iBuss Management Vol. 3, No.

2, (2015) 99-108

The Impact of Brand Equity on Brand Preference and Purchase Intention

in Indonesia’s Bicycle Industry: A Case Study of Polygon

Felly Liliyana Soenyoto

International Business Management Program, Petra Christian University
Jl. Siwalankerto 121-131, Surabaya
E-mail: [email protected]

ABSTRACT

In the midst of stiff competition of local and foreign brands in Indonesia’s bicycle industry,
understanding the role of a brand in influencing consumer’s brand preference and purchase
intention is becoming more important as more choices are available for consumers. Thus, taking
the case study of Polygon, this research aims to investigate the impact of brand equity on
consumer’s brand preference and purchase intention as well as the possibility of brand preference
as the mediator between brand equity and purchase intention.
Through an online platform of questionnaire, 121 respondents were gathered from both
Polygon existing consumers and other bicycle consumers whom have not had Polygon bicycle.
The data gathered was analyzed using the linear regression and sobel test. The result shows that
brand equity has significant impact on brand preference and purchase intention. Brand preference
is also proven to significantly impact consumer’s purchase intention. Finally, brand preference is
confirmed to mediates the impact of brand equity on purchase intention. Thus, bicycle companies,
especially the PT Insera Sena as the company of Polygon brand, should focus on improving their
brand equity to increase consumer’s brand preference and purchase intention of their brand.

Keywords: Brand Equity, Brand Preference, Purchase Intention, Consumer Behavior, Brand
Management.

ABSTRAK

Di tengah ketatnya persaingan antara merek lokal dan merek asing di industri sepeda
Indonesia, pemahaman akan peran merek dalam mempengaruhi preferensi merek dan minat beli
konsumen sangatlah penting seiring dengan bertambahnya pilihan yang tersedia bagi konsumen.
Oleh karena itu, dengan studi kasus tentang merek sepeda Polygon, penelitian ini bertujuan untuk
mengetahui pengaruh ekuitas merek pada preferensi merek and minat beli konsumen sekaligus
kemungkinan preferensi merek berperan sebagai mediator antara ekuitas merek dan minat beli.
Penelitian ini dibuat berdasarkan 121 responden dari konsumen Polygon saat ini dan
konsumen sepeda yang belum memiliki sepeda Polygon, yang dikumpulkan melalui platform
kuesioner online. Data yang diperoleh kemudian dianalisa dengan menggunakan regresi linear
dan Sobel Test. Hasil penelitian menunjukkan bahwa ekuitas merek mempunyai dampak signifikan
pada preferensi merek dan minat beli konsumen. Preferensi merek juga terbukti secara signifikan
mempengaruhi minat beli konsumen. Oleh karena itu, perusahaan sepeda pada umumnya dan PT
Insera Sena sebagai perusahaan pemilik merek Polygon pada khususnya, perlu fokus untuk
meningkatkan ekuitas merek untuk meningkatkan preferensi merek dan minat beli konsumern
terhadap merek mereka.

Kata Kunci: Ekuitas Merek, Preferensi Merek, Minat Beli, Perilaku Konsumen, Brand
Management.

INTRODUCTION consumer for bicycle in ASEAN (Inabike, 2015). Based on

data from Indonesian Bicycle Industry Association, in 2010,
Bicycle industry in Indonesia has been growing over the market demand of bicycle is 5.5 million units which
the years both in demand and supply side. In the demand then increased 10% in 2011 to 6 million units (Caturini,
side, as mentioned in Inabikes (Indonesia International 2011). Finally, according to the Head of Indonesian Bicycle
Bikes Exhibition) official websites, Indonesia is the largest Industry Association (AIPI), Rudiyono, the market demand

99
 iBuss Management Vol. 3, No. 2, (2015) 99-108

of bicycle in 2014 has reached 6.5 million and in the future A good brand management could lead to a high brand
is projected to grow 20% annually (Issetiabudi, 2014). Not equity.
only that, international bike event and exhibition also Past researches has shown the importance of brand
annually being held in Indonesia, such as Inabikes equity. Brand equity can help company to outperform
exhibition which is also the ASEAN’s largest international competitors in market by differentiating the product and
trade show for bike, part, and accessories industry. lead to competitive advantage (Aaker, 1991). Beside brand
From the supply side, it seems that the opportunity equity, brand preference and purchase intention are chosen
raised from the growing market demand in Indonesia is to be analyzed in this research based on certain
captured by both local bicycle manufacturer and foreign considerations. Brand preference is important in indicating
manufacturers, making significant increase in the number of consumer’s option on a brand among other brands,
imported bicycle is inevitable. According to Indonesian especially in midst of stiff competition from both local and
Statistics Bureau (BPS), total value of imported bicycle foreign brands, while purchase intention is essential as it
increase 223.4% from US$ 42.8 million in 2012 to US$ will influence company’s financial performance through
138.5 million in 2013 (Yogatama, 2014). The increase in sales. Past researches (Allaway et al., 2011; Buil et al.,
the number of imported bicycle is also supported by a cheap 2008) found that “a strong brand with positive equity is
import tariff for the Completely Built Unit (CBU), which is considered to have high strategic value and advantages of
only 10%. It is cheaper compared to the import tariff for the higher margins, brand extension opportunities, insulation
bicycle parts which later on still need to be assembled, against competitors and more effective communicative
which is 10% to 15%. It makes the price of foreign bicycle power, as well as stronger consumer preferences, purchase
more competitive than the local brand. As a result, most of intentions and customer loyalty” (in Vincze, 2012, p.440).
local bicycle manufacturers choose to allocate 40% of the In this case, consumer preference refers to the brand
production for the International market, such as Europe and preference of consumer. In addition, brand preference is
America, and leaving 60 % of the production for the proven to have positive impact on purchase intention (Buil,
domestic market (Yogatama, 2014). Martinez, & Chernatony, 2013). It means the more
Therefore, the bicycle market in Indonesia is preferable a brand is, the higher the probability that the
dominated by imported bicycle. According to Head of customer will buy a product from that brand.
AIPI, Rusdiyono, in 2014, from 6.5 million units bicycle However, there has not been many researches
sold in the market, only 3 million units was local brand analyzing the possibility of brand preference as mediator
while the rest are imported brand (Issetiabudi, 2014). between brand equity and purchase intention, which make
Seventy percent of the imported bicycle are coming from the researcher want to find out the answer. A research by
China, 20% from Taiwan, and the rest are coming from US Sanjaya (2012) shows that strong brand equity will trigger
and South East Asia, such as Malaysia (Yogatama, 2014). some kind of favoritism of certain brand that will increase
Some of the top foreign brands that are widely sold in consumer willingness to buy the product of that brand.
Indonesia are Phoenix, Pacific, and XDS for the middle-low However, the research focused on service sector,
class and Giant, Merida, Specialized for the middle-up class specifically an insurance industry. Different industry may
(Bike Taiwan, 2013). For the local supplier, there are three has different result according the nature, consumer
biggest bicycle manufacturers that have been dominating behavior, and competition within the industry. In this
the market for years, namely PT Wijaya Indonesia Makmur research, the researcher try to analyze the manufacturing
Bicycle Industries with its brand “Wimcycle”, PT Insera sector, specifically bicycle industry, in which customer can
Sena with its brand “Polygon”, and PT Terang Dunia see, touch, and use the real product. Therefore, there might
Internusa with its brand “United Bike”. Those three be a possibility of different result obtained.
companies consistently achieved highest Top Brand Index Thus, after looking the above condition of bicycle
(TBI) among other bicycle brands in Indonesia from 2012- industry and past researches about the importance of brand
2014 (Top Brand Award, 2015). equity, brand preference, and purchase intention, in this
Based on the exploration about the market condition, research, the researcher decided to analyze the impact of
the concern arise as more companies entering the market brand equity on brand preference and purchase intention in
and selling similar products which are bicycles and Indonesia bicycle industry by taking a case study of one of
merchandise of bicycle-related product to the same market, top local bicycle brands, Polygon.
customers are having more choices of product as well.
Therefore, to win such a stiff competition arising from both LITERATURE REVIEW
local and overseas brand, of course relying on product
quality only is not enough for the bicycle companies. Brand There are three main concepts used as the backbone
is definitely a very powerful key distinction among products of this research, which are brand equity, brand preference,
in the market. According to American Marketing and purchase intention.
Association, brand is defined as “a name, term, design, To understand the concept of brand equity, it is
symbol, or any other feature that identifies one seller’s good important to first understand the basic concept of a brand
or service as distinct from those of other sellers” (American itself. According to American Marketing Association
Marketing Association, 2015). Therefore, as a key (2015), brand is defined as “a name, term, design, symbol,
distinction, a brand need to be carefully built and managed. or any other feature that identifies one seller’s good or

100
 iBuss Management Vol. 3, No. 2, (2015) 99-108

service as distinct from those of other sellers”. From that Table 1. Dimensions of Aaker’s Brand Equity Concept
definition, it is clear that a brand serve as a key distinction Dimension Description
among product of other sellers in the market. Therefore, a Brand awareness “the ability of a potential buyer to
brand that is carefully built and managed can actually recognize or recall a brand”
leverage the position and value of a product in consumer’s Brand association “anything linked in memory to a
mind which will eventually benefit the company as well. A brand”
well-established brand can be called as having a high brand Perceived quality “customers’ perception of the
equity. overall quality or superiority of a
There has been various definitions of brand equity product with respect to its intended
defined by many researches in the past due to growing purpose, relative to alternatives”
popularity of brand equity. Simon and Sullivan (1990) Brand loyalty “the attachment that a customer has
defined brand equity as the incremental discounted future to a brand”
cash flows of a product caused by a brand name which can Other proprietary e.g. trademarks, patents, channel
be measured through the value of tangible and intangible assets relationships
assets (in Keller, 1993, p.1). Aaker (1991) defined brand Source: Aaker (1991, p.15)
equity as a set of five brand assets (brand awareness, brand
association, perceived quality, brand loyalty, and other However, among those 5 dimensions of brand equity,
proprietary assets) that add value to a product. Other only three will be used in this research, which are brand
researchers, Lassar, Mittal, & Sharma (1995), defined brand awareness, brand association, and perceived quality. Brand
equity as consumers’ perception of product’s superior value loyalty dimension is not used as this research will
as a result of brand name attached to it, which can be investigate the impact of brand equity on brand preference
measured through 5 dimensions (performance, social and purchase intention not only to the existing consumer of
image, value, trustworthiness, and attachment). Anderson Polygon, but also to the potential consumer of Polygon
(2011) described brand equity as the financial value of which have not had Polygon bicycle. Therefore, it is not
customers’ response to the marketing of a brand. relevant to include ‘brand loyalty’ dimension. The “other
From the various definitions above, the researcher proprietary assets” dimension is not used as it is not related
concludes that there is indeed a red line among all those with consumer’s perspective.
definitions that is, brand equity is all about “value added” Brand preference is defined as “the biased behavioral
that lies in a product as a result of the brand attached to it. tendencies reflecting the consumer’s predisposition toward
The main point that differentiate one research and another is a brand” (Ebrahim, 2011) which refers the extent to which a
from which perspective the “value added” is seen and consumer favors one brand over another. Creating a strong
dimension used in measuring that value. Generally, there bond between brand and customer is ultimately one of the
are two perspectives in defining brand equity, which are goals of marketing managers. Understanding brand
firm’s perspective and consumer’s perspective. From firm’s preference is crucial in achieving that goal as it is believed
perspective, brand equity is seen more in terms of financial to influence consumer intention to purchase a product or
value that a brand can give to the company as a measure of conveying a positive word-of-mouth (Ebrahim, 2011). A
success and performance of the company. Brand equity strong preference toward a brand can help the brand to
defined from the perspective of consumers, focus on the sustain in the long run despite of stiff competition from
interaction of customer and brand as well as the other brands. The reason is that as there are more brands
consequences result in that relationship. coming into the market selling the same product category,
One of the argument in defining brand equity from consumers’ have more choices of brands; therefore, a more
consumer’s perspective is that financial value of brand favorable brand will be more likely to be chosen by
equity is only the outcome of the consumer’s response to consumers compared to those who are not.
the brand name itself (Christodoulides & Chernatony, 2009) Past researchers (Sirgy, 1997; Buil, Martinez, &
while the power of a brand actually lies in what customers Chernatony, 2013) measured brand preference using
have learned, felt, seen, and heard about the brand as a result indicators such as “I like brand X better than other brands of
of their experiences over time (Keller, 2003). Therefore, product category”, “I would use brand X product category
“for a brand to have value, it must be measured by more than other brands of product category”, and “In the
customer”, which mean that “if the brand has no meaning to product category, brand X is my preferred brand”. Other
the customer, none of other definitions is meaningful” researchers (Chen & Chang, 2008; Sanjaya, 2012) used
(Atilgan, Aksoy, & Akinci, 2005, p.238). That also become four indicators such as, “I feel that brand X is appealing”, “I
the underlying reason for the researcher to focus on prefer this brand X to other brand of its type”, “If I was to
analyzing the impact brand equity from consumer’s buy a product from this category, I would prefer this brand
perspective in this research. X if everything else was equal”, “Overall, I have a very
This research will use the dimensions of brand equity strong preference for X”. Five out of those seven indicators
introduced by Aaker (1991) to measure brand equity. will be used as the indicators of brand preference in this
research. The third indicator from Sirgy (1997) and Buil,
Martinez, & Chernatony (2013) is actually similar with the

101
 iBuss Management Vol. 3, No. 2, (2015) 99-108

second and third indicator of Chen & Chang (2008) and intention has been confirmed by Buil, Martinez, &
Sanjaya (2012), therefore, only one of them will be used. Chernatony (2013), Sanjaya (2012), and Prabhawedasttya
Intention is “motivational factors that influence a & Yasa (2013) who found that brand preference have
behavior” (Ajzen, 1991). Therefore, according to Dodds et positive influence toward purchase intention. It means the
al. (1991), purchase intention can be defined as the more preferable a brand is, the higher the probability that
possibility for consumers to buy a product (in Huong, the customer will buy a product from that brand. The theory
2012). Similarly, Sam & Tahir (2009) also defined purchase planned behavior by Ajzen (1991) explained that a
intention as the probability that a consumer will purchase favorable attitude towards a brand leads to purchase
the product. In brand management, purchase intention is intention. So, when consumer like certain brand, they are
one of the important factor in predicting consumer behavior more motivated to buy the product rather than buying a
(Durianto & Liana, 2004). product from the brand that they do not like.
Purchase intention can be classified into two, which The relation among brand equity, brand preference,
are purchase intention of new consumers and purchase and purchase intention as explained above, leads to a
intention of existing consumers. Purchase intention of new possibility that brand preference may serve as the mediator
consumers (potential consumers) is essential to identify the between brand equity and purchase intention. Sanjaya
likelihood of future sales to occur and to know whether the (2012) tried to investigate that possibility in the insurance
current brand position is strong enough to motivate new industry. In the result, brand preference was proven to
consumers to buy the product since each consumer has mediate the impact of brand equity on purchase intention. It
different perception and perceived value toward a brand. means that brand equity actually affect consumers’
Identifying purchase intention of existing consumer is purchase intention indirectly by forming a strong preference
useful to predict the possibility of buying more products of a brand on consumers’ mind which then influence them
from the brand (Madahi & Sukati, 2012) which eventually to purchase the product of that brand. In this research, the
will contribute to company’s financial performance as well researcher try to analyze the same issue but in different
as giving an insight for improvement of the brand position. industry, which is bicycle industry to see whether the same
In this research, purchase intention from both type of result is shown or not.
consumers will be investigated. Besides the theories explained in the above section,
Two past researches done by Erdem et al. (2006) and the researcher also use four relevant researches with similar
Buil, Martinez, & Chernatony, (2013) measured purchase research framework as depicted in Figure 1, as reference
intention based on the degree of seriousness to buy the and additional knowledge for conducting this research.
product, such as the intention of consumer to buy the Those four relevant researches are research by Buil,
product, the intention to consider buying the product Martinez, & Chernatony (2013) on 6 brands of sportswear,
seriously, and the possibility of them to actually buy the consumer electronics, and cars products, Chen & Chang
product. (2008) on airline industry, Prabhawedasttya & Yasa (2013)
on smartphone products , and Sanjaya (2012) on insurance
Relationship between Concepts industry. The result of all those four relevant researches
support the theorical background used in this research about
brand equity, brand preference, and purchase intention.
On the basis of the relation among brand equity, brand
preference, and purchase intention concepts, the four
hypotheses below will be tested in this research.
H1: Brand equity significantly impacts brand preference
of Polygon
Figure 1. Relationship between Concepts H2: Brand preference significantly impacts purchase
intention of Polygon
Myers (2003) found that brand with higher equity H3: Brand equity significantly impacts purchase intention
generated greater brand preference. Other researchers, of Polygon
Prasad and Dav (2000) also found that high brand equity H4: Brand preference mediates the impact of brand equity
lead to a high brand preference. It shows that the more on purchase intention of Polygon
added value the consumers get from a brand, the more
likely they will like the brand and favor that brand over RESEARCH METHOD
another. Besides, brand equity is also confirmed to bring According to Cooper & Schindler (2014), there are
positive influence on consumers’ purchase intention (Cobb- four types of research, which are reporting, descriptive,
Walgren, Ruble, & D, 1995). The higher the brand equity it explanatory, and predictive research. This research can be
means the more consumers can easily recognize the brand, categorized as an explanatory research because it tries to
consumers are highly connected with the brand, consumers investigate the impact of brand equity on brand preference
perceive the brand as high quality; therefore, the more likely and purchase intention and explain why that relationship
consumer will like the brand, choose the brand over other occur with the related theory as the base of the research and
brands, and motivated to actually buy the product of that to generate hypothesis.
brand. The relation between brand preference and purchase
102
 iBuss Management Vol. 3, No. 2, (2015) 99-108

In this research, there are three types of variables used, please the interviewer or tend to avoid answers that they
which are independent variable (brand equity), mediating perceived to be socially unacceptable answer (Garland,
variable (brand preference), and dependent variable 1991). By using the 6-point Likert scale, it is easier to
(purchase intention). The measurements used for each identify whether the respondents express a (an)
variable are explained in the following tables. “disagreement” or “agreement”.
There are two types of data scale used in this research
Table 2. Indicators of Brand Equity which are nominal and interval data. Nominal data can be
Indicator Reference found in the screening question part, which include gender,
“I am aware of Polygon brand” Prabhawedasttya & category of age, occupation, category of monthly income,
Yasa (2013) category of intensity in cycling, and brand of bicycle
“I can recognize Polygon bicycle Prabhawedasttya & owned. The answers to these questions are considered as
among other competing brand of Yasa (2013) nominal data as it is presented in the form of multiple
bicycle” choices, in which the numbers representing the answers are
“I consider Polygon as a trusted Buil, Martinez, & only codes which shows no particular order, distance, or
brand” Chernatony (2013) natural origin. The interval data is generated from the
“I associate Polygon as bicycle Pappu, Quester, & answers of the target questions regarding brand equity,
for middle-up market” Cooksey (2005) brand preference, and purchase intention. It is classified as
“Polygon offers very good quality Buil, Martinez, & interval data as the average of the data measure the same
bicycle” Chernatony (2013) distance between any two points (Cooper & Schindler,
“Polygon offers bicycle with Buil, Martinez, & 2014).
excellent specification” Chernatony (2013) The data for this research is collected from three types
of sources, which are primary sources, secondary sources,
Table 3. Indicators of Brand Preference and tertiary sources. The primary source used is online
Indicator Reference questionnaires, which is filled in by the respondents that is
“I feel that the design of Polygon Chen & Chang part of the population. Other than primary sources, the
researcher also collect information from the secondary and
bicycle is appealing” (2008); Sanjaya
tertiary sources to support the analysis of this research. The
(2012)
secondary sources used are books (both online and offline),
“I like Polygon better than other Sirgy (1997); Buil,
journals, articles from website, online newspaper, and
brands of bicycle” Martinez, &
dissertation. This research take advantage of internet search
Chernatony (2013)
engine (e.g. Google scholar) and journal database (e.g.
“I would use Polygon bicycle Sirgy (1997); Buil,
Science Direct, Emerald) to help identify and local online
more than other brands of Martinez, &
books and journals online.
bicycle” Chernatony (2013)
The sampling method used is the simple random
“I would prefer to buy Polygon Chen & Chang
sampling under the probability sampling method in which
rather than other brands of (2008); Sanjaya
each element of the population has equal chance to be
bicycle.” (2012) selected into the sample. The population of this research
“Overall, I have a very strong Chen & Chang include both potential and existing consumer of Polygon in
preference for Polygon” (2008); Sanjaya Indonesia. The potential consumer is defined as people who
(2012) like cycling and know (ever heard and ever seen) Polygon
bicycle but have not had Polygon bicycle. They can be
Table 4. Indicators of Purchase Intention regarded as potential consumer of Polygon as they might be
Indicator Reference a possibility for these people to buy Polygon in the future as
“I will buy Polygon bicycle” they like cycling, know Polygon brand, and have not had
“I will seriously consider Erdem et al. (2006); Polygon bicycle. The existing consumer of Polygon is
buying Polygon bicycle” Buil, Martinez, & defined as people who have had Polygon bicycle. To obtain
“It is very likely that I would Chernatony, (2013) a more reliable result, the researcher set the minimum age of
buy Polygon bicycle” 18 to be the respondents of this research since they are
considered to be fully capable to understand the
Likert scale is used in measuring all variables because questionnaire’s questions and have the buying power to be
it is easy and quick to be answered by the respondents and the consumer of Polygon. So, the population of this research
most importantly it is believed to provide a more reliable will be all people who like cycling, know about Polygon
and a greater volume of data than any other scale (Cooper bicycle, have or have not had Polygon bicycle whose age
& Schindler, 2014). Specifically, in this research the 6-point are above 18, and reside in Indonesia.
Likert scale is used because the researcher wants to avoid A good sample need to represent the characteristics of
the “neutral” answers that often dominating the result when the population well (Cooper & Schindler, 2014). Therefore,
an odd Likert scale (e.g. 5-point Likert scale) is being used. the questionnaire will be distributed randomly to people
Omitting mid-point or neutral answer might minimize who meet those criteria. To reach a greater number of
social desirability bias, in which the respondents try to respondents, the questionnaire will be distributed through an

103
 iBuss Management Vol. 3, No. 2, (2015) 99-108

online platform, Google Docs, to several online cycling needed to answer the hypothesis, instead of doing it
communities in Indonesia. The simple random sampling separately for each of the analysis. The conceptual
method enable an opportunity for the researcher to framework used as the basis of the mediation analysis is as
generalize the findings to the population of interest from follows:
sample population.
The number of the sample size will be determined
based on the formula from Green (2001) which is N ≥ 50 +
8k where N is the number of minimum sample size needed
and k is the number of predictor or independent variable (as
cited in Brooks & Barcikowski, 2012). In this research,
there is only one independent variable, which is brand
equity, therefore, the minimum sample size needed is 58
(50+8(1)).
In order to answer the four hypothesis in this research,
an additional package of IBM SPSS 21 macros downloaded
from afhayes.com is used as a mean to analyze the data
collected from the questionnaires.Before testing all Figure 2. Framework of Mediation Analysis
Source: Modified from Preacher & Hayes (2008, p.880)
hypothesis, the validity and reliability of the data need to be
tested for justification of the data. After the data is found to
be both valid and reliable then the classic assumption test Panel A shows a direct effect of X on Y. This simple
need to be fulfilled as requirements for running the relation between X and Y is called as total effect with c as
mediation analysis. the coefficient value. Panel B represents a simple mediation,
which is the indirect impact of X on Y through M as the
A validity and reliability test are performed to ensure
mediator. Coefficient c’ is the direct effect of X on Y after
that the data collected from the questionnaire is both valid
controlling M.
and reliable. The validity of the indicators is measured by
comparing r-data of each indicator and r-table with the According to Baron and Kenny (1986), a variable can
degree of freedom (df) = n-2, where n is the number of be indicated as a mediator if:
X significantly impact Y (c path is significant)
sample (Ghozali, 2013). The r-data of each indicator can be
X significantly impact M (a path is significant)
seen from the Cronbach Alpha output in the SPSS,
M significantly impact Y by controlling X (b path is
specifically under Corrected Item-Total Correlation
significant)
Column. If the r-data greater is greater than the r-table at the
specified degree of freedom, it means that the indicator used The impact of the independent variable on dependent
is a valid measurement of the variable. The reliability is variable of each path is tested through the t-test. The
significance of each path can be seen by looking at the p
tested based on the Cronbach Alpha value. If the Cronbach
value of each regression. For the decision rule, if the p-value
Aplha value is higher than 0.7 the indicators are said to be
is higher than the significance level (α) of 0.05 then the null
reliable (Ghozali, 2013).
As the mediation analysis will include some hypothesis is failed to be rejected. It means that there is no
regression models, some classic assumption tests need to be significant impact between variables in the regression
performed to ensure a correct regression analysis is used. model. On the other hand, if the p-value is lower than the
significance level (α) of 0.05, then the null hypothesis will
The classic assumption tests include normality test,
be rejected, which mean that there is significant impact
multicollinearity test, heteroscedasticity test, and
between variables in the regression model (Lind, Marchal,
autocorrelation test (Ghozali, 2014). However, in this
research, the autocorrelation test does not need to be & Wathen, 2010).
conducted as it aims to test whether or not there is a The significance of the indirect effect, which is the
product coefficient of and b path, is then tested directly
correlation between the residual on period t and the residual
using the Sobel test to reveal whether or not the mediating
in the previous period (t-1) (Ghozali, 2014). Autocorrelation
variable really mediates the impact of the independent
test is only needed when a longitudinal research with time
variable on the dependent variable. The hypothesis for
series data is conducted. This research is considered as the
cross-sectional research as it is only conducted once and testing the significance of the indirect effect (ab) is as
represent a snapshot of one point in time, therefore, an follows:
H0: ab = 0
autocorrelation test can be omitted (Ghozali, 2014).
H1: ab ≠ 0
In this research, the mediation relationship is analyzed
The result of the Sobel test can be seen in the normal
using Sobel test. Sobel test will investigate directly whether
the indirect effect is significant or not and how much is the theory section. As for the decision rule, the null hypothesis
indirect effect (Ghozali, 2014). The Sobel test is performed is failed to be rejected if the p-value of the indirect effect is
through an additional macro of IBM SPSS 21 (downloaded higher than the significance level (α) of 0.05 or the z-value
is lower than the z-table at that significance level, which is
from afhayes.com). The additional macros is used as it can
1.96, meaning that the mediation impact is not significant.
analyze the direct effect, total effect and indirect effect of
On the other hand, the null hypothesis will be rejected if the
each variable at once and give the big picture of all answers
p-value of the indirect effect is lower than the significance
104
 iBuss Management Vol. 3, No. 2, (2015) 99-108

level (α) of 0.05 or the z-value is higher than the z-table at Brand equity has significant impact on purchase
that significance level, which is 1.96, meaning that the intention
mediation impact is significant (Preacher & Hayes, 2004).
If the mediation impact is proven to be significant Table 5. Path Testing Results
from the result of Sobel test, the type of the mediation will Test Path Coefficient SE p-value
then be identified. A perfect mediation happens when a BE to BP a 0.8776 0.0771 0.0000
path, b path, and c path are proven to be significant but Direct effect b
when M is included in the computation (c’ path), the impact 0.6654 0.0464 0.0000
of BP to PI
of X on Y becomes zero or insignificant (p-value > α). If the Total effect c
impact of X on Y decreases after including M, but not equal 0.9078 0.0644 0.0000
of BE to PI
to zero or still significant (p-value < α), a partial mediation Direct effect c’
is said to happen (Preacher & Hayes, 2004; Ghozali, 2014). 0.3238 0.0564 0.0000
of BE to PI

RESULTS AND DISCUSSION As brand equity is proven to be significantly impact

There are 125 respondents filled out the online brand preference of Polygon, it means the first hypothesis
questionnaires in which four of their responses were (H1) of this research in confirmed. The current result is in
immediately excluded from the responses pool as they do line with the result of the four relevant researches of Buil,
not like cycling and do not know about Polygon, which Martinez, & Chernatony (2013), Chen & Chang (2008),
means that they are not the member of the population to be Prabhawedasttya & Yasa (2013), Sanjaya (2012). The
taken as the sample. Thus, in total there are 121 responses to significant impact of brand equity on brand preference can
be processed further. be explained as follows: the higher the brand equity it
First of all, a validity and reliability testing is means the easier consumers can recognize the brand,
conducted. The result of validity test shows that the r-data of consumers are highly connected with the brand, consumers
each indicator is bigger than the r-table (with df=119), perceive the brand as high quality; therefore, the more likely
which is 0.1786. It means that all indicators are valid. The consumer will like the brand and favor that particular brand
result of reliability test shows that all indicators are reliable over another. Conversely, it less likely for a brand that is
as the Cronbach’s Alpha value are all higher than 0.7, which barely even known by the consumers, not associated to
are 0.885 for brand equity, 0.963 for brand preference, and certain positive image, and perceived as a brand with poor
0.925 for purchase intention. quality, to be chosen as a preferred brand by the consumers.
The data also has passed all the classic assumption Therefore, building a strong brand is important to get a high
tests, which are normality test, multicollinearity, and overall preference over another competing brand.
heteroscedasticity, which mean that the data is good enough Brand preference is proven to have significant impact
for regression that will be used as part of the mediation on purchase intention. It confirms the second hypothesis
analysis. The mediation analysis is performed based on the (H2) of this research which stated that brand preference
following conceptual framework: significantly impact purchase intention of Polygon. The
result of this research also confirms the four past researches
of Chen & Chang (2008), Sanjaya (2012), Buil, Martinez,
& Chernatony (2013), and Prabhawedasttya & Yasa (2013).
The four past researchers also found that brand preference
have significant impact to purchase intention. Significant
impact means the more preferable a brand is, the higher the
probability that the customer will buy a product from that
brand. When consumer like certain brand, they are more
motivated to buy the product rather than buying a product
from the brand that they do not like. The theory of planned
behavior by Ajzen (1991) also supports this argument, as
Figure 3. Framework of Mediation Analysis of the stated that one of the construct of individual’s behavioral
Research intention is attitude toward behavior. The more favorable
attitude toward performing certain behavior, the stronger the
The path testing as illustrated in Figure 3 is conducted intention to perform that behavior. In the case of purchase
to examine whether or not the criteria of a mediator variable intention, it means when consumer feels positive about
is fulfilled. The result of the t-test shown in Table 5 shows choosing, favoring, or buying the product of a brand, the
that the p value of a path, b path, c path are all lower than higher their willingness to actually choose and buy the
0.05. Thus, the null hypothesis is rejected, meaning: product of that brand. Thus, it is important for a company to
Brand equity has significant impact on brand form a strong preference of its brand in consumer’s mind so
preference that more consumers will buy the product of that brand and
Brand preference has significant impact on purchase for the existing consumers, the more likely they will keep
intention buying the product in the future.

105
 iBuss Management Vol. 3, No. 2, (2015) 99-108

Brand equity is proven to significantly influence Brand preference as the mediator of brand equity and
purchase intention, which means the third hypothesis (H3) purchase intention means that brand equity actually affect
of this research is also confirmed. Comparing with the past consumers’ purchase intention indirectly by forming a
researches, the current result supports the result of the past strong preference of a brand on consumers’ mind which
researchers who also found that brand equity have then influence them to purchase the product of that brand.
significant impact on purchase intention. Significant impact Consumer’s knowledge, association, and perception of a
of brand equity on purchase intention means that the higher brand are transformed into some favoritism and biased
the “added value” that consumers felt from a brand, or in behavioral tendencies toward certain brand which then
other words, the more familiar they are with the brand, the influence them to prefer particular brand over another. In
more positive their association towards the brand, and the accordance with the theory of planned behavior of Ajzen
more positive their perception about the quality of the (1991), this preference (favorable attitude) is then the one
product; therefore, the more likely they will buy the product that increase their motivation to purchase the product from
from that brand. On the other hand, brand with low equity that brand.
means that consumers are not really aware of the brand, In this research, however, the mediation is proven to
does not associate the brand with certain positive image, be partial, it means that the impact of brand equity does not
does not perceive the brand as a high quality product; always necessarily to be transformed into brand preference
therefore, the less likely they will buy the product from that to motivate consumer to purchase the product of certain
brand. brand. The knowledge, association, and perception of
The result of Sobel Test for testing the significance of consumers towards certain brand also can directly motivate
the indirect effect are presented in Table 6. It shows that the them to buy the product. This phenomena can be explained
p-value is 0.0000, which is lower than the significance level by referring to the theory of planned behavior of Ajzen
(α) of 0.05, and the z-value is 8.9650 which is higher than (1991) which explained that there are two other constructs
the z-table (1.96) showing that the null hypothesis is of behavioral intention, which are subjective norm
rejected. It means that the mediation impact of brand (perceived social pressure to perform or not to perform a
preference is proven to be significant. behavior) and perceived behavioral control that a consumer
has. Thus, even when the preference is not yet formed in
Table 6. Sobel Test Results for Indirect Effect consumer’s mind, but if the knowledge, association, and
Test Path Effect z-value p-value perception of the consumer toward a brand is positive,
adding with supportive social pressure or enough resources
Indirect effect to the buy the product, there is possibility that the consumer
of BE on PI ab 0.5840 8.9650 0.0000 will buy the product.
through BP
CONCLUSION
The type of mediation (full or partial mediation) will
After analyzing data and interpreting the result, the
then be identified by examining the direct effect of the researcher will then summarize the content of this whole
independent variable on dependent variable after including research. An introduction of the condition and problems of
the mediating variable (c’ path). The result in Table 5 shows Indonesia’s bicycle industry is presented as the background
that the impact of brand equity on purchase intention after of the research which then enable the researcher to come up
including brand preference decreases from 0.9078 to with the four statements of research problem to be analyzed
0.3238, but is not equal to zero. In addition, the p-value is in this research. The research problems are basically focus
also lower than the significant level (α) of 0.05, showing on revealing the impact of the brand equity on brand
that the direct effect of brand equity on purchase intention preference and purchase intention as well as the possibility
after including brand preference is still significant. Thus, it of brand preference acting as the mediator between brand
can be concluded that there is a partial mediation between equity and purchase intention. To address each of the
brand equity and purchase intention with brand preference research problems, four hypothesis are then developed
as the mediator. In the case of partial mediation, the accordingly. One hundred and twenty one responses are
independent variable have both direct and indirect effect to used as the sample of this research. Several tests, such as
the dependent variable. reliability-validity test, classical assumption test, and
The above results confirms the fourth hypothesis (H4) mediation test are conducted to process the data obtained
of this research which stated that brand preference mediates from the questionnaires. The result shows that brand equity
the impact of brand equity and purchase intention of has significant impact on brand preference, brand
Polygon. The current result confirms the past research of preference has significant impact on purchase intention,
Chen & Chang (2008) that also found brand preference brand equity has significant impact on purchase intention,
mediates the impact of brand equity on purchase intention and brand preference mediates the impact of brand equity
with an indirect effect of 0.40. In addition, the current result on purchase intention. It means the four hypothesis of this
also confirms the latest research of Sanjaya (2012) that research are all confirmed. Thus, by revealing the result of
brand preference act as the mediator of between brand the hypothesis testing, all research objectives have also been
equity and purchase intention. accomplished.

106
 iBuss Management Vol. 3, No. 2, (2015) 99-108

During the completion of this research, the researcher Regression. Multiple Linear Regression Viewpoints,
realizes that there are also several limitations of this 38(2), 1-16.
research. Firstly, regarding the research method, since all Buil, I., Martinez, E., & Chernatony, L. d. (2013). The
variables being investigated (brand equity, brand influence of brand equity on consumer responses.
preference, purchase intention) are actually considered as Journal of Consumer Marketing, 30(1), 62-74.
latent variables, which are the variables that cannot be Caturini, R. (2011, December 21). Sepeda booming, bisnis
observed directly. It is suggested for the future research to pun melejit. Retrieved February 24, 2015, from
use Partial Least Square (PLS) or Generalized Structured Kontan.co.id:
Component Analysis (GSCA) instead of regression. PLS http://lipsus.kontan.co.id/v2/sepeda/read/21/Sepeda-
or GSCA is better fitted analysis method for research with booming-bisnis-pun-melejit
latent variables, one way casuality, and reflective and Chen, C. F., & Chang, Y. Y. (2008). Airline brand equity,
formative indicators (Lohmöller, 2013; Solimun, 2015). brand preference, and purchase intentions-The
Secondly, in this research, the researcher does not moderating effects of switching costs. Journal of Air
specifically ask about the geographical area of the Transport Management, 14(1), 40-42.
respondents in the screening questions, therefore, there is a Christodoulides, G., & Chernatony, L. d. (2009). Consumer
possibility that the sample might cover or might not cover
Based Brand Equity Conceptualization &
various area of Indonesia. Since different geographical area
Measurement: A Literature Review. International
might have different characteristics of consumers, different
Journal of Market Research.
result might be obtained for different coverage area
included in the sample. For further research, it is suggested Cobb-Walgren, C., Ruble, C., & D, N. (1995). Brand
to include specific question asking about the geographical Equity, Brand Preference, Purchase Intention. Jornal
area of the respondents to ensure a broad coverage area of of Advertising, 24(3), 25-40.
sample is reached. Cooper, D. R., & Schindler, P. S. (2014). Business
Finally, the scope of this research is limited to a single Research Method (12 ed.). New York: McGraw-
brand of bicycle, which is Polygon, while there are many Hill.
other bicycle brands available in Indonesia’s bicycle Ebrahim, R. (2011). A Study of Brand Preference: An
industry. Therefore, future researchers who wants to Experiential View. London: Brunel Business
conduct similar research regarding bicycle industry or School.
similar topic of brand equity is suggested to conduct the Garland, R. (1991). The mid-point on rating scale: Is it
research on several brands (2 or 3 brands) so that a better desirable? Retrieved April 6, 2015, from
picture of market condition in Indonesia’s bicycle industry http://rangevoting.org/MB_V2_N3_Garland.pdf
could be obtained. Then, the result can be compared with Ghozali. (2013). Analisis Multivariate dengan Program
the current result of this research. IBM SPSS 21 (7th ed.). Semarang: Badan Penerbit
Universitas Diponegoro.
REFERENCES Ghozali, I. (2014). Ekonometrika : Teori, Konsep, dan
Aplikasi dengan IBM SPSS 22. Semarang: Badan
Aaker, D. A. (1991). Managing brand equity. New York: Penerbit Universitas Diponegoro.
Free Press. Huong, N. T. (2012). Key Factors Affecting Consumer
Ajzen, I. (1991). The theory of planned behaviour. Purchase Intention: A study of Safe Vegetable in Ho
Organizational behavior and human decision Chi Minh City, Vietnam.
processes, 50, 179-211. Inabike. (2015). Indonesia International Bike Exhibition.
American Marketing Association. (2015). Dictionary. Retrieved 2 24, 2015, from Inabike.net:
Retrieved March 3, 2015, from American http://www.inabike.net/market-
Marketing Association: information#axzz3SfbVC9Ec
http://ama.org/resources/Pages/Dictionary.aspx?dLe Issetiabudi, D. E. (2014, July 30). Pasar Sepeda: Pemerintah
tter=B&dLetter=B Diminta Adil Perlakukan Produk Impor. (Y. B.
Anderson, J. (2011). Measuring the FInancial Value of Widagdo, Editor) Retrieved February 24, 2015,
Brand Equity: the Perpetuity Perspective. Journal of from business.com:
Business Administration Online, 10(1), 1-11. http://industri.bisnis.com/read/20140730/257/24671
Atilgan, E., Aksoy, S., & Akinci, S. (2005). Determinants of 4/pasar-sepeda-pemerintah-diminta-adil-perlakukan-
the brand equity: A verification approach in the produk-impor
beverage industry in Turkey. Marketing Intelligence Janice , J., & Leosaputro, J. I. (2014). The Impact of
& Planning, 23(3), 237-248. Corporate Image of PT XYZ towards Job Applicant
Bike Taiwan. (2013). The Indonesian Bicycle Market. Attraction in Surabaya. iBuss Management, 2(2),
Retrieved 2 24, 2015, from Biketaiwan.com: 12-20.
http://www.biketaiwan.com/resource/article/2/28/art Keller, K. L. (1993). Conceptualizing, Measuring, and
icle-16.pdf Managing Customer-Based Brand Equity. Journal
Brooks , G. P., & Barcikowski, R. S. (2012). The PEAR of Marketing, 57, 1-22.
Method for Sample Sizes in Multiple Linear

107
 iBuss Management Vol. 3, No. 2, (2015) 99-108

Keller, K. L. (2003). Strategic Brand Management: (E. Djumena, Editor) Retrieved February 24, 2015,
Building, Measuring and Managing Brand Equity from Kompas.com:
(Second ed.). Englewood Cliffs: Prentice-Hall. http://bisniskeuangan.kompas.com/read/2014/09/03/
Lassar, W., Mittal, B., & Sharma, A. (1995). Measuring 134213026/Sepeda.Impor.Membanjiri.Indonesia.In
Customer-Based Brand Equity. Journal of dustri.Lokal.Pilih.Ekspor
Consumer Marketing, 12(4), 11-19.
Lohmöller, J.-B. (2013). Latent Variable Path Modeling
with Partial Least Squares. Berlin: Springer Science
& Business Media,.
Madahi, A., & Sukati, I. (2012). The effect of external
factors of purchase intention amongst young
generation in Malaysia. International Business
Research, 5(8).
Myers, C. A. (2003). Managing brand equity: a look at the
impact of attributes. Journal of Product & Brand
Management, 12(1), 39-51.
Pappu, R., Quester, P. G., & Cooksey, R. W. (2005).
Consumer-based brand equity: improving the
measurement - empirical evidence. Journal of
Product & Brand Management, 143-154.
Prabhawedasttya, I. G., & Yasa, N. N. (2013). Pengaruh
Ekuitas Merek terhadap Preferensi Merek dan Niat
Beli Konsumen pada produk iPhone di Kota
Denpasar. E-Jurnal Manajemen Universitas
Udayana, 2(11), 1525-1551.
Prasad, K., & Chekitan, S. D. (2000, June 30). A customer-
centric framework for assessing performance. Hotel
Brand Equity, pp. 22-31.
Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS
procedures for estimating indirect effects in simple
mediation models. Behavior Research Methods,
Instruments, & Computers, 36(4), 717-731.
Preacher, K., & Hayes, A. (2008). Asymptotic and
resampling strategies for assesing and comparing
indirect effects in multiple mediator models.
Behavior Research Methods, 40(3), 879-891.
Sam, M. F., & Tahir, M. N. (2009). Website Quality and
Consumer Online Purchase Intention of Air Ticket.
International of Journal of Basic & Applied
Sciences, 9(10), 4-9.
Sanjaya, E. (2012). Pengaruh Brand Equity terhadap
Purchase Intention Melalui Brand Preference pada
Produk PT. Sunlife Financial di Surabaya. Jurnal
Universitas Katolik Widya Mandala Surabaya, 1(5).
Solimun. (2015). Analisis Multivariat: Generalized
Structured Component Analysis. Pendidikan dan
Pelatihan Dosen "Generalized Structured
Component Analysis". Malang: Universitas
Brawijaya.
Top Brand Award. (2015). Top Brand Award. Retrieved
from topbrand-award.com: http://www.topbrand-
award.com/
Vincze, U. H. (2012). Consumer-based brand equity and
top-of-mind awareness: a cross-country analysis.
Journal of Product & Brand Management, 21(6),
439 - 451.
Yogatama, B. K. (2014, September 3). Sepeda Impor
Membanjiri Indonesia, Industri Lokal Pilih Ekspor.

108