Master thesis
One brand strategy does not fit all: the international success of branding in emerging and developed markets
Keywords: international management, global brand strategy, local brand strategy, market share, developed markets, emerging markets
Name: Gabriel Leonard (Jelle) van Oosten Slingeland Student number: 6120407
University of Amsterdam – Business administration – International Management Supervisor: Dr. Vittoria Scalera
Statement of Originality
This document is written by Gabriel Leonard van Oosten Slingeland, who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document are 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
Two specific brand strategies have emerged since the globalization of products. Both a global brand strategy and a local brand strategy have brought multinational enterprises successes. A global brand strategy is effective via integrating homogeneous features all over the world, while a local brand strategy responds to local cultures and demands. Up to now, many of previous researches regarding these strategies have been performed in a setting with developed markets. In contrast, this thesis researches if a global brand strategy is more successful than a local brand strategy. Second, this result is tested in the context of developed and emerging markets, where I hypothesize that a global brand strategy is more effective in a developed market than it is in an emerging market. A sample filled with data of firms operating in the beer, spirit and wine industry in South East Asia, from the period 2008-2015, is used to test the hypotheses. A random effects analysis tests the panel dataset. The results of both hypotheses were significant. A global brand strategy is more successful, in terms of market share, than a local brand strategy. Subsequently, this effect is positively moderated if a global brand strategy is implemented in a developed market. Implications are proposed for managers how to use these outcomes. This thesis is ended with limitations and advises for future research on this topic.
Table of content
Introduction 5
Literature review 8
The importance of market share 8
Brand strategies 11
Global brands 12
Local brands 14
Emerging and developed markets 16
Research question 18
Theoretical framework 20
The relation between a local or global brand strategy and market share 20 The moderating effect of developed and emerging markets on
hypothesis 1 23 Conceptual model 26 Methodology 26 Empirical setting 26 Variables 29 Dependent variable 29 Independent variable 30 Moderating variable 30 Control variables 32 Empirical strategy 35 Results 37 Descriptive analysis 37 Correlations 38 Hausman test 40
Random effects analysis 40
Robustness check 42
Discussion and conclusion 44
Managerial implications 47
Limitations and future research 48
Acknowledgement 52
Introduction
Globalization has made it easier for firms to internationalize. This resulted in the internationalization of firms, and the introduction of brands in different countries. Main drivers for globalization are the diminution of barriers for investment and trade, and technological innovation. This enhanced international trade, movement of capital, dissemination of knowledge and migration. Friedman (2005) metaphorically argued that the world is flat and that from a commercial perspective all competitors in the marketplace have the same amount of opportunity.
Perhaps with this thought in mind, the fast food chain Taco Bell internationalized to China in Shenzhen and Shanghai in the 80’s. Taco Bell expected a success after earlier success stories like KFC in China, but the opposite happened. Taco Bell had to shut down their stores, finding out that the Chinese were not so hungry for Mexican food. However, Taco Bell did not give up. They expanded to South Korea in 2010, and this time it resulted in a success. It pointed out to be a strategic move, since Taco Bell opened their new stores in cities with a big nightlife scene, attracting foreigners who are yet familiar with the brand.
This example shows that firms can be successful in more than one country. Some benefits go hand in hand with being present in multiple countries. One of these benefits can be economies of scale: producing products or services at a lower cost per unit. Economies of scale give firms a competitive advantage and therefore market share (Buzzell, Gale & Sultan, 1975). Market share has, besides economies of scale, multiple drivers for growth. Return on investment, customer loyalty and a proper brand strategy are examples of variables that enhance the growth of market share (Miller & Camp, 1985; Smith & Park, 1992; Rust & Zahorik, 1993; Abratt &
Motlana, 2002). Thus it is essential for firms to invest in these aspects, and with the case of brands, invest in a proper brand strategy (Urde, 1994).
Brands differentiate products and link the product or service with certain associations. From an international perspective, brands can be positioned across different geographical regions, while others are positioned in one country only. These are known as global and local brands (Van Gelder, 2004). Global brands are well known for their consistency in quality and firms can make use of economies of scale in for instance advertising. On the other hand, local brands have a close connection to local cultures and are perceived as unique. In international strategy, focusing on a global brand is called a global strategy, focusing on a local brand is known as a multinational strategy. Implementing the best of both is a so-called transnational strategy (Bartlett & Ghoshal, 1988). The global strategy focuses on the integration of earlier successes in new markets in order to obtain cost efficiencies. The multinational strategy has a focus on responsiveness. Local demand drives this strategy, and products should be customized according local demand.
Previous research regarding brand strategies and fundamental thoughts with regards to branding has often been performed in a western setting (Aaker, 1991, Keller, 1993, among others). Emerging markets have very different characteristics than developed markets, which makes doing business there differently. Emerging markets are less developed in factor markets, infrastructure or institutions (Hoskisson , Wright, Filatochev and Peng, 2013). Emerging markets also have less branded products sold in relation to developed markets, and therefore less branded competition (Sheth, 2011). These differences make it highly important to take into account when firms internationalize and managers have to choose between implementing a local or global brand strategy, because consumer demand changes per region. Differences
between countries are bigger than generally acknowledged, and should be identified by managers in order to be successful across borders (Ghemawat, 2001). Up to now, there is inadequate scientific based information on this topic. Contemporary literature suggests that there is a link between the stage of development of a country and the profitability of the multinational enterprise (Makino, Beamish & Zhao, 2004). However, there is no information available on how the stage of development of countries moderates the success of local or global brand in relation to change in market share for multinational enterprises. This is important for managers to know if they make use of a country portfolio analysis. If firms internationalize, they want to know where their firm can be successful, and with which strategy (Perlitz, 1985, Ghemwat, 2001). The stage of development of a country on these effects is a subject of further research (Brouthers & Xu, 2002). Therefore, this thesis will focus on how implementing a local or global brands strategy affects the market share growth of the brand, when moderated by doing business in either a developed or emerging market.
The outcomes of this thesis will contribute to the international business theory by bringing nuance in the discussion of seeking for either integration or responsiveness when a business goes cross borders. This thesis will clarify which of the two strategies is more successful for brands, and if this success if moderated by different stages of development of countries. Managers have thus specific information for the optimal branding strategy if they decide to internationalize. Secondly, managers have new and more specific insights for their country portfolio analysis via segmenting countries on a hard measure of segmentation of the development stage of countries (MSCI, 2017).
In order to bring these insights, this thesis will first elaborate why market share is important for firms, and how a global or local brand strategy can potentially
contribute to market share. This relation is also put in a different context. Indeed, this thesis will elaborate if the effect between a global or local band strategy and market share is moderated by the development stage of a country. The focus will be on looking at two stages of development, namely a developed market or an emerging market (Hoskisson et al., 2003) After elaborating on present theories regarding these topics, I propose two hypotheses to test the relation between these concepts. I will follow this up with a methodology in which the researched variables and empirical strategy is explained. This is followed by the results. Lastly, I present the discussion and conclusion, implications and limitations of this research.
Literature review
The importance of market share
One of the most researched topics in business is market share. “Market share is the percentage of a market (defined in terms of either units or revenue) accounted for by a specific entity” (Farris, Bendle, Pfeifer & Reibstein, 2010, p. 32). The main reason for the importance of market share in business literature might be that market share is one of the determinants for firm profitability (Anderson, Fornell & Lehmann, 1994). In fact, three factors explain the main reasons for this correlation (Buzzell et. al, 1975). The first aspect of this correlation is that firms obtain economies of scale in for instance procurement, recruitment and advertising. The obtained experience curve, introduced by the Boston Consultancy Group, leads to more sales versus a lower cost per produced unit, leading to growing profits (Boston Consultancy Group, 1972 in Day & Montgomery, 1983). The second aspect originating from a high market share that should lead to profitability is market power. The size of a firm should give it the advantage of bargaining more effectively, and thus realize higher sales prices for their
products or services. Losing bargaining power leads to losing profits as a firm (Foss and Pedersen, 2004). Thirdly, the quality of management relates links a high market share to profitability. Skilled managers are more successful in obtaining higher market shares in their respective markets. Besides, they are also better in creating optimal output from their employees, lower production costs and train themselves in being more emotional intelligent. Moreover, when a firm obtains a higher market share, it is easier to sustain that position than it is for the competition to catch up (Day & Montgomery, 1983).
The importance of market share and the correlation to profitability differs per industry. Infrequently purchased goods are more often durable and have a higher cost per unit. Consumers tend to pre-evaluate the purchase of these sorts of goods longer and more often. The higher costs lead to a higher risk for the consumer. For that reason the consumer is willing to pay a premium for the assured quality. With consumer goods, often frequently purchased, there is a lower risk attached. This means that the risk occurring when buying the product in unequal to the satisfaction of the consumer. Relative market share is of less importance in the market of products that have lesser risks attached. This is because it is easier to catch up with other firms with high market share in these markets (Day & Montgomery, 1983).
So if firm profitability is obtained via a high market share, it should be the aim for firms to grow. Why do not all firms achieve this high market share? Probably because they lack the capabilities to achieve what it takes to have a relative high market share. A higher retention rate, customer satisfaction and thus consumer loyalty result in higher market share (Rust & Zahorik, 1993). This implies that firms need customers that keep on buying from their firm, and therefore avoid the competition. Aaker & Shansby (1982) show a relation between effective positioning by the end
product user or attribute and growth of market share. Effective in this context means that the product suits the demand of the market. And perhaps quite obvious, Miller and Camp (1985) found a positive relation between return on investment and market share.
Abratt & Motlana (2002) find, among others, that there is also evidence that brand strategies lead to higher market share. More specifically, a brand extension is more successful in contrast to positioning an individual brand (Smith & Park, 1992). A brand extension means that a firm makes “use of established brand names to launch new products” (Völckner & Sattler, 2006, p. 18). The implementation of Coca-Cola Zero in relation to the, at that given time, already existing Coca-Cola product brands is an example of such a brand extension. Sullivan (1992) showed the opposite for pioneers, saying that introducing a product via brand extension leads to smaller market share in relation to introducing it via a new brand. One reason to support this view might be that innovative consumers perceive less risks and greater decision confidence (Klink & Athaide, 2010). Thus, these consumers switch easier to new brands.
However, firms should be vigilant for trying to satisfy all customers because customer satisfaction can also lower market share (Rust & Zahorik, 1993). Trying to serve all can result in a firm to be stuck in the middle. Data can give deep insights in the demand of consumers, and hence a firm can choose the most profitable customer segment (Van den Driest & Weed, 2014). Also, firms can apply the strategic solution of larger or a more diversified brand portfolio. Firms can introduce different brands that all serve different profitable markets, and it does not risks the brand to become to diversified in their offer (Rego, Morgan and Fornell, 2013).
Firms should also invest in their brands if consumers in the market find brands important in their decision-making. Moreover, firms lose market share when they underinvest in their brands (Fischer, Völckner & Sattler, 2010) Thus, the existence of a brand, and a purposeful strategy behind it are essential to grow a firms or brands market share
Brand strategies
Brands bring significant financial contributions for firms. Starbucks and Apple can charge premiums for their products because of their brand value. Buyers also tend to be loyal to brands, meaning that brands create future sales for the firm (Simon & Sullivan, 1993). Brands also point out the difference between what a firm and the competitor offers, in which brands can contribute to the success of firms (Wood, 2000).
Brands are all over civilized parts of the world. A brand, or so-called brand description “is a name, term, sign, symbol or design, or a combination of these that identifies the goods or services of one seller or group and differentiates them from those of competitors” (Kotler & Armstrong, 2010, p. 985). This can also be named as the brand description. This description is tailored to the needs and wants of a consumer via the marketing mix, well known of consisting of price, place, product and promotion, and sometimes also complemented with people, physical evidence or process (Kotler et al., 2010). The success of the implementation of this marketing mix will determine the brand strength. The stronger the brand, the more loyal consumers it will attract. This means that the brand value will rise, and create future revenues for the firm. From a strategic point of view, brands should be managed as a long-term
asset. A proper measurement of brand description, brand strength and brand value can be modified in a brand strategy (Wood, 2000).
The marketing mix can lead to an increase of brand knowledge. Consumers conceptualize this knowledge within an associative network. This network is composed of two components; brand awareness and brand image. These associations can be widely chosen by the managers, and should be in line with customer preferences and corporate goals (Keller, 1993).
Brand strategists can choose to extend existing brand names into new products if existing brand names have a positive association with the target group. This is known as brand extension, a well-known source for strategic growth for many firms since it helps companies to reduce advertising expenses. Also, it raises the acceptance of new products since the brand is already in the consumers mind (Aaker, 1991). Multinational enterprises can effectively use brand names if they introducing new products, since the brand is already well known. On the other hand, multinational enterprises seek for local connectedness when introducing new brands. This thesis will continue to elaborate on the difference between global and local brands as a brand strategy.
Global brands
The globalization of corporate brands has been a huge success. Brands like Toyota, Coca-Cola, Vodafone and Siemens are globally well known (Sheth, 2011). The positive global recognition allows companies to use their corporate brand name all over the world. A global brand strategy is partly the result of the globalization of the worldwide economy. Brands of products can be easier implemented in new economies because of the high brand recognition, and thus many multinational
enterprises favor global brands (Steenkamp, et al., 2003). Global brands are advantageous for some reasons. Firstly, it gives firms the chance to exploit economies of scale (Aaker & Joachimsthaler, 2000). Economies of scale are gained if a higher amount of production results in a lower cost per unit. Lower costs in for instance marketing are obtained because of the global identity of the brand. Secondly, global brands can be used to create synergies between countries (Holt, Quelch, & Taylor, 2004). Macro-economical factors can be influenced by the globalization of certain brands. Tesla for instance disrupts political thinking and infrastructural decision making by establishing a paradigm shift in transportation. Lastly, global brands are associated with quality, status and prestige. Global customers perceive global brands as high quality products because of their international status, and therefore perceive a good price to quality ratio (Holt et al., 2004).
Multinational enterprises should pay attention to some aspects in order to successfully create a global brand. According to Aaker et al. (2000), multinational enterprises must identify the components of creating a brand, and the capabilities of the firm in order to establish global brand leadership. The components of the brand refer to the brand identity of the position of the brand in the market in relation to the competition, and the brand architecture. Brand architecture is the structure of brands within the organization and gives relations between the brands. Firms can for instance choose to give their parent firm the same name as their main product, like Coca-Cola and Heineken do. However, brand names can also be names differently than the parent name. Multinationals like AB inBev and Diageo do not position product brands with either AB inBev or Diageo. However, these firms are well known for brands like Budweiser and Johnnie Walker. And the capabilities of the firm refer to building brand programs by accessing multi media and measure the results, and organizational
structure and process, clearing responsibility for the brand strategy and integrate the strategy within the management.
One of the ways a global brand is more successful is when they tend to adapt to local needs and features. Managers can use insights of national cultures in order to create new brand strategies if firms have the knowledge and capabilities to do so (Roth, 1995). A brand strategy that does so is the glocalization strategy. The idea of the glocalization strategy is that multinational enterprises position global brands, but adapt their features according to the local markets (Hollensen, 2007). Economies of scale and local responsiveness go hand in hand in this strategy. On of Heineken’s famous global brands Amstel adapts the alcohol percentage according to the demand of the local market. Fanta, one of Coca-Cola’s global brands, changes the color and sugar amount per market. Southern European and Middle Eastern countries prefer higher amounts of sugar in their Fanta than North European countries do.
As mentioned earlier, global brands are positioned in in countries and regions all over the world. However, definitely not all firms decide to position their brands globally. Their antagonists are called local brands, and have their own features.
Local brands
Local brands are defined in this thesis as brands that are implemented in one country only (Van Gelder, 2004). Multinational enterprises acknowledge the positive outcomes of local brands. One of these outcomes is the high local recognition to the brand. Local brands achieve this by making use of the nation brand advantage. This nation brand advantage is the representation of emotions, cultures and identification of a nation, and represents this in marketing communication. This concept is relatable to the nation as a whole; “it describes the country’s intangible assets without any
explicit links with a product” (Fan, 2006, p. 6). Nation branding relates more to cultures and believes than it does to products. Firms can identify nations identities, and create new products according to the identity of specific cultures.
Three product market factors determine or influence the preference for making use of a local brand strategy as a firm (Douglas, Craig & Nijssen, 2001). The first factor is the target market. Local brands are more effective if target markets are internationally heterogeneous, because demand varies heavily. This means that niche markets can be targeted via a local brand. These markets are known for the aversion regarding the homogeneous identity of a global brand. Some markets are relatively homogeneous, such as the luxury market. Brands like Lamborghini are purchased all over the world in this market. But others markets have heterogeneous demand, varying per country. Restaurants with a well-known local image do not often seek for internationalization (in contrast to the fast food market).
The second factor is cultural embeddedness. Some cultures have deep-rooted preferences for certain products, product variances or products that are an integral part of a culture. Football (or soccer for Americans) is in many European countries the biggest sport. The mania for football events is far less in Australia or North America. And on the other hand is American football one of the biggest sports in the United States of America. Some American Football brands are well established there, but have little value in the rest of the world.
The third factor is the competitive market structure. This depends on the degree of domination of local or global brands within a certain market (Douglas & Craig, 1996). Some markets are more profitable for local brands than others. If a market is the most profitable by positioning local brands, multinational enterprises can anticipate by differentiating their brand portfolio by creating local brands and
acquiring local brands. Conclusively, by doing this a multinational enterprise can be profitable in more markets, only by adapting their brand strategy to a local brand strategy.
In contrast to local brands, the success of global brands can be addressed to the fact that consumer groups are relatively homogenous across markets, and so can standard products be developed (Steenkamp et al., 2003). Sheth (2011) disagrees on this insight that international markets are homogeneous, especially between emerging and developed markets.
In some international product markets, local brand are perceived to be more successful like in the food industry. This is mainly because of cultural roots of the country, and thus local brands can be better tailored to the unique needs and wants in the market (Dimofte, Johansson and Ronkainen, 2008). On the other hand, global brands can be perceived to be successful in other product markets because of the standard it delivers, among others. These factors do not only relate to product markets, but also to emerging or developed markets (Özsomer & Altaras, 2008).
Emerging and Developed markets
Hoskisson et al. (2013) propose that emerging and developed markets differ from each other on several factors, which can be classified in two groups. The first group consists of factor market developments and infrastructural developments. The amount of development of these aspects determines the economical opportunity of a nation (Porter, 1990). The higher the factor market development and infrastructural development, the higher the economical opportunity of a nation. Factor markets and infrastructural development directly influence the production of goods and services. These thoughts are in line with the industry-based view of Porter (2008). This view
proposes that the financial success of a firm depends on the profit potential of the industry and the actions taken to obtain a part of this profit potential. The variance in the latter two determines the variance in the financial performance of the firm. The second group consists of institutional development. Institutions are the formal and informal rules of the game (North, 1990). They are humanly devised constraints that “are formed to reduce uncertainty in human exchange” (North, 2016, p. 73). Formal rules like laws and regulation are guidelines for the short term, where informal rules like habits and cultures are often inherited and established centuries ago (Williamson, 2000). The idea that institutions matter is in line with the institutional-based view (Peng, 2002). This view got more recognition with the rise of importance of the emerging market context for multinational enterprises, since institutional frameworks are mostly heterogeneous around the world.
Both institutions and factor markets help firms to capture economical opportunities like for instance foreign direct investment (Hoskisson et al., 2013). They classified four different markets according to the groups. Developed markets have a high factor market and infrastructural development, and as well a high institutional development. Examples of these are for instance Germany, Canada and Japan. Firms in these countries are used to constantly improve their value chain in the domestic country, and continuously pursue to do so in foreign countries (Porter, 1990, Chang & Hong, 2000).
The group on the other side of the spectrum is the group with traditional emerging economies. This group experiences a low degree of factor and infrastructural development, and also a low degree of institutional development. Most emerging countries would be categorized in this group 20 years ago (Hoskisson et al., 2013). However, much have changes and a large group of emerging countries have
developed themselves on either institutional development or factor or infrastructural development. These are defined a mid-range emerging economies.
Sheth (2011) proposed five fundamental distinctions in characteristics between developed markets and emerging markets. Businessmen and marketers should thus change their assumptions of marketing when targeting emerging markets as western multinational enterprise. The first characteristic of emerging markets is the heterogeneity of markets. This is because markets in emerging markets are often locally oriented, small-scaled enterprises having the lead and operations are low scaled. Markets are also skewed towards the bottom-of-the-pyramid with consumers who live of an income of less than two dollars per day. The second aspect is social political governance. Emerging markets are more influenced by for example, religion, NGO’s and local communities. This results in more monopolistic markets, where the monopolist is often state owned. The third aspect is unbranded competition. According to Sheth (2011) 60% of the competition in emerging markets is unbranded. Unbranded products are especially sold in rural areas due to a lower average income and poor infrastructure. Also, households are both consuming and producing units, making them less dependent on external products. The fourth difference is the chronic shortage of resources in emerging markets. Examples are electricity, lack of skilled labor and insufficient supply of materials. This influences the consistent quality of production. The last difference is an inadequate infrastructure. This is not limited to proper road and rails, but also access to banks and point-of-sale terminals.
Research question
Traditional and contemporary literature has brought us a paramount of insights in brand strategies, market share, and the difference between emerging and developed
markets. Branding giants such as Aaker (1990) and Keller (1993) among others have discussed the importance and potential success of a proper brand strategy. And these brand strategies can also be seen from an international perspective. If a firm decides to position its brand in one country, one can call it a local brand strategy, while if a brand is positioned in more than one region, a firm applies a global brand strategy. A proper brand strategy leads to a growth in market share, and underinvesting in brands makes firms lose market share (Fischer et al., 2010).
Many characteristics decide if a global or local brand strategy is the most successful strategy for a firm to use. Specific market demand makes it necessary to adopt specific brand strategies. Global brands have a well-known reputation in other countries, which might help introducing it in another country. In contrast, local brands can be more easily adapted to local needs, and might therefore suit specific demand better. Also, the demand of the market varies per region. There is especially a huge difference between emerging and developed markets. Developed markets often have capitalistic systems, where uniformity in the offer has been the norm for many decades, and thus created economies of scale for the firm. However, marketers can make use of more and more data nowadays to adapt the products to the local needs. Emerging markets often have markets with unbranded competition. Globalization of markets however means that also these markets experience branded competition nowadays.
There is up to now no scientific evidence if global or local brand strategies are more successful in obtaining market share when taking the development stage of international markets into account. The development stage might significantly moderate the gap between the successes of different international brand strategies. This development also changes over time, and thus needs constant reexamination.
Brand strategies might be changed over time in order to be successful in different international markets (Hsieh, 2002). Also, earlier work (Batra, Ramaswarny, Alden, Steenkamp & Ramachander., 2000, Heinberg, Ozkaya & Taube, 2017) only focus on the success of local of global brand strategies in emerging markets, but do not compare this with developed markets. A specific brand strategy might be successful in an emerging market, but can be even more successful in another. This thesis comes up with the following research question in order to test the best international setting for either a global or local brand strategy:
“How does the choice of a local or global brand strategy affect the market
share for a firm, and is the relation affected by the development stage of the country where the brand is positioned?”
Theoretical framework
The relation between a local or global brand strategy and market share
Brand strategies are a well-known tool for marketers to add value to the brands and thus to the firm. As earlier specified, two brand strategies are specifically relevant in an international business context. A firm can either position the brand as a local brand, which is only available in one country only, or position the brand as a global brand, which is positioned in multiple regions over the world. Global brands have certain advantages. One of these advantages is that global brands leverage economies of scale (Aaker et al., 2000). Economies of scale give costs advantages to the parent firm, meaning that the costs per produced unit decreases if more production is made. Global well-known brands are yet well known, and therefore less advertisement is needed in a more globalized world. As a result, it can create even more awareness for the brand, since the brand familiarity is yet higher since the brand
has crossed the consumers mind. The more money a firm can spend on brand advertisement, the more market share it will have (Weiss, 1968).
Another benefit of global brands is that it creates synergies between countries In cultural terms speaking, brands can appeal to different feelings. Heineken positions itself as a beer for people all over the world, for those who like an alcoholic beverage, but live and drink responsibly. The effect of people all over the world who want to be part of such a brand, together with the country of origin effect that might be relevant for some products, give the brand an advantage in another country when it competes to its local antagonists (Holt et al., 2004). Thirdly, global brands are associated with quality, status and prestige. Some cultures, as for instance the Spanish and Russian culture among others over the world, pay high attention to what others purchase, and see this as a quality mark. Global brands are perceived as products of high quality, and might thus motivate others to buy it too, resulting in a higher market share (Steenkamp et al, 2003, Holt et al., 2004). In the end, global brands deliver quality for a reasonable price, and give the buyer a certain feeling to be a world citizen, regardless of where the brand originated.
Global brands are perceived as very impactful and powerful assets of multinational enterprises. They can offer consumers the feeling of joy and being part of more than just their own country of nation. On the other hand, these global institutions can sometimes do more harm than joy. Some perceive global brands as a severe interference with their own culture, believes and freedom. Anarchists demonstrate against the capitalistic change the world experiences. Even though people share many things, nations have its own believes and values. Huge shifts are made to local brands, which can more easily adapt to local demands (Dimofte et al., 2008).
Local brands should not only be described as “only available in a specific geographical region” (Dimofte et al., 2008, p. 118), but local brands also have the unique proposition of being perceived as original and unique, and proudly represent the local culture and country. Local brands have traditionally benefited from the fact that they have a close relation with their consumers. This makes some local brands icons for countries, since they have such close relations with the local culture and people (Dimofte et al., 2008). Also, local brands have a higher amount of responsiveness to the specific demand. Global brands cannot directly react to radical changes in demand with regards to their products. McDonalds has experienced difficulties in reacting to a demand for healthier food. Local brands can do so easier because they only sell to one country, and can therefore adapt their features to this demand only.
However, Global brands have shown to be successful in multiple regions, since it gave them the resources to constantly internationalize. This growth in resources gives them the power to innovate and invest in parts where growth can be created, such as integrating big data (Van den Driest & Weed, 2014). The marketing or management team also has more data available, since it has operated in more regions yet. Local brands do have the powerful connection with the local community, which able them to purposeful position themselves. But in general this is backed with less resources, to create the total experience consumers demand nowadays. Global brands can invest in multiple touch points and personalized offers, which makes it easier for country brand teams to acquire new customers (Aaker & Joachimsthaler, 2000).
And even though global brands experience protests and they sometimes lack the ability to fully adapt to the local needs of the consumers, they do know how to
create and facilitate a global conversation, based on shared views. Loyal consumers have a stake in the aggregation of market share, which often initiate these conversations (Rust et al., 1993). Knowing that the earlier success of global brands able them to achieve economies of scale results in the following hypothesis:
“H1: A global brand strategy is more successful in terms of market share than a local brand strategy”.
The moderating effect of developed and emerging markets on hypothesis 1
Firms can positively increase the perception of the brand by linking the characteristics of a country to the brand. Dutch firms can for instance put a French flag on the label of a baguette. Consumers can also automatically have a certain perception of a country to certain products or brands. People tend to have a natural preference for Italian pasta, German cars or Swedish design. This is called the country of origin effect. Traditional analyses mention that consumers have a natural preference for local or regional products over foreign products (Nagashima, 1970 & Bannister & Saunders, 1978). On the other hand, there is also evidence that global brands are preferred over local brands (Steenkamp et al, 2003, Holt, Quelch, & Taylor, 2004). The reason for this is that global brands are associated with prestige and quality. However, there might be a different strategy than using global brands needed when targeting emerging markets (Yu, 2003).
Marketers should rethink existing perspectives and practices between developed countries and emerging countries in marketing (Sheth, 2011). In the western world one focuses on the country of origin effect, which is closely related to products. The effect is meaningless without the product. In line with the characteristics of emerging markets, marketers should focus on the nation brand
advantage when targeting emerging markets. The nation brand advantage concept is relatable to the nation as a whole; “it describes the country’s intangible assets without any explicit links with a product” (Fan, 2006, p. 6). Nation branding relates more to cultures and believes than it does to products. Companies can use the country’s name or certain logo’s to emphasize the country of origin effect. The main purpose of using the nation brand effect is to leverage exports or sales. There are cases where one can see that the country of origin effect can be exploited all over the world. Selling pasta as an Italian product works all over the world. However, the nation brand effect is nation dependent. It differs per nation which intangible assets should be highlighted. Examples are race, language, social institutions, famous persons or a nations history. Firms should diversify their product portfolio in order to be successful in emerging markets when the goal is to highlight the nation brand effect (Fan, 2006).
It is questionable if classical approaches in marketing and branding can be utilized all over the world, knowing that characteristics are different between emerging and developed countries (Sheth, 2011). Previous brand extension research performed by leading scientists like Aaker & Keller (1990) and Sunde & Brodie (1993) all use mainly western brands. Other branding strategies or product diversification strategies should be taken into account since demands are different all over the world.
There are different perspectives in contemporary literature about how branding in emerging countries should be done to achieve growth in market share. As mentioned before, Sheth (2011) suggests that emerging countries can be better approached with a non-branding strategy. This idea is in contrast to the results of Brouthers & Xu (2002). They found support that Chinese exporting firms should implement a branding product strategy to less developed countries in order to achieve
a significant increase in performance satisfaction. However, countries and firms develop, and in line develop the appropriate optimal product strategy that fits the market the most. Up to now there is no information on how the economic evolvement in Asia influences the brand strategy that is optimal to fit the market.
Not all countries are the same and therefore brand strategies should be adapted to specific countries. The development of a country has effect on profitability of the firm (Makino, Beamish & Zhao, 2004). The importance of changing characteristics makes it essential to identify these differences.
In the end, branding nowadays can work both in emerging and developed markets. But there is more than one brand strategy to be chosen by a multinational enterprise. The characteristics of global brands are that they create economies of scale, have a perceived higher quality, and that they create synergies between countries. Consumers in developed markets find quality perception more important than consumers from emerging markets (Morgeson III, Sharma & Hult, 2015). One can also argue that globalization by itself is more experienced in developed markets than emerging markets, and have thus more experience and exposure to merging with different countries.
Emerging markets do develop themselves. Midrange emerging markets distinguish themselves from traditional emerging markets in either a development in factor markets and infrastructure, or a development in institutions. This means that the gap to developed markets shrinks, and differences in characteristics like social political governance, unbranded competition and inadequate infrastructure are fading away. However, there is still a gap, which makes it less easy for global brands to successfully operate in emerging markets than in developed markets. This is why this thesis proposes to following hypothesis:
“H2: The relation in hypothesis 1 between a global brand strategy and market share is stronger for developed markets than it is for emerging markets”.
Conceptual model
Methodology
Empirical setting
This thesis has the purpose to research the relationship between either local or global brands on market share. Furthermore, as a moderating effect, this relationship will be tested in both a developed market and emerging market context. A panel dataset is used in order to test the proposed hypotheses. The dataset includes firms present in the liquor, wine and beer industry in Japan, Singapore, Hong Kong, Kazakhstan, India and China during 2008 – 2015. These six countries are chosen for several reasons. Firstly, these countries are chosen because of their geographical location. The beer market is very fast growing in Asia. The beer consumption in Asia has grown with 50% from 2003-2012, and it accounts for 33% of the total beer Global brand strategy
versus
Local brand strategy
Developed market versus Emerging market Market share H1 H2
consumption in the world (Ascher, 2012). Also the wine consumption in the Asia pacific region has grown with 31.5% from 1997-2011 (Dal Bianco, Boatto, Caracciolo, 2013). This growth is mainly due to changes in global economics, resulting in the globalization of wine and therefore a growing consumption. (Tang, Tchetchik & Cohen, 2015). The economic growth compensates the high import tariffs in Asia, which are the highest in the world (Del Bianco, Boatto, Caracciolo & Santeramo, 2015). However, the government of Hong Kong decided to dismiss the payment of duties on wine, resulting in a growth of wine consumption of 54% between 2008-2012. Secondly, I chose these six countries for research validity reasons. In order to test a sample with sufficient brands I chose for all three developed countries in Asia, which are Japan, Singapore and Hong Kong (MSCI, 2017). As emerging markets, China and India were picked as the two biggest Asian countries within this classification. These two countries have both experienced high economical growth during the period 2008-2015, and have also accepted foreign direct investment from multinational firms. Kazakhstan on itself was added as a third country to give a balance in global and local brands positioned in the emerging market group in the dataset. This thesis further elaborates on this choice in the section Moderator on page 29 and 30.
The dataset is composed out of different databases to test the hypotheses. As also further explained later in this chapter, a dependent, independent, moderating and control variables are included in the dataset. For the original dataset I retrieved the variables market share, brand strategy and the nationality of the firm from the Euromonitor Passport database. The Euromonitor Passport database provides information on economies, industries and consumers worldwide, from over 106 countries (Euromonitor). I complemented this dataset with the variables market
development, firm size, firm age and regulatory quality. Market development was decided upon the classification of MSCI. MSCI is “an independent provider of research-driven insights and tools for institutional investors” (MSCI). MSCI classifies countries as a developed, an emerging or a frontier markets. Firm size and firm age were derived from the Orbis database. The Orbis database provides financial information of firms, varying from income statements to balance sheets, together with information on international ownership structures for more than 130 million firms around the world (Kalemli-Ozcan, Sorensen, Villegas-Sanchez, Volosovyeh & Yesiltas, 2015). And finally the data regarding regulatory quality is retrieved from the Worldwide Governance Index. The Worldwide Governance Index provides individual and combines governance indicators for over two hundred countries from the period 1996-2016. The index provides six indicators; Voice and Accountability, Political Stability and Absence of Violence, Government Effectiveness, Regulatory Quality, Rule of Law, and Control of Corruption.
The initial sample includes 2115 observations for 367 brands. This means that each brand per country has on average 5.8 observations, varying from 1 to 8 observations per brand per country. Missing data had been manually added to the dataset, based on annual reports and information from Nikkei. However, not all data could be added to complete this dataset. Thus, I excluded the observations where there was no information available for total assets and market share. This results in an unbalanced dataset. The level of analysis of the dataset is brand level, and distinguished per country.
Variables
In the following part I further explain the different variables used for this analysis, grouped as a dependent, independent, moderating or control variable.
Dependent variable
The dependent variable is market share. The need to have figures on market share for a firm is because sales forecasts can either be attained by a growing market share or a in total growing market. The first is always more difficult to achieve. Strategic and tactical actions can change the competitive landscape and thus the allocation of market share (Farris et al. 2010).
Firms seek for sustainable competitive advantage to foster long-term profits. It is essential to be national or international successful as a firm (Porter, 1990). Market share is one of the factors that determine sustainable competitive advantage and sustainable profitability (Buzzell et al., 1975). Market share is increasingly important for international firms because it enhances market power and economies of scale. Among other factors, a successful branding strategy relates to a growing market share (Fischer et al., 2010).
The dependent variable will be measured via the relative market share per brand per country during the time period 2008-2015. The Euromonitor Passport database has numbers on market share available for brands in the six countries that are analyzed and will thus be used in this thesis. Market share is calculated as sales revenue divided by the total market revenue.
Independent variable
For the independent variable I distinguish two types of brand strategies: a local and a global brand strategy. The independent variable is called brand strategy since firms actual consider if a brand should be a local or global brand, based on the potential across cultures, markets and countries (Van Gelder, 2004). Making a choice for either a local or global strategy can have serious implications for a firm. A firm might not use the full potential of a brand when it is not introduced in more than one country, or the investment of the brand placement in different countries has been not the optimal one, which leads to higher opportunity costs (Liu & Buck, 2009).
The data regarding brand strategies are derived from The Euromonitor Passport database. Brands from firms located in 22 different countries were used. These countries are Japan, Hong Kong, Singapore, China, Kazakhstan or India. The 22 home countries are respectively Australia, Belgium, Chile, China, Denmark, France, Germany, Hong Kong, India, Italy, Japan, Kazakhstan, Mexico, Netherlands, Philippines, Portugal, Russia, Singapore, South Korea, Spain, United Kingdom and the United States of America. A global brand strategy is calculated as a ratio of the firm’s global brands to the total number of brands that it positioned in the host country. Brands that had a global brand strategy were allocated a 1; local brand strategies were allocated a 0.
Moderator
The moderating variable is the stage of development of a country and is created to test hypothesis 2. The development stage of a market can in this thesis either be an emerging or developed stage, in line with the classification of MSCI. This thesis focuses on countries in East Asia. In this region there are three developed
markets: Japan, Hong Kong and Singapore (MSCI, 2017). I include all three countries in this analysis for several reasons. Firstly, more countries added in this analysis increase the external validity of the results. Adding more countries in East Asia to the dataset improves the generalizability to other countries within this region. Secondly, due to lack of data, more countries are needed in this analysis to find statistically representative results.
I added three emerging markets to the dataset in order to see the difference between the brand strategies in both emerging and developed countries. These countries are China, India and Kazakhstan. The choice for these countries is mainly based on an equal distribution of observations between the developed and emerging market group. The MSCI also has a frontier market classification, in which Kazakhstan is included. I decide that Kazakhstan can also be included in the emerging market group for this thesis, based on the earlier mentioned classifications of Hoskisson et al. (2013). Kazakhstan has experienced a certain amount of both institutional growth and infrastructural growth, making it an emerging market. Also, the MSCI mainly categorizes the countries based on investment decisions. However, the GDP per capita in Kazakhstan was $10506. For China this was $8069, and for India $1596 (The World Bank, 2017). GDP is recognized as one of the main determinants for managers to choose a country for doing foreign direct investment. The higher the GDP, the better is generally seen the choice of investment (Ghemawat, 2011). Also this reason justifies the choice to put Kazakhstan as a home country in this dataset. The MSCI classification of markets is the best classification that can be used for the purpose of this thesis, but I acknowledge the flaws of their classification.
A new variable is created to segment the host countries of the brands into a dummy variable, where Japan, Hong Kong and Singapore (developed markets) are allocated a 1; China, India and Kazakhstan (emerging markets) are allocated a 0.
Control variables
A list of control variables is added to the research in order to optimize the research model and to gain more robust result. These control variables are: Nationality of the firm, firm size, firm age and regulatory quality.
An important factor for international brand strategies is the nationality of the firm. If firms did not originate in the home country it might experience some problems in obtaining market share. One of these problems is that it might experience psychic distance, which makes it harder for a firm to be successful (Johansson & Vahlne, 1977). Dummy variables are created in the dataset for the variable nationality of the firm. A 1 indicates that the host country is the same as the home country of the firm, a 0 indicates that the home and host country are not in line.
The size of a firm is quantified in this thesis by the total amount of assets of a firm. Total amount of assets are computed as total assets of the parent firm divided by their brands. Data for the total amount of assets is derived from the Orbis databank. The firm size variable is correlated with the market share variable (Johannisson & Lindström, 1971) and is thus needed to add as a control variable. Namely, among potential other reasons, the size of a firm can be related to exploiting economies of scale (Chao & Kumar, 2010) that may lead to a higher market share (Buzzell et al., 1975). I used a logarithm for the variable firm size to control for outliers and skewness in the distribution.
Another added control variable is the age of the firm, measured in terms of the years after establishment of the parent firm of the brand. There is a robust and significant negative relationship between firm age and profitability, namely because of rent seeking behavior and organizational rigidities of the aging firms (Loderer & Waelchli, 2010). Younger firms are more flexible and can better capture organizational value and growth, measured in terms of market share, than their older competitors. This is mainly due the lack of structure and guidance in their decision-making, thus making them more flexible and reactive to the market (Anderson & Eshima, 2013). The age of the firm is measured via the incorporation date of the firm, and then adjusted for the relevant year in the dataset (2008-2015). I used a logarithm for the variable firm age to control for outliers and skewness in the distribution.
Political stability is also added to the model as a control variable. The Worldwide Governance Indicators provide information on governance. Governance consists of the traditions and institutions by which authority in a country is exercised (The World Bank, 2017). Institutions partly define the extent of a country to be developed or not (Hoskisson et al., 2013). And institutions not only affect a firm’s strategy (Peng, Sun, Pinkham & Chen, 2009) but also the performance of the firm. If firms do not prepare themselves for entering countries with weak institutions, they will have a disadvantage to those who have this knowledge (Luo & Peng, 1999). As mentioned earlier, The World Governance Indicators consist of 6 indices. For this thesis is chosen to add the variable Regulatory Quality – “capturing perceptions of the ability of the government to formulate and implement sound policies and regulations that permit and promote private sector development.” (Kaufmann, Kraay & Mastruzzi, 2011). The main reason to add this variable is because of the last part of the explanation. Foreign direct investment develops private sectors. And if
governments promote this as well, it is easier for foreign brands to be successful in the host country. However, emerging markets might have weaker institutions than developed markets (Hoskisson et al., 2013). For this reason this variable is added to control for this effect. The composite measures of the regulatory quality follow a normal distribution, providing values from -2.5 to 2.5. The higher the value, the better the regulatory quality. I used a logarithm for the variable regulatory quality to control for outliers and skewness in the distribution. By picking some countries out of the dataset I risk that the chosen variables are to skewed. For than reason the variable is converted via a common logarithm.
The variables that are incorporated in the dataset and therefore used for the research and analysis of the hypothesis are represented in table 1. The table includes, in line with the structure of the methodology, the dependent variable, the independent variable, the moderating variables and the control variables. For each variable is given the name of the variable, a short description of the variable and the original source of the variable.
Table 1 variable overview
Name of the variable Description of the variable Source of the variable
The independent
variable
Global brand strategy The brand is positioned in more
than one region Passport database The dependent
variable
Market share
The total market share of a brand in the home country, specified per year
Passport database The moderating
variables
Developed market The host country of the brand positioning is either Japan, Hong Kong or Singapore
MSCI
The control variables
Nationality of the firm
Converted to a dummy variable; 1 if the host country is in line with the nationality of the parent firm, 0 if not
Passport database
Firm size Total assets of the firm divided by
the amount of brands per firm Orbis database Firm age
Year of observation - the incorporation date of the parent firm
Orbis database
Regulatory quality
Values from - 2.5 to 2.5,
following a normal distribution, the higher the value, the better the regulatory quality of the country
The World Governance Indicators database
Empirical strategy
I will use a random effects analysis to test the hypotheses with the panel dataset. The analysis is performed in Stata. Stata is chosen as analytical tool because it can perform analyses for time series, in opposition to SPSS. Both a fixed effects analysis and a random effects analysis can be performed in order to test the research models if a panel dataset is used. “The random effects model assumes exogeneity of all the regressors and the random individual effects. In contrast, the fixed effect model allows for endogeneity of all the regressors and the individual effects” (Baltagi,
Bresson & Pirotte, 2003, p.361). A Hausman test can be performed in advance to test which of the two suits for the analysis of the models. A fixed effects analysis should be performed if the Hausman test rejects the null hypothesis (Hausman, 1978). Also I chose for a log transformation of the variables Firm age, Firm size and Regulatory quality in order to make the distributions less skewed.
This thesis will test three models to explain the potential significance of the two hypotheses. The first model will test the effect of the control variables and the moderating variable on the dependent variable. The second model will test the effect of the independent variable on the dependent variable. Thus the independent variable is added in to model 2. The third and last model tests the moderating effect of market development on the outcome of model 2. The interaction term (Global brand strategy*Developed market) is added to the model, which is model 3.
1. Y market share = α + β Firm age + β Firm size + β Nationality of the firm + β Regulatory quality + β Developed market
2. Y market share = α + β Firm age + β Firm size + β Nationality of the firm + β Regulatory quality + β Developed market + β Global brand strategy
3. Y market share = α + β Firm age + β Firm size + β Nationality of the firm + β Regulatory quality + β Developed market + β Global brand strategy + β (Global brand strategy * Developed market)
Results
Descriptive analysis
I performed a descriptive analysis, represented in Table 2. In this analysis I integrated the dependent, independent, moderating and control variables. In the second row one can see the number of valid observations per variable, followed by the total amount of valid observations. The table also shows the minimum value, maximum value, mean value and standard deviation of each variable. The values of market share are presented in percentages, meaning that a value of 0.10 is in line with 0.1% market share for a brand in the respective market. The variables Global brand strategy, Developed market and Nationality of the firm are dummy variables. This explains the values between 0 and 1 in Table 3. I included these to check the distribution of the values. The means of the values are respectively 0.33, 0.57 and 0.47. This implies for instance for the variable developed market that 57% of the values are from an emerging market, since emerging market is listed as 0, and developed market as 1. The data should be adjusted if one of these values would be extremely high or low. This is not the case, so the data can be used for further analysis.
The variables Firm age, Firm size and Regulatory quality are converted according to a common logarithm. The values are changed accordingly, and thus do not show the actual minimum and maximum values that are originally retrieved from the other databases.
Table 2 Descriptive statistics
N Minimum Maximum Mean Std.
Deviation Market share 2202 0,10 24,90 1,51 3,06 Global brand strategy 2202 0 1 0,33 0,47 Developed market 2202 0 1 0,57 0,50 Nationality of the firm 2163 0 1 0,47 0,50 Firm age 2156 0 5 3,53 1,300 Total assets 2200 5,576 18,78 15,16 2,70 Regulatory quality 2202 1,5 4,4 3,62 0,98 Valid N 2115 Correlations
The second test I perform is to test for linear correlation. This test is performed to measure the strength and the direction of linear relationship between the variables used in the models.
Table 3 Correlation matrix Market share Global brand stategy Developed market Firm
nationality Firm age
Total assets Regulat ory quality Market share Pearson Correlati on 1 Global brand strategy Pearson Correlati on 0.185* 1 Developed market Pearson Correlati on 0.020 0.175* 1 Firm nationality Pearson Correlati on -0.065* -0.590* -0,037 1
Firm age CorrelatiPearson on 0.093* 0.236* 0,029 -0.196* 1 Total assets Pearson Correlati on 0.038 0.435* 0.301* -0.257* 0.112* 1 Regulatory quality Pearson Correlati on 0.018 0.044* 0.699* 0.100* 0.086* 0.159* 1
*. Correlation is significant at the 0.05 level (1-tailed).
Table 3 represents the correlation matrix. The output shows a positive and significant correlation of 0.185 (p<0.01) between the independent variable global brand portfolio and the dependent variable Market share. Other significant outcomes on the dependent variable Markets share are Firm nationality, -0.065, and Firm age, 0.093, (p<0.05) Firm nationality shows a negative relationship, which is in contrast with what this thesis earlier proposed. One of the reasons for Firm nationality to be negative might be that the parent firm had followed the Uppsala model of internationalization. This implies that firms first acquire knowledge and capabilities to overcome their psychic distance, and thus build enough competencies to successfully internationalize (Johanson et al., 1977). In this case, their combined ability to
overcome psychic distance and their experience from other markets makes them perform better than local firms.
All variables in the model have a significant correlation with the dependent variable Global brands. Market share, Developed market, Firm age, GDP per capita and Total assets show a positive effect, all as expected so far. This is again in contrast with firm nationality.
Hausman test
The data analyses were performed in Stata/SE 15.0. The reason for this is that a panel data set is used, and Stata can analyze panel datasets, in contrast with for instance SPSS. There are two ways Stata can analyze a panel data set. This is either via a fixed effect or a random effect analysis. I tested the most preferable test via the Hausman specification test (Hausman, 1978). I first tested model 3, including all the variables and the interaction effect, via a fixed effect analysis, subsequently via a random effect analysis. The outcome of the Hausman specification test was prob > chi² = 0.5470. This implies that the null hypothesis is supported and thus a random effect analysis is preferred.
Random effect analysis
This thesis will analyze three models, as discussed in the methodology part, to test the two hypotheses. The outcomes of the random effects analyses are represented in Table 4 and further discussed per model.
Model 1 includes the five control variables, the moderating variable and the dependent variable. I run a random effect analysis in Stata to test the significance of the model, and the variance explained by the control variables and the moderating
variable on the dependent variable. The results of model 1 is that it is significant at p=0.041. The only significant variable in this model is Total assets (β = 0.071, p=0.032). Total assets have a positive significant effect on market share in this model. All other control variables are insignificant at a p=0.05 level. The R² of model 1 is 0.008. This outcome is very low, since model 1 only explains 0.8% of the market share.
Model 2 includes the four control variables, the moderating variable, the dependent variable and the independent variable. This model is used to test the outcome of hypothesis 1. The outcomes of the random effect analysis show that model 2 is significant at p=0.000. The test of hypothesis 1 shows a positive and significant outcome with a p=0.000 and a Z-score of 3.68. The β of the independent variable Global brand is 1.424. Therefore, Hypothesis 1 is supported. All other variables in this model are insignificant. The R² of model 2 is 0.044. This means that 3,6% of the market share of a firm is explained by the global brand strategy of a firm. Model 3 includes the control variables, the moderating variable, the dependent variable, the independent variable and the interaction term between the independent variable and the moderator. This model is used to test the significance of hypothesis 2. The outcomes of the random effect analysis show that model 3 is significant at p=0.000. The outcomes of hypothesis 2 shows a positive and significant outcome with a p=0.000 and a Z-score of 4.44. The β of the interaction effect is 2.728. Therefore, hypothesis 2 is supported. The only other variable besides the interaction effect that is significant in this model is the variable Developed market (β=-1.028, p=0.007). The R² of the model is 0.089. This means that 4.5% of market share is explained by global brands that are positioned in developed markets.
