.05). The ratio has has a positive correlation with the independent variable of market size (r = 0.099, p = 0.01). There is also a positive correlation with the control variable on country level; the correlation coefficient (r) of the population size of the host country is 0.071 (p = 0.01); for the GDP, r = 0.139 (p = 0.01), and for the GDP per capita, r = 0.107 (p = 0.01).
The control variables on company level also reveal a positive correlation for company age, the correlation is r = 0.052 (p = 0,05) and for company size, the correlation coefficient r = 0.277 (p = 0.01).
As to the additional psychic distance, the following correlation coefficients with the dependent variable are found: for educational difference, r = .316 (p = 0.01); for industrial development difference, r = 0.313 (p = 0.01); and for political difference, r = 0.326 (p = 0.01).
The highest correlation observed between an independent variable and a control variable is, as expected, the correlation between the control variable host country GDP and the independent variable market size (r = 0.710, p = 0.01). There is also a high correlation observed between the control variables host country GDP and host country population (r = 0.789, p = 0.01). Although these correlations are the highest, they are not high enough to eliminate the GDP variable.
Table 3: Correlation Matrix Variable 1 2 3 4 5 6 7 8 9 10 11 12 1. Adaptation Ratio 1 2. Market Size ,099** 1 3. Language difference ,053* ,122** 1 4. Religious difference ,323** ,090** -,112** 1 5. Educational difference ,316** -,021 -,119** ,425* 1
6. Industrial development difference ,313** ,008 -,082** ,279** ,689** 1
7. Political difference ,326** -,005 ,145** ,279** ,471** ,552** 1
8. Host Country Population ,071** ,532** -,005 ,169** ,189** ,126** -,057** 1 9. Host Country GDP ,139** ,710** ,101** ,104** ,049* -,031 -,204** ,789** 1 10. Host Country GDP per cap ,107** -,005 -,131** ,016 -,117** -,250** -,057** -,426** -,037 1 11. Company Age ,052* -,133** ,211** ,085** ,005 ,047* ,174** -,180** -,159** ,041 1 12. Company Assets ,277** -,073** ,119** ,123** ,197** ,159** ,282** -,117** -,068* ,036 ,196** 1
Note:
* Correlation is significant at the 0,05 level (2-tailed) ** Correlation is significant at the 0,01 level (2-tailed)
5.3. Regression Analysis
For this study, several linear regressions are performed. The first model includes only the control variables and the psychic distance ratios of linguistics and religious difference. The second model includes the first independent variable, market size. Next the first interaction is included. The fourth model includes all the control variables, the first independent variable, and the second interaction. Finally, a regression with all the mentioned variables is performed. The results of these regressions are presented in table 4.
The first regression included only the control variables (language difference, religious difference, educational difference, industrial development difference, political difference, host country population, host country GDP, host country GDP per capita, company age, company assets, year dummy, and home country dummy). This model has a significant (p-value <0.01) F (p-value of 75.253 and R-squared (p-value of 0.689. The coefficients for the psychic distance variables are positive, as expected, but only religious and educational distance are significant. The company experience and size coefficients are also positive and significant. The coefficient of GDP per capita in the host country is also positive and significant. The coefficients for host country GDP and population size are not significant.
The second model tested Hypothesis 1, and for this the independent variable market size is included in the regression together with the control variables of the first model tested. Model 2 has the following statistically significant (p-value <0.01) values: F = 73.215 and R- squared = 0.689. The significance level of the independent variable market size is p-value = 0.993. This result means that we cannot accept hypothesis 1. Also, the coefficient of market size is 0, meaning that market size would have no effect on the degree of brand adaptation. The variables on company level have statistically significant coefficients (p-value < 0.01) and are positive, as expected, in line with what the literature indicates. We can also conclude that more experienced and bigger companies tend to use a customized approach when it comes to brands in host countries (O’Donnell & Jeong, 2000)—the company size as measured by company assets has a coefficient of 0.131, reflecting the highest impact on the degree of brand standardization. The GDP per capita proves to have a positive influence on the degree of market adaptation, with a p-value < 0.01 and a coefficient of 0.073. We can thus conclude that companies in wealthier countries tend to develop an adapted, tailored approach when it
comes to branding, in an effort to get closer to the customers and increase sales (Alashban et al., 2002). In line with the results on the market size variable, the host country GDP and population size have no significant coefficients. This can be explained by the fact that these variables are all highly correlated with market size. The cultural difference control variables all have positive coefficients, as expected, religious difference being the only statistically significant coefficient (p-value < 0.01).
Model 3 tested the second hypothesis. To test this, the first interaction term was created, that is, the interaction between market size and the Douglas Dow index of linguistics difference. The model tested has a statistically significant (p-value <0.01) F value of 73.437 and an R-squared value of 0.695. Although the coefficient of the interaction variable is significant (p-value <0.01), the coefficient is negative (-0.131), a result that is contrary to what had been predicted. We predicted a positive moderating effect of linguistic difference between host and home country, but the results actually support contrary negative relation. This means that in a bigger market, linguistic difference will have a negative moderating effect on the degree of brand adaptation. Therefore, hypothesis 2 is not supported.
The control variables have a similar outcome as in model 2. Company size and experience have significant (p-value <0.05), positive coefficients. GDP per capita also has a significant (p-value <0.01) and positive coefficient of 0.075, with a limited impact on the degree of brand adapation. Host country GDP and population size have no significant coefficients. The psychic distance control variables all have positive coefficients, as expected, language and religious difference having the only statistically significant coefficients (p-value <0.01). The coefficients have a limited effect on the dependent variable, the language coefficient being the highest among the psychic distance ratios (0.096); religious difference has a coefficient of 0.085.
The third hypothesis is tested in model 4, and for this, the second interaction term is included in the regression. This interaction term is market size and the religious difference measured according to the Douglas Dow ratio. Model 4 has a significant (p-value <0.001) F value of 74.529 and the R-squared of the model is 0.689. The interaction 2 coefficient is statistically significant (p value <0.01). The coefficient is positive (0.159), as predicted. From these results, we can infer that in a larger market, a higher religious difference between the host
and home country has a positive moderating effect on the degree of brand adaptation. Therefore hypothesis 3 is supported.
The control variables have the following outcome: positive (0.116) and significant (p-value <0.001) for company experience, the variable that has the highest influence on the degree of brand adaptation. Although the coefficients for the psychic distance factors are positive, they are not significant.
The final model (#5) included all the independent and control variables. The F value is 74.675 and the R-squared value is 0.695.
Table 4: Linear Regression Results Models 1–5
Model 1 M1 Model 2 Model 2 Model 3 Model 3 Model 4 Model 4 Model 5 Model 5
Std Beta Sig Std Beta Sig Std Beta Sig Std Beta Sig Std Beta Sig
Control Variables
Language difference .018 .391 .018 .395 .096 .000 -.003 .884 .072 .006
Religious difference .082 .000 .082 .000 .085 .000 -.012 .622 -.007 .762
Educational difference .059 .034 .059 .035 .037 .182 .028 .319 .008 .784
Industrial development difference .002 .948 .002 .948 .009 .729 .003 .903 .010 .692
Political difference .014 .557 .014 .607 .005 .847 .011 .671 .003 .911
Host Country Population .012 .769 .012 .787 .046 .301 .007 .876 .040 .365
Host Country GDP -.004 .918 -.004 .944 -.013 .798 .003 .951 -.006 .903
Host Country GDP per cap .073 .003 .073 .003 .075 .002 .039 .113 .041 .092
Company Age .086 .001 .086 .001 .078 .003 .116 .000 .108 .000
Company Assets .131 .003 .131 .003 .100 .022 .090 .039 .061 .158
Year dummy included Yes Yes Yes Yes Yes
Company origin country Yes Yes Yes Yes Yes
Independent Variables Market Size (H1) .000 .993 -.016 .593 -.002 .948 -.017 .564 Interaction 1 (H2) -.131 .000 -.126 .000 Interaction 1 (H3) .159 .000 .155 .000 Model Details N 1297 1297 1297 1297 1297 R Square .689 .689 .695 .689 .695 F-value 75,253 .000 73,215 .000 73,437 .000 74,529 .000 74,675 .000
