STRATEGIC BRAND MANAGEMENT IN EMERGING MARKETS: CONSUMER PERCEPTIONS OF BRAND EXTENSIONS

. The purpose of this paper is to examine consumers’ perceptions of the parent brand and their attitudes towards the brand extensions by exploratory researching the topic in the new context of one emerging European market. Theoretical part gives a comprehensive description and analysis of the main features of the brand extension strategy. The purpose of the primary research is to explore and understand Croatian consumers’ perceptions about the global parent brand (Coca-Cola) and their attitudes towards its extensions on the Croatian market. Research results support the assumption that managers should introduce brand extension when they are sure of the position of the parent brand and its former extensions in the consumers’ mind. As such, results corroborate previous findings, showing how it is necessary to know the existing level of brand affect, brand loyalty, and brand trust since they have different impacts on consumers’ perceptions of the brand extension. The main contribution of the paper stems from the replication and extension of the previous research in the new context (in terms of the level of the market development and consumer culture). Research findings add to the existing knowledge with insights and results from the new market (Croatia), and as such might help in increas-ing understanding of the brand extension phenomena in the field of the strategic brand management in emerging markets.


Introduction
Considerable evidence in marketing supports the assumption that the brand extension strategy allows the company to expand its business in the new areas, primarily thanks to the security that a well-known parent brand brings. According to Nielsen's global research (2015, p. 14), brand extensions are "approximately three to four times more common than new manufacturer and new brand launches combined. " However, this strategy should be used with caution -it cannot be applied to all products, and it can happen that consumers do not accept the brand extension.
Despite existing numerous research dedicated to the brand extension strategy in the developed markets, the knowledge about consumers' attitudes toward the brand extensions in the emerging markets is still limited, insufficient and based mainly on research findings from the context of large Asian markets. For example, Fu et al. (2009) empirically tested the extended model of Aaker and Keller brand extensions evaluation framework in China and have found additional significant factors for consumers' evaluation of brand extensions in that particular emerging market. In researching the specificities of the model of the brand extension impact on parent brand-equity in India, Dwivedi & Merrilees (2013) findings have corroborated previous ones from developed markets showing that brand extension attitude significantly influences change in brand-equity of the parent brand. In addition, Joshi and Jadav (2017) have examined the effects of brand extension strategy on parent brand equity in Indian FMCG industry and have proved that brand extensions affect parent brand equity. In the context of other emerging markets, this research area is still neglected. To the best of the authors' knowledge, there are no empirical contributions on the consumers' perceptions of brand extensions from European emerging and developing markets, i.e., from Central and Eastern Europe, South Eastern Europe and/or Baltic countries. Therefore, researching consumers' attitudes toward the brand extensions in the emerging markets within European context presents an actual and timely topic which has not been sufficiently researched so far.
Furthermore, Bottomley & Holden (2001) warn the professionals about the importance of familiarizing global brands' managers with different approaches that customers may have in evaluating the same brand extensions in different markets, and they conclude that (2001, p. 499) "…the impact of cross-cultural influences on consumer evaluations of brand extensions cannot be entirely ruled out. " In addition, according to Park et al. (1991), brand extensions are more demanding for function-based rather than the prestige-based brands. Finally, data (Nielsen's global research, 2015) show that there are differences in brand extension acceptance between emerging and developed markets -e.g., 22% of consumers from emerging markets would purchase a new product because it is related to the brand they like, versus 17% of consumers from developed markets. In line with the aforementioned, researching consumers' perceptions of and attitudes towards the parent brand and its extension(s) in emerging markets, especially when it comes to FMCG brands, is very important and must not be ignored.
Like other emerging European markets, the Croatian market began to face strong competition from companies with well-known international and global brands in the 1990s, which has been accompanied by numerous and rapid changes in value system, attitudes and behavior of Croatian consumers (Ozretic-Dosen, 2005). The use of the brand extension strategy in Croatia is growing continuously, but little is known about Croatian consumers' reactions to and attitudes toward the brand extensions regardless the global or local character of the parent brand, and regardless the parent brand type (function or prestige-based). In Croatian economy, imports in 2017 are estimated at 21.2 billion USD (CIA, The World Factbook, 2018), and foodstuffs are among the five most imported products. Global FMCG (parent brands and their extensions) have been present on the Croatian market for decades, and are constantly trying to keep, even strengthen their leading positions in many categories. Therefore, the Croatian market can serve as an appropriate context for the exploration of brand extensions in emerging European markets.
The aim of the paper is twofold. The theoretical part gives a summary of the specific characteristics of brand extension strategy as a usual marketing practice of companies with strong and well-positioned brands. It is followed by detailed results of primary research, i.e., of preliminary, exploratory efforts to examine and understand Croatian consumers' perceptions of and attitudes toward a global parent brand (Coca-Cola) and its extension strategy on the Croatian market. Research results are summarized and analyzed; the effects of certain variables on consumers' perceptions of brand extension are discussed. Conclusions, theoretical and practical implications, research limitations and possible future research streams are presented in the last section.

Characteristics of the brand extension strategy
The company that decides to use the brand extension strategy must be a good judge of its decision. It must thoroughly investigate whether there is a positive opinion among consumers about the existing, core, primary or parent brand, and in which area to set its extension. The parent brand plays an important, a pivotal role in the brand extension strategy. It provides security to the success of expansion, as well as the greater probability of trial because consumers of the parent brand will probably try the brand extension. However, it does not guarantee the success of a brand extension. Evidence shows that a large number of extensions fail -even seventy percent of them do not achieve the expected success, with the most common argument and explanation how the extension is too much away from the main, parent brand (Kotler & Keller, 2008). When managed well, brand extension represents a new source of revenue and improves the image and meaning of the parent brand (Shahrokh et al., 2012).
Previous research (e.g., Hem et al., 2003;Meyvis & Janiszewski, 2004) confirmed that the greater similarity between the parent brand and its extension leads to the larger transfer of positive (or negative) effects on the brand extension. Powerful brands provide greater benefits for extensions in comparison with the weak brands. When the compatibility between the parent brand and its extension is small, the consequences for the brand image are more negative. If the compatibility is minimal, the extension should be seen as a product subcategory which is distant from the parent brand, so that the effect of dilution of the associations will be smaller than expected, or it will not exist at all (Martinez & Pina, 2003). Consumers significantly affect the success of the strategy of brand extension. Their positive assessments of parent brand and its extensions lead to the development of brand equity (Grime et al., 2002). Different consumers think differently about brand extension, i.e., their methods of processing information vary depending on their familiarity with the parent brand. In the case of a low familiarity with the brand, information processing occurs at the level of a product category, while in the case of a high level of familiarity with the brand, information processing is related to the characteristics of the product itself (Thorbjørnsen, 2005). Consumers' attitudes about the parent brand and its extension are influenced by marketing communications, too. According to Kim (2003) and Martinez & de Chernatony (2004), the emphasis of communication campaigns should be more on the parent brand name than on the characteristics of the extension itself, in order to ensure consumers' positive perceptions about the extension.
The company may follow the steps for the successful introduction of brand extension and pay attention to the components of success, but success is never guaranteed. Seltene & Brunel (2008) argue that the brand extension's success depends on two components: the distance between the brand extension and the parent brand regarding consumption and the distance that separates the brand extension of the parent brand regarding brand associations. Five strategic components for the brand extension's success, according to Athanasopoulou et al. (2015), are quality of distribution strategy, quality of positioning, quality of product development strategy, the extent of promotional investment and extent of market research. In international marketing, brand extension's success is strongly influenced by the brand origin -extension fit, while perceived brand globalness and brand origin image affect brand extension's success to a lesser extent (Sichtmann & Diamantopoulos, 2013).

Advantages and disadvantages of the strategy of brand extension
The knowledge on how to expand the brand helps to maximize the advantages and minimize the disadvantages of the strategy of brand extension. The success (or failure) depends on market research conducted by the company, nature of the extension itself and consumers' attitudes towards extension (Völkner & Sattler, 2006;d' Astous et al., 2007). An advantage may easily turn into a disadvantage because much depends on consumers' perceptions about the parent brand and its extension.
The brand extension strategy may bring many advantages to a company. Aaker (1990) in his seminal article stated that consistent brand extensions could reinforce the brand image, which results in the higher visibility of brands. Well-known parent brand eliminates risk and increases the chances for the product's trial. When the well-known parent brand expands in the product category, that is considered risky, it can serve as an important tool to minimize or eliminate the risk, to signal an acceptable level of quality and therefore to increase the likelihood of extension's acceptance by consumers (Hem et al., 2003). Also, brand extensions of well-known brands can have easier access to distribution channels. The successful brand extension may increase the brand equity, and contribute to the development of meaning and importance of brand to consumers (Spiggle et al., 2012).
Although different gains of exploitation of the parent brand seem attractive, the percentage of failure of brand extensions is high and may be more than 80% (Shahrokh et al., 2012). Precisely because of the high rate of failure, it is necessary to study thoroughly possible disadvantages related to the brand extension strategy. According to Sharp (1993, p. 12), disadvantages are: 1. Spillover: brand extension exposes a greater range of brands to the possible spillover of negative publicity. For example, Balachander & Ghose (2003, p. 5) point out that "if a low-quality product is offered with an umbrella brand name, it leads consumers to conclude that all other products with the same brand name are also of low quality. " 2. Cannibalization: the brand extension, when set too close to the parent brand, could cannibalize existing product(s) of the parent brand. According to Taylor (2004, p. 25) "this is the risk of an extension eating up other family members. The biggest risk occurs with range extensions that are 'brand clones' lacking differentiation. " If the brand extension fails, this will have a significant impact on the image, sales, and consumers' perceptions about the parent brand. One of the major negative impacts of unsuccessful brand extension is a dilution of the parent brand's image (Martinez & de Chernatony, 2004). Dilution appears in the cases where the parent brand name is no longer associated with a specific product or some similar products, so it fades in the minds of consumers (Kotler & Keller, 2008). The strength of the dilution depends on how the existing consumers' beliefs change under the influence of the information about the brand extension which is not consistent with the perceptions of parent brand (Loken & John, 1993).
Brand managers use different approaches to minimize risk and maximize potential rewards of brand extension strategy. They may perform brand extension pre-testing in a laboratory environment in order to avoid possible undesirable, negative effects. If such pre-testing is not possible, then it is advisable to introduce extension on the market where the familiarity with the parent brand is weak, because the negative feedback effects, if they happen, will be less pronounced. If it turns out that the brand extension strategy is successful in the markets of low familiarity with the parent brand, it is almost certain that it will be successful and provide positive feedback effects on markets with high brand awareness and strong image of a parent brand (Thorbjørnsen, 2005).

Types of brand extension strategy
A company should carefully choose the type of brand extension strategy which will enable maximization of advantages and minimization of disadvantages. Brand extensions can be broadly classified into one of two types -horizontal and vertical brand extensions (Sheinin & Schmitt, 1994 cited in Chen & Liu, 2004). Another possible categorization is line extensions -where a new product is offered within an existing brand category; and category extension where a new product is introduced in a different, new product category (Keller, 1998, cited in Kim, 2003. In the vertical brand extension, the new brand is in the same product category as the parent brand but differs in price and quality (Chen & Liu, 2004). Research evidence shows that vertical step-up extensions increase price and brand image of the parent brand, while vertical extensions step-down have reducing effects on price and image of the parent brand (Chen & Yang, 2013). Consumers may see the significant difference in price and quality as inconsistency, which may lead them to re-evaluate their initial opinion on the price and quality of parent brand (Kim et al., 2001). The second or descriptive brand name is usually added to the parent brand name, to reflect the link between extension and the parent brand (Chen & Liu, 2004), aiming to create a connection in the minds of consumers who buy the parent brand and to motivate them to buy the brand extension.
When introducing the horizontal brand extension, the existing brand name is given to new products in the related or the new category of products (Chen & Liu, 2004). The company may use the strategy of horizontal brand extension to position new products as close as possible to the parent brand, taking the advantages of its brand equity. The horizontal brand extension does not provoke consumers' negative attitudes because this type of extension means the same level of prestige, status or quality as the parent brand (Kim et al., 2001). Consumers do not have to re-evaluate the parent brand. There are two types of horizontal brand extension: line extensions and franchise extensions. Line extensions involve "a current brand name to enter a new market segment in its product class" (Pitta & Katsanis, 1995, p. 60). However, if extension develops too far, the brand can lose its meaning in the minds of consumers, who become confused. In the franchise extension strategy, the company uses the well-known brand name for the products in the category that is new for the company (Tauber, 1981). Doing so, the company can minimize the costs and maximize the sales of the parent brand due to almost the same brand name. Minimum investment in marketing communication is required because consumers are already familiar with the brand. Through franchise extension strategy, the company may take advantage of all the positive aspects of the parent brand and increase its brand equity. Successful product extension may reap benefits, but management should not forget the risk of extension's failure (Pitta & Katsanis, 1995).

Research Hypotheses
The purpose of the primary research is to explore and understand Croatian consumers' perceptions about the global parent brand (Coca-Cola) and their attitudes towards its brand extensions.

Hypotheses
The reviewed literature suggests that consumers' perceptions of the parent brand and its extensions may boost the company for further expansion or may influence the company's decision to stop using this strategy. Understanding consumers' perceptions about the parent brand and how the associations about the parent brand are transferred to its extensions is crucial to the success of brand extensions.
Accurate tools are needed to determine which brand extensions are consistent with the parent brand (Viot, 2011). Brand affect is often operationalized as consumer's overall favorable or unfavorable evaluation of the brand (Keller, 1993). Furthermore, brand affect concept, as defined by Chaudhuri & Holbrook (2001), refers to "brand's potential to elicit a positive emotional response in the average consumer as a result of its use" (Chaudhuri & Holbrook, 2001, p. 82). As stated by Taylor (2004), a strong parent brand creates brand affect and brand awareness more easily; therefore an extension will benefit from the familiarity and confidence related to the well-known parent brand in the process of value creation. Since every company wants its brand and brand extension to be accepted by the target market, it is necessary to develop brand trust. Consumers will support brand extension upon having brand trust in the parent brand. Furthermore, gaining consumer confidence in the brand, as well as his/her loyalty to the brand is very important for the positive assessment of brand extension. Loyalty to the brand is formed from the positive perceptions and feelings towards the brand (Shahrokh et al., 2012). Chaudhuri & Holbrook (2001, p. 83) suggest that "brand trust and brand affect are each related to both purchase and attitudinal loyalty. " On the basis of these observations, we posit that: According to Reast (2005), trust is a part of the brand's credibility in the process of the acceptance of its extension, it is crucial for the development of loyalty, critical for maintaining a successful relationship between the company and its customers, and influences brand equity. Therefore, we propose the following research hypothesis:

H2: The relationship between the brand affect (BA) and the brand loyalty (BL) is mediated by the brand trust (BT).
Delgado-Ballester & Manuera-Aleman (2001) were among the first who empirically tested and analyzed the relationship between concepts of brand trust (BT), satisfaction and brand loyalty (BL), and have found that brand trust (BT) is a key variable for generating consumers' brand commitment and subsequent future intentions. According to Matzler et al. (2008), brand loyalty (BL) is an important consequence of brand trust. In line with the research reviewed here, we propose that:

H3: The relationship between the brand trust (BT) and the consumers' attitude towards brand extension (BE) is mediated by the brand loyalty (BL).
Variables brand affect (BA), brand trust (BT) and brand loyalty (BL) represent consumers' perceptions of the parent brand (Coca-Cola), while brand extension (BE) refers to the consumers' perceptions of the brand extensions of Coca-Cola.

Data collection and research instrument
In order to collect data, a convenience sample was used. The research respondents were students of business at the Faculty of Economics & Business, University of Zagreb in Croatia. Student sample was chosen because of the accessibility, convenience, and low cost (Hakala et al., 2012). Furthermore, university students, in general, are frequent users of Coca-Cola and its extensions. They represent a population of interest for this particular research; therefore, as suggested by Peterson & Merunka (2014), they are the appropriate research subjects. Questionnaire data were collected from 242 respondents.
"Considered by many to be the quintessential international brand, Coca-Cola transcends ethnicity, gender, age, education and social class." (Slater, 2000, p. 202). The decision to focus on a soft drink as a specific category of FMCG, specifically on Coca-Cola brand as appropriate for this research was made after reviewing publications dedicated to brand management and brand extensions in refereed scientific marketing journals, where a significant number of papers reported about examining different aspects of brand management using Coca-Cola brand. Since the research aimed to bridge the gap of deficiency in contextual contribution (European emerging markets), Coca-Cola was chosen as an adequate brand.
A self-reported structured questionnaire was used. The questionnaire consisted of 34 questions. Eight questions were dedicated to collecting data about the status of the user of Coca-Cola, the frequency of its use, the associations which respondents have about Coca-Cola brand, the awareness of the latest Coca-Cola extension (Coca-Cola Cherry) and demographics. All the remaining questions (26) were the statements regarding variables brand affect (BA), brand trust (BT), brand loyalty (BL) and brand extension (BE), and were measured by using a five-point Likert-type scale (indicating 1 = strongly agree to 5 =strongly disagree) for the level of the agreement with a particular statement. The scales used to measure the variables brand affect (BA), brand trust (BT), brand loyalty (BL) and brand extension (BE) were already used and tested in the marketing literature. Brand affect (BA) was measured through the 3-item scale employed in research of Joji (2011), where it showed good reliability (Cronbach alpha coefficient 0.889). Brand trust (BT) was assessed through a 9-item scale; the items were extracted from the previous research of Delgado-Ballester (2004) and Joji (2011), adapted and combined. The original scales of Delgado-Ballester (2004) and Joji (2011) for brand trust (BT) showed Cronbach alpha coefficient greater than 0.7 and 0.886, respectively. Measurement of brand loyalty (BL) was performed by a combined 6-item scale, consisting of 4 items from Joji's research (2011) and 2 items from Brand Loyalty (Soft Drink) Scale in Bruner & Hansel (1992); with the Cronbach alpha coefficients 0.755 and 0.75, respectively. Finally, brand extension (BE) was assessed through an 8-item scale, 3 of which were adapted and composed of Brand Extensions Fit scales proposed in Bruner et al. (2005), with original reliability coefficients alpha of 0.90 and 0.95, three were adapted form Joji (2011), and the two remaining dimensions were proposed by the authors. The questionnaire was pre-tested on fifteen respondents, who were not included in the main research. Table 1 shows scale dimensions and items as well as their respective source.
Of the 218 respondents, 26, or 11.93%, stated that they do not drink Coca-Cola. They were asked to skip other questions and answer only those related to demographic characteristics. The remaining 192 respondents (88.07%) filled in the questionnaire completely and their answers were taken for further testing and analysis. Most of them, 82 respondents, or 42.71%, consume Coca-Cola products only on occasion, followed by 71 respondents, or 36.98%, who drink Coca-Cola once a week. The smallest number of respondents, 7, or 3.65%, drink Coca-Cola every day. Interestingly, respondents reported 272 associations about Coca-Cola. Most of them mentioned the associations connected with a special shape of the Coca-Cola bottle, the main characteristics regarding the taste of the beverage, food related to Coca-Cola consumption, followed by the various associations of Christmas holidays. A high number of respondents, 181 (or 94.27%), heard of Coca-Cola Cherry, which was the latest (introduced in 2013) brand extension of Coca-Cola on the Croatian market, while the remaining 11, or 5.73%, never heard of this brand extension. Of those who heard of Coca-Cola Cherry, the majority learned about the extension through advertising (75 respondents, or 39.06%), followed by WOM -i.e., friends as a source of information (57 respondents, or 29.69%). However, only 102, or 53.13% of all the respondents, tried the extended product, while 90, or 46.88%, did not. Of these 90 respondents, 32, or 16.67% of all the respondents (or 17.68% of those respondents who heard of Coca-Cola Cherry), do not even want to try a brand extension. The remaining 58, or 30.21% of all the respondents (or 32.04% of those respondents who heard about the extension), had not tried it but wanted to.
A factor analysis of 26 items, grouped into four variables (BA -3 items, BT -9 items, BL -6 items and BE -8 items) was conducted. Kaiser-Meyer-Olkin measure of sampling adequacy was used for measuring the appropriateness of applying factor analysis. Also, Bartlett's Test of Sphericity was used to test the assumption of homogeneity of variance. Table 3 shows the data for both tests. The KMO index of .886 and p-value  of Bartlett's test of .000 confirm the appropriateness of using factor analysis to test the collected data. Regarding the reliability statistics of the scales used, Cronbach alpha coefficients are greater than 0.7 for all the variables, i.e., they are: 0.908 for BA, 0.879 for BT, 0.893 for BL and 0.769 for BE. Data presented in Table 4 show the average agreement with the statements in the questionnaire. There is a certain level of brand trust BT (mean 2.43), which shows that respondents, on average, agree with statements that describe the confidence in the Coca-Cola brand. Brand affect BA also reveals the agreement with the statements, depicting positive feelings of respondents when using Coca-Cola (mean 2.78). Likewise, brand loyalty BL shows agreement with the statements (mean 2.79), thus showing that respondents, on average, are loyal to Coca-Cola. The mean for the statements indicating the brand extension BE is 2.90, showing the reluctance regarding those statements, i.e., respondents are unsure about the brand extensions of Coca-Cola.  Table 5 shows the correlation coefficients for the relationships between individual variables: brand trust BT, brand affect BA, brand loyalty BL and brand extension BE. They reveal whether there is a correlation between variables and how strong the relationship is. As presented in Table 5, all correlation coefficient values for variables BT, BA, BL and BE are greater than zero (meaning that connection between the variables exists) and are less than one (meaning the connection is not complete but has a positive direction). Table 6 presents the results of the regression analysis for variables: brand trust BT, brand affect BA, and brand loyalty BL, conducted for the purpose of testing hypotheses H1a, H1b, and H1c. Results show that brand affect (BA) and brand trust (BT) positively influence consumers' attitude towards brand extension (BE), which allows confirming hypotheses H1a and H1b. Brand affect (BA), with the unstandardized β coefficient (.430), has a positive and greastest influence on the perceptions of the Coca-Cola brand extension. Consumers have positive feelings towards Coca-Cola, because it is a well-known global brand, with long presence and tradition on the Croatian market. The unstandardized β coefficient for the variable brand trust BT is low, (.224). Nevertheless, this variable has positive, very little effect on the consumers' perceptions about the brand extension BE. H1c is rejected; the unstandardized β coefficient for brand loyalty BL (-.159), with the significance .110, shows that brand loyalty BL diminishes consumers' perception on brand extension and as such does not influence positively the consumers' attitude towards brand extension.  (IBM SPSS,version 19) Regression analysis was also used for the purpose of testing hypothesis H2. Since this hypothesis aims to find out the influence, i.e. significance of the mediation effect of the variable brand trust BT in the relationship between the variables brand affect BA and brand loyalty BL, the Sobel test was used. In order to get the necessary information to implement the Sobel test, two regression analyses (depicted in Table 7 and Table 8) were conducted. Data on the relationship between the variables brand affect BA and brand trust BT, and data on the relationship of the variables brand loyalty BL and brand affect BA with the variable brand trust BT, which has the intermediary role in hypothesis H2, were obtained. The Sobel test was conducted, the level of significance was α = 0.05, and the range of acceptance of the hypothesis was: │z│< z α/2 , │z│< z 0,025 , │z│< 1.96.

Testing hypotheses and discussion
The Sobel test results for hypothesis H2 are: z = 5.723, standard error = 0.785 and p-value = 0.000. Since "z" is greater than 1.96, hypothesis H2 is rejected. The relationship between the variables brand affect BA and brand loyalty BL is strong (.613), but according to the results obtained, brand trust BT does not have influence (has no mediation effect) on this relationship.   (IBM SPSS,version 19) Hypothesis H3 testing was done through regression analysis and the Sobel test, too. Data were obtained by regression analysis of the variables brand loyalty BL and brand trust BT (Table 9), and regression analysis of variables brand extension BE, brand trust BT and brand loyalty BL (Table 10). Sobel test's values for hypothesis H3 are: z = -0.495, standard error = 0.263 and p-value = 0.620. Hypothesis H3 is confirmed ("z" value is lower than 1.96). The relationship between the variables brand trust BT and brand extension BE, which according to the correlation coefficient is slightly lower (0.285) than other correlations, is affected, i.e., is mediated by brand loyalty BL.   (IBM SPSS,version 19) Research results of this study point to the conclusion that, for Croatian consumers, brand affect (BA) may be taken into consideration as a primary variable that positively influences their attitudes towards brand extension. Brand trust (BT) does not mediate the relationship between brand affect (BA) and brand loyalty (BL). However, brand loyalty (BL) mediates the relationship between brand trust (BT) and consumers' attitude towards brand extension (BE).
As such, research results confirm that managers should introduce brand extension only when they are sure of the position of the parent brand and its previous extensions in the consumers' mind, corroborating the findings of Joji (2011). Furthermore, results are in line with the research findings of Chaudhuri & Holbrook (2002), showing how it is necessary to know the existing level of brand affect and brand trust since they have a different influence on brand-customer relationships. The research reveals that brand loyalty has an important role in the determination of consumers' attitude towards brand extension (BE), which is in line with Chaudhuri & Holbrook (2001), who found the potential importance of brand loyalty as a link in the determination of brand performance outcomes. Also, results support to a certain degree the conclusion of Joshi and Yadav (2017), implying that brand extension strategy should help to strengthen the parent brand.

Conclusion
Brand managers invest many efforts in positioning their brand extensions on the market. In spite of different risks that might occur when the company introduces the extension, there are also many possible advantages related to the brand extension strategy. Marketing research should help minimize the risk or disadvantages and maximize the benefits or advantages of brand extension strategy.
This research extends the examination of variables affecting brand extensions in emerging markets and confirms previous findings on how consumers significantly influence whether a brand and its extension(s) will be successful in the future or will collapse. Since brand extension's destiny depends heavily on consumers' perceptions about the parent brand, it is necessary to conduct continuous research about con-sumers' perceptions and explore how certain perceptions affect the company's business and how they are interconnected.

Theoretical implications
More than 80% of the world's consumers live in emerging markets and transitional economies (Steenkamp & Burgess, 2002). Emerging markets are rather heterogeneous, and their classification, which is based dominantly on economic factors, does not reveal sufficiently the differences that exist among them (Schlager & Maas, 2013). Burgess and Steenkamp (2006) call for more academic research regarding theory development, acquisition of meaningful data, analysis of the data to test one's theories and learning, through which research in emerging markets can contribute to the advancement of marketing science and practice.
Therefore, in terms of theoretical implications, this research adds to the existing knowledge about brand extension strategy with insights and results from the new research context of the emerging European market (Croatia), where companies (global, international and local), especially from different FMCG industries, often use the brand extension strategy. As such, and in spite of the fact that empirical results confirm the outcomes that exist in studies conducted previously in developed markets, this research contributes to enriching the knowledge from the wider area of consumer behavior in emerging markets.

Managerial implications
In terms of managerial implications, research findings might be useful for marketing professionals either being in the field of the strategic brand management on the emerging markets already or planning to internationalize their future business activities in those markets.
The results also imply that it is of particular importance to find out the perceptions of consumers about the parent brand in order to implement the new brand extension successfully or improve its performance if it is already on the market. This kind of market research helps the company to find out both its advantages and drawbacks on the market, and to plan its future activities strategically. Knowledge about the parent brand and its extension(s) is valuable for making decisions about, e.g., desirable integrated marketing communications, marketing channels choice, etc. Finally, the results suggest that brand managers operating on the Croatian market need to devote their resources to developing and maintaining brand affect since brand affect was found to be the most important variable that positively influences consumers' attitudes towards brand extension(s). To achieve this goal, brand managers have to constantly search for innovative ways to attract the attention of the target segments by using traditional and new media in the frame of the integrated marketing communications.

Research limitations and future research
Several research limitations should be taken into account. The first is related to the fact that the convenience sample was used (with predominantly female respondents), i.e., the research respondents were only students of one university institution. University students' perceptions cannot be considered as the general ones of all consumers of Coca-Cola and its extensions on the Croatian market. Also, it cannot be said with certainty that all the respondents understood the questions completely and replied honestly and truthfully. Moreover, the research is limited to only one brand, and it cannot be argued that the results would be the same when exploring some other brand(s) of soft drinks or FMCG.
Due to the exploratory nature of this research, future studies should re-evaluate the strength of these findings. Also, future studies should approach a greater number of respondents (non-student samples) and brands (and other products categories, too) to obtain results that are more representative. The replication of the research with other emerging (European or other) markets and their consumer cultures may reveal new insights, too. Other field research methods, such as interviews and/or experiments, would be appropriate to provide further explanations of the topic. All this will enable a better understanding of brand extension strategy and explain its role in different industries on the emerging, developing markets.