ISSN: 1550-7521
Abdulla Al-Towfiq Hasan* and Mahamudul Hasan
Department of Marketing, Faculty of Business Administration and management, Patuakhali Science and Technology University, Bangladesh
Received date: Janaury 18, 2019; Accepted date: Janaury 26, 2019; Published date: Janaury 31, 2019
Citation: Hasan AAT, Hasan M. Secondary Brand Association Influences on Brand Preference and Purchase Intention. Global Media Journal 2019, 17:32.
Brand associations; Brand preference; Purchase intention; Bangladesh
Visit for more related articles at Global Media Journal
Everything is being branded in 21st centaury with a view to positioning in customer’s mind. The main objective of this research is to analyze the sources of secondary brand association impacting on brand preference and purchase intention. The research was conducted through, simple random sampling; structured questionnaires distributed among 250 respondents of different brands users in Bangladesh. The research area was Khulna division (Khulna University, Khulna medical college, Khulna govt. B. L. College) with five (05) months duration. The validity and reliability of questionnaires were checked by composite reliability, Cronbachs Alpha, average variance extracted. Subsequently, Pearson correlation and multiple regressions were analyzed. The results of the study showed that company, and country of origin had significant positive relationship and positive impact; channel of distribution, licensing, celebrity endorsement, and third party sources had insignificant positive impact; co-branding, sporting-cultural and other events had insignificant negative impact on brand preference and purchase intention.
Brand, is not a new term in 21st century, is the set of signs, symbols, logos, colors, design, shapes, model, version etc Keller [1]. Desai and Keller [2] found that modern marketing concept emphasizes on brand because, attractive and convincing brand can maximize company profits and image, make long term relationship with customers; and consumers become aware of brand rather than itself called Secondary brand association. The researcher is going to identify secondary brand association sources which influence the consumer’s preference and purchase intention most by proposing and testing the hypothesis; developing a structured questionnaire which is checked by pilot study, expert opinion, KMO tests, composite reliability, and average variance extracted; and subsequently analyze the Pearson correlation between independent variables and the dependent variable, multiple regression analysis to calculate the impact of independent variables on the dependent variable.
Company
Aaker [3] found that Company is the separate entity which represents the product owner, authority, and liabilities of the products/brands. Company plays a role of family brand or corporate brand to introduce sub-brand into new markets or in existing market [4]. The main task of a company is to create a new brand, adopt or modify an existing brand, and combine an existing and a new brand [5]. This generates secondary brand associations.
Country of origin
Yasin [6] identified that country of origin indicates the geographic location from which a product is originated. Some countries may have reputation for expertise of some specific product categories which can generate secondary brand associations [7]. Keller [1] suggested that consumers prefer products from expertise countries which can be trusted. For example, Germany, France, Japan, England, and Switzerland, are considered expertise country for BMW, Chanel perfume, Kikkoman soy sauce, Cadbury, and Mont blanc pens [1].
Channel of distribution
Kumar [8] found that retailers have a strong relationship with consumers. So, consumers have a strong linked with retailers [8]. This indicates that channel of distribution via retailers, wholesalers, agents, affects consumer’s mind as a secondary brand association [1]. A consumer can choose a product on the basis of where it is sold. Retailers or some other distributors signify the quality of product, because consumers perceive same brand differently depend ending on whether it is sold in a store seen as prestigious and in a foot path seen as a low quality [9].
Co-branding
Blackett and Boad [10] argued that co-branding is a brand extension strategy by which an umbrella brand creates a set of sub brands in the family, a wing of brand alliance where, two or more established brands are combined together with a view to producing and marketing products in same fashion. When, new entrant brands make partnership or alliance with established renowned brand then, new entrant brands get opportunities to become popular with the established renowned brands [2].
Licensing
Licensing is the mechanism of making contractual agreements with firms to use logos, names, characters to market own products or brands with fixed price [11]. Licensing is, the short cut way of building brand equity, easy of building popularity, legal certification of building brand awareness, the way of renting another brand to contribute on brand equity of its own product [12].
Celebrity endorsement
Choi and Rifon [13] found that celebrity endorsement is the mechanism of using well known personalities in advertisements to promote products among the consumers. Celebrity endorsement, increases brand image, brand awareness, and brand response; indicates that some celebrities or trustworthy personalities talk about the brand so that fans of those celebrities are influenced to purchase and consume the brand [14].
Sporting, cultural, or other events
Keller [1] found that sporting, cultural, or other events are the set of combination of different events which are sponsored by a firm with a view to promoting of its brands among the consumers. Chalip at el (2003) argued that sponsorship contributes to brand associations improving brand awareness, image development, and improving the brand strength. Consumers get secondary brand knowledge from favorite sporting, cultural or other events what they like most [15].
Third-party sources
Kotler [16] found that third-party sources are the supporting wing which links brand knowledge with consumers and mechanism of getting functions operated by other than the firm. Third-party sources are the strong secondary brand association wing because it indicates the quality of products, popularity of products, awareness of products, and credibility of products [9].
Brand Preference and Purchase Intention
Fishbein and Ajzen [17] identified that brand preference states a situation where, consumers are determined about the quality and performance of the brand. It indicates the strength of consumer’s assumption under which consumers try to satisfy them through the brand [18]. Kotler and Keller [19] indicated that Purchase intention is the mechanism of willingness of consumers to buy certain products/services. Purchase intention varies according to stimulus, outcome expectation, aspirational value, recommendation, and emotional association [20].
The study is descriptive in nature through proposing, and testing the hypothesis. The research was conducted by developing structured five point Likert scale questionnaires for primary data collection where, (1) indicates strongly disagree and (5) indicates strongly agree. Secondary data was collected from websites, newspapers, books, and journals. Demographic data was collected in terms of gender, region, age, occupation, marital status, and income. The research duration was five (5) months from July 1, 2018 to December 1, 2018. The primary data was collected from Khulna division (Khulna University, Khulna medical college, Khulna govt. B. L. College). The sampling frame was the list of students of honors 1st year to 4th year and master’s of different brand users. Simple random sampling was the sampling technique and sample size was 250. Systematically thirty (30) respondents of different brand users had been selected and the pilot study had been conducted on them in order to detect any ambiguities or questions that were not easily understood and checked by expert opinion before the study was administered. The researcher conducted Kaiser-Meyer-Olkin (KMO) sample adequacy test and Bartlett’s sphericity tests to identify whether it is suitable for factor analysis or not. When, the value of KMO is greater than 0.5 and “p” value is less than 0.05 in Barlett’s test, it implies that the data is adequate for factor analysis [21].
Conceptual framework
Secondary brand association (Figure 1).
Statement of hypothesis
H1: The Company has a significant positive effect on Brand preference and purchase Intention.
H2: Country of origin has a significant positive effect on Brand preference and purchase Intention.
H3: Channel of distribution has a significant positive effect on Brand preference and purchase Intention.
H4: Co-branding has a significant positive effect on Brand preference and purchase Intention.
H5: Licensing has a significant positive effect on Brand preference and purchase Intention.
H6: Celebrity endorsement has a significant positive effect on Brand preference and purchase Intention.
H7: Sporting, cultural or other events has a significant positive effect on Brand preference and purchase Intention.
H8: Third-party sources have a significant positive effect on Brand preference and purchase Intention.
65% were male and 35% were female among the 250 respondents. Majority of respondents (84%) were unmarried and were aged below 30 years and undergraduate (79%) having income 5000 taka to 15000 taka. 100% of the respondents were students.
Table 1 shows that Kaiser-Meyer-Olkin Measure of Sampling Adequacy is 0.696 and significance value is 0.000. This indicates the data set is adequate to factor analysis because KMO value is greater than 0.5 and “p” value is less than 0.05 in KMO and Barlett’s test [22].
Table 1: KMO and Bartlett’s Test.
KMO and Bartlett's Test | ||
---|---|---|
Kaiser-Meyer-Olkin Measure of Sampling Adequacy | 0.696 | |
Bartlett's Test of Sphericity | Approx. Chi-Square | 3389.844 |
df | 741 | |
Sig. | 0 |
Table 2 shows that the composite reliability of all variables is greater than 0.8. That indicates that the construct of questionnaire is valid [23]. Average variance extracted of all variables is greater than 0.40. This indicates that the statements are reliable and valid for conducting further research (Clauser, 2006). Cronbach’s Alpha of all the variables is greater than 0.60 which indicates that all items of a particular variable are internally consistent [24].
Table 2: Reliability and factor analysis.
Factors | Items | Cronbach’s Alpha | Composite Reliability | Average variance Extracted |
---|---|---|---|---|
Company | 5 | 0.684 | 0.897583 | 0.636929 |
Country of origin | 5 | 0.785 | 0.847279 | 0.526016 |
Channel of distribution | 4 | 0.812 | 0.867455 | 0.622054 |
Co-branding | 4 | 0.809 | 0.862965 | 0.61282 |
Licensing | 3 | 0.807 | 0.875578 | 0.701537 |
Celebrity endorsement | 5 | 0.766 | 0.832267 | 0.499327 |
Sporting, cultural and other events | 4 | 0.713 | 0.817188 | 0.529126 |
Third-party sources | 4 | 0.704 | 0.809759 | 0.518117 |
Brand preference and purchase intention | 5 | 0.837 | 0.878966 | 0.59279 |
The Table 3 shows that company has a strong association and positive significant correlation with Brand preference and purchase intention. The correlation coefficient is r=0.209, p=0.001<0.05.
Table 3: Correlation of coefficients.
Correlations | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|
Company | Country Of origin | Channel of distribution | Co-branding | Licensing | Celebrity endorsement | Sporting cultural and other events | Third party sources | Brand preference and purchase intention | ||
Company | Pearson Correlation | 1 | 0.085 | 0.109 | 0.094 | 0.116 | .200** | 0.038 | -0.002 | .209** |
Sig. (2-tailed) | 0.181 | 0.087 | 0.14 | 0.067 | 0.001 | 0.554 | 0.981 | 0.001 | ||
N | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | |
Country of origin | Pearson Correlation | 0.085 | 1 | -0.051 | .130* | -0.014 | -0.065 | -0.016 | 0.103 | .198** |
Sig. (2-tailed) | 0.181 | 0.421 | 0.04 | 0.825 | 0.31 | 0.798 | 0.104 | 0.002 | ||
N | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | |
Channel of distribution | Pearson Correlation | 0.109 | -0.051 | 1 | 0.121 | 0.098 | -0.025 | 0.04 | 0.061 | 0.116 |
Sig. (2-tailed) | 0.087 | 0.421 | 0.056 | 0.121 | 0.695 | 0.531 | 0.335 | 0.067 | ||
N | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | |
Co-branding | Pearson Correlation | 0.094 | .130* | 0.121 | 1 | 0.071 | .162* | 0.05 | -0.044 | 0.014 |
Sig. (2-tailed) | 0.14 | 0.04 | 0.056 | 0.263 | 0.01 | 0.431 | 0.492 | 0.824 | ||
N | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | |
Licensing | Pearson Correlation | 0.116 | -0.014 | 0.098 | 0.071 | 1 | 0.122 | 0.033 | 0.091 | 0.095 |
Sig. (2-tailed) | 0.067 | 0.825 | 0.121 | 0.263 | 0.054 | 0.6 | 0.152 | 0.134 | ||
N | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | |
Celebrity endorsement | Pearson Correlation | .200** | -0.065 | -0.025 | .162* | 0.122 | 1 | .134* | -.135* | 0.061 |
Sig. (2-tailed) | 0.001 | 0.31 | 0.695 | 0.01 | 0.054 | 0.034 | 0.032 | 0.337 | ||
N | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | |
Sporting cultural and other events | Pearson Correlation | 0.038 | -0.016 | 0.04 | 0.05 | 0.033 | .134* | 1 | 0.015 | 0.009 |
Sig. (2-tailed) | 0.554 | 0.798 | 0.531 | 0.431 | 0.6 | 0.034 | 0.811 | 0.889 | ||
N | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | |
Third party sources | Pearson Correlation | -0.002 | 0.103 | 0.061 | -0.044 | 0.091 | -.135* | 0.015 | 1 | 0.063 |
Sig. (2-tailed) | 0.981 | 0.104 | 0.335 | 0.492 | 0.152 | 0.032 | 0.811 | 0.318 | ||
N | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | |
Brand preference and purchase intention | Pearson Correlation | .209** | .198** | 0.116 | 0.014 | 0.095 | 0.061 | 0.009 | 0.063 | 1 |
Sig. (2-tailed) | 0.001 | 0.002 | 0.067 | 0.824 | 0.134 | 0.337 | 0.889 | 0.318 | ||
N | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 | 250 |
**Correlation is significant at the 0.01 level (2-tailed).
*Correlation is significant at the 0.05 level (2-tailed).
Country of origin has a strong association and positive significant correlation with Brand preference and purchase intention. The correlation coefficient is r=0.198, p=0.002<0.05.
Channel of distribution has a positive insignificant correlation with Brand preference and purchase intention. The correlation coefficient is r=0.116, p=0.067>0.05.
Co-branding has a weak association and positive insignificant correlation with Brand preference and purchase intention. The correlation coefficient is r=0.014, p=0.824>0.05.
Licensing has a positive insignificant correlation with Brand preference and purchase intention. The correlation coefficient is r=0.095, p=0.134>0.05.
Celebrity endorsement has a weak association and positive insignificant correlation with Brand preference and purchase intention. The correlation coefficient is r=0.061, p=0.337>0.05.
Sporting, cultural, and other events have a weak association and positive insignificant correlation with Brand preference and purchase intention. The correlation coefficient is r=0.009, p=0.889>0.05.
Third-party sources have a weak association and positive insignificant correlation with Brand preference and purchase intention. The correlation coefficient is r=0.063, p=0.318>0.05.
The Tables 4-6 showed the results of regression analysis to examine the strength of relationship between independent variables and the dependent variable. The R-Square is 0.097, F-value is 3.229 and significance level is 0.002 which is <0.05. This indicates that the regression equation has achieved a satisfactory fitness level for explain the effects of independent variables on the dependent variable. R-Square indicates that the 9.7% variation of dependent variable caused by independent variables. The all independent variables have the positive relationship with the dependent variable; and company (β=0.272) has the highest positive impact on Brand preference and purchase intention followed by country of origin (β=0.234), Channel of distribution (β=0.069), Celebrity endorsement (β=0.053), Licensing (β=0.039), Third-party sources (β=0.025), Sporting cultural and other events (β=-0.004), and Co-branding (β=-0.034).
Table 4: Model summary
Model Summary | ||||
---|---|---|---|---|
Model | R | R Square | Adjusted R Square | Std. Error of the Estimate |
1 | .311a | 0.097 | 0.067 | 0.74272 |
aPredictors: (Constant), Third-party sources, Company, Sporting cultural and other events, Co-branding, Licensing, Channel of distribution, Country of origin, Celebrity endorsement
Table 5: Anova
ANOVAa |
||||||
---|---|---|---|---|---|---|
Model | Sum of Squares | df | Mean Square | F | Sig. | |
1 | Regression | 14.251 | 8 | 1.781 | 3.229 | .002b |
Residual | 132.942 | 241 | 0.552 | |||
Total | 147.193 | 249 | ||||
aDependent Variable: Brand preference and purchase intention
bPredictors: (Constant), Third-party sources, Company, Sporting cultural and other events, Co-branding, Licensing, Channel of distribution, Country of origin, Celebrity endorsement
Table 6: Coefficients.
Coefficientsa | ||||||
---|---|---|---|---|---|---|
Model | Unstandardized Coefficients | Standardized Coefficients | t | Sig. | ||
B | Std. Error | Beta | ||||
1 | (Constant) | 1.268 | 0.581 | 2.181 | 0.03 | |
Company | 0.272 | 0.102 | 0.169 | 2.664 | 0.008 | |
Country of origin | 0.234 | 0.075 | 0.196 | 3.113 | 0.002 | |
Channel of distribution | 0.069 | 0.04 | 0.107 | 1.703 | 0.09 | |
Co-branding | -0.034 | 0.043 | -0.051 | -0.796 | 0.427 | |
Licensing | 0.039 | 0.039 | 0.062 | 0.993 | 0.322 | |
Celebrity endorsement | 0.053 | 0.07 | 0.049 | 0.747 | 0.456 | |
Sporting cultural and other events | -0.004 | 0.048 | -0.005 | -0.084 | 0.933 | |
Third-party sources | 0.025 | 0.044 | 0.036 | 0.57 | 0.569 |
aDependent Variable: Brand preference and purchase intention
The strength of relationship between independent variables and the dependent variable is examined by the research model in this study. Company has the highest positive impact and country of origin has the second highest positive impact on Brand preference and purchase intention. This implies that consumers prefer and intend to purchase products from renowned company [5]. Wall et al. [7] suggested that consumers also prefer products/brands which are from renowned country in the case of particular products/brands indeed. Channel of distribution, licensing, celebrity endorsement, and third-party sources have the positive impact on Brand preference and purchase intention. This indicates that consumer buying decision process is influenced through brand ambassador, permission of business, delivery system, and out sourcing process. Keller [1] identified that consumers prefer products/brands which increase image of them among society and particular region.
Subsequently, co-branding, sporting cultural and other events have the negative impact on Brand preference and purchase intention. This implies that when consumers don’t prefer or intend to purchase products on the basis of co-brand, sponsored by sporting, cultural, and other events. Keller [1] suggested that consumers prefer and purchase products/brands according to the brand’s quality, price, manufacturers, and country of origin.
The research takes significant contributions on the field of marketing especially in strategic brand management. The first contribution is to know the factors for which consumers prefer particular products/brands or not. Moreover, the research will contribute to understand which factors of secondary brand association which motivate consumers in the case of product preferences and purchase intention. Fishbein and Ajzen [17] found that consumers choose products according to not only brand itself but also secondary brand knowledge of that particular brand. Therefore, the research gives insight knowledge about secondary brand association knowledge and its impact on Brand preference and purchase intention.
Copyright © 2024 Global Media Journal, All Rights Reserved