The effects of brand relevance and self-expansion on brand loyalty in green

cosmetics industry

Introduction

Brand relevance has never been less important in the green cosmetics industry. Women buy
cosmetics not only to embrace their skin but also to define themselves, so they seem to be
more inclined to choose brands that are closely related to their needs. Once society has
developed modern technology that has allowed shoppers to have thousands of shopping
options, why do they choose and stay loyal to your brand? It is the degree of relevance of
the brand with the customer, the more customers find them in the brand, the more engaged
and intimate they are.

On the other hand, nowadays theory of self expansion also plays an important part in
building relationships between customers and brands. Customers become aware of
opportunities for self-expansion when they interact with brands that offer new perspectives,
new resources, and new identities (Reimann & Aron, 2009). It means that the more
customers feel the value of green products of a brand, the more these will positively
influence their self expansion. This research argues that if a green cosmetic’s brand can
provide self-expansive opportunities such as new knowledge, experience, self-improvement
about environment, green lifestyle…, it can motivate fans to engage in its brand.

This research will discuss two main things which are the relationship between brand
relevance and self-expansion model and the relationship between brand relevance and self
loyalty in the green beauty industry.

Research Objectives and Research Question ( viết thành đoạn)

The broad objective of this research is to examine how brand relevance and self-expansion
affect brand loyalty in the green cosmetic industry. Two studies that will be concluded in this
research are the relationship between brand relevance and brand loyalty, and another one is
the relationship between self-expansion and brand loyalty in the green cosmetics industry.

For this research to achieve its objective, the research questions that need to be answered
are “Is there a positive relationship between brand relevance and brand loyalty in the green
cosmetics industry?” and “Is there a positive relationship between self-expansion and brand
loyalty in the green cosmetics industry?”.

Literature review (Tách thành từng phần = thêm Hyo + Conceptual)

Green cosmetics don’t mean your beauty products will be green in color. Green here is all
about nature. You might be hearing or seeing terms like “Organic”, “Natural”, “Herbal”,
“Cruelty-free” “Eco-friendly”, “Vegan”, “Biodegradable”, “Recyclable”, etc. in advertisements
or printed on products which is what going green is all about. All these terms give us the idea
that they are cosmetics generally manufactured using all-natural, non-toxic ingredients and
are safer for human use and gentler to the environment.

Brand relevance is a company's ability to connect with people's emotions and become
personally relevant to them. It is not about how customers think but how they feel about the
brand. (Thibaud Clement, 2020). What previous studies have done very well is to show the
brand relevance in categories, brand relevance applying in education (From research
excellence to brand relevance, 2016), brand relevance in brand equity (The value relevance
of brand equity in the financial services industry: an empirical analysis using quantile
regression, 2012), as well as talk about how to win brand relevance among competitors (Win
the Brand Relevance Battle and then Build Competitor Barriers, 2012). However, there is
one aspect they have not exploited, which is the effect of brand relevance on brand loyalty,
one of the aspects that contributes a lot to brand equity. Therefore, this research is
determined to show the relationship between brand relevance and brand loyalty, especially
their mutual influence.

Self-expansion model is a conceptual framework that seems as a platform for researching

relationships and relationship-linked phenomena. The ideas in the self-expansion model
renew the view of “similarity-attraction effect”, “contact hypothesis”, and related perspectives
about close relationships and intergroup relationships. Furthermore, it is helpful to some
actual activities such as smoking abstinence. Although the concept of self-expansion was
first used to describe romantic relationships, it has since been expanded to encompass
relationships between consumers and brands (Patwardhan & Balasubramanian, 2011).
When interacting with a brand, one might expand the scope of their self-definition and adopt
fresh viewpoints (Reimann & Aron, 2009). According to a research by Reinman et al. (2012),
incorporation of the brand in the self can lead to self-expansion in intimate interactions
between brands and consumers, much like in human relationships. Customers assimilate
the resources, personalities, symbols, and social standing of their preferred brand into their
identities. (Reinman et al. 2012; Kerviler and Rodri-Guez 2019)

Brand loyalty is the emotionally-charged decision of a consumer for purchasing a particular

brand again and again. The consumer has the perception that the particular brand has the
qualities that will meet their expectations and identifies with the consumer at a personal level
(Ahsan Ali Shaw). Brand loyalty is also perception-based due to image and experience.

Jacoby and Chestnut proposed one of the most extensive and often recognized conceptual
definitions of brand loyalty (1978). Brand loyalty is defined as the biased, behavioral
response displayed over time by some decision-making unit with respect to one or more
alternative brands out of a group of such brands, and is a result of psychological
decision-making, evaluative processes, according to Jacoby.

Brand loyalty, according to David A. Aaker, is the level of attachment that a consumer has to
a brand. According to him, brand loyalty demonstrates how likely a consumer is to shift
brands when that brand changes the pricing or features of its products. According to Aaker,
consumer loyalty lies at the heart of a brand's equity. As a result, if buyers purchase based
on features, affordability, and convenience rather than brand name, there may be little equity
(Aaker, 1991).

Literature gap
According to the studies about the relationship between brand relevance, self - expansion
and brand loyalty, there is no concrete research about those 3 factors affecting each other
with the topic related to the green cosmetics industry. We do this research with the purpose
of understanding more about the association between 3 components influencing green
cosmetics and finding out the solution for brands to increase their profit through
understanding the insight of customers.

1. Research Design
We decided on running an online survey with several detailed and selective questions to
collect data for our research for more understanding about brand relevance and its
association with brand loyalty in the green cosmetics industry. The methodology used in this
research is a descriptive method.

The sampling method we use is Purposive Sampling. This is a type of non-probability

sampling technique. In other words, we only choose people who have characteristics that we
need to do the research. We want to do research about Green Cosmetics, so we only
choose the one who has knowledge about this topic. If they have no clue about this industry,
we will delete them from our sample. The reason for us to choose this method is because we
want to collect in-depth information from people who already have a background about
Green Cosmetics. Another reason is we have already had a certain amount of background
information about this topic so we know which one is best suited to help us answer the
research question. This sampling method allows us to get high sample quality, as a result
the information we gain for the research also has high quality.

We are going to conduct quantitative research by using an online survey of relevant

respondents with various kinds of questions linked directly to our desired topic. This
research will survey 400 youngsters who use green cosmetics. The survey will be uploaded
on social media networks, especially in beauty-related groups/fanpages, green cosmetics
using groups… which most of our target audiences follow.

SPSS is the tool we use to do the test for this research. SPSS shows a lot of benefits in
solving complicated data. In our research, there is a lot of data needed to analyze so that we
can base it on it to come to a conclusion. SPSS assists us to make that process much
easier. We do online surveys and SPSS do help to analyze the responses. With SPSS, we
are able to uncover powerful insights from responses to open-ended survey questions.
Another reason we use SPSS is We may easily produce a broad range of graphics from our
collected survey data such as density charts or radial boxplots. SPSS also offers data
management options that let researchers choose cases, produce derived data, and reshape
files. Additionally, SPSS provides data documentation, which enables us to keep a metadata
dictionary, which can connect to other data, show the meaning, the format or the usage of
data. Besides, we also use AMOS which is an extension of SPSS to drawing model and
check

In this research, ANOVA one way test and Post-hoc test are two types of tests used to test
the validity of hypotheses. This choice is made because the purpose of the research is to
test the interaction between the pairs of variables: brand relevance & brand loyalty;
self-expansion & brand loyalty. Theory shows that, when we want to find out if there is any
difference between two or more variables, we can use a one way ANOVA test. However, it is
worth noting that a limitation of this type of test is that it cannot show a difference, it can only
show if there is any difference. The research not only wants to show if there is any
difference, but it also wants to show how brand relevance and self-expansion affect brand
loyalty. That is also something that other research studies seem to be unable to conclude.
The proposed hypothesis is that brand relevance and self expansion are both positively
related to brand loyalty. To prove/reject these hypotheses, data were collected and ANOVA
was run on SPSS. If the hypothesis is rejected, the Post-hoc test will do the rest.

Data Analysis
1. Test for reliability using Cronbach's alpha
Cronbach's alpha calculates the internal reliability of a set of survey items.

Figure 1: Compute Cronbach's alpha using SPSS

 ● Factor 1: Self-expansion

Figure 2: Cronbach’s Alpha of independent variable “Self-expansion’’

Figure 3: Cronbach’s Alpha of observed variables “Self-expansion’’

According to the test results:

- Only observed variable “SEposquali” is less than 0.3, so we removed this variable.
- All other observed variables have a suitable Corrected Item-Total Correlation (> 0.3).
- Cronbach's Alpha coefficient = 0.915 > 0.6 meets the reliability requirements.
Factor 2: Brand Relevance

Figure 4: Cronbach’s Alpha of independent variable “Brand Relevance’’

Figure 5: Cronbach’s Alpha of observed variables “Brand Relevance”

According to the test results:

- All observed variables have a suitable Corrected Item-Total Correlation (> 0.3). Only
observed variable “BRSponsor” is less than 0.3, so we removed this variable.
- Cronbach's Alpha coefficient = 0.898 > 0.6 meets the reliability requirements.
 ● Factor 3: Brand Loyalty

Figure 6: Cronbach’s Alpha of dependent variable “Brand Loyalty’’

Figure 7: Cronbach’s Alpha of observed variables “Brand Loyalty’’

According to the test results:

- All observed variables have a suitable Corrected Item-Total Correlation (> 0.3).
- Cronbach's Alpha coefficient = 0.819 > 0.6 meets the reliability requirements.
2. Exploratory factor analysis (EFA)

Figure 8: Perform Exploratory factor analysis steps

● Exploratory factor analysis for independent variables

Figure 9: Rotated Component Matrix

Since all of the factor loading coefficients are more than 0.5 and no variable has ever
uploaded data to both factors at the same time with loading coefficients that are similar to
one another, the factors ensure convergent and discriminant validity when analyzing EFA.
Furthermore, the components are not mixed, the question belonging to one factor is not
mixed up with the question belonging to the other. Therefore, following factor analysis, these
independent factors are kept unchanged.

● Exploratory factor analysis for dependent variable “Brand Loyalty”

Figure 10: Component Matrix of dependent variable “Brand loyalty”

The variables BLnotavailwhere and BLreputegood are removed from the rotation matrix's
results because they each include two elements with a factor loading's difference less than
0.3. We rerun the component matrix after removing the two variables BLnotavailwhere and
BLreputegood. The result is below:

Figure 11: Component Matrix of dependent variable “Brand Loyalty” after removing variables
 3. Confirmatory Factor Analysis (CFA)

Figure 12: Measurement diagram in CFA

We'll consider the P value (p-value). This number shows whether or not the
observed variable demonstrates the latent variable's characteristics. The observed
variable is significant in the model if the p-value is less than 0.05. The AMOS 3-sign
*** is 0.000. We will go back to the CFA diagram and remove the observed variable
from the model if the observed variable is not significant in the model (p-value >
0.05), rerun a CFA analysis, and then reassess the observed variable's quality.
Regression Weights: (Group number 1 - Default model)

Estimate S.E. C.R. P Label

SElargerper <--- SE 1.027 .043 24.145 ***

SEincreknowl <--- SE .963 .038 25.101 ***

BRPerformance <--- BR .947 .048 19.531 ***

BRSociety <--- BR .940 .049 19.321 ***

BLtrust <--- BL .772 .082 9.463 ***

BLhaveused <--- BL .818 .089 9.164 ***

SEexpandsense <--- SE 1.000

SEgreaterawe <--- SE .947 .038 24.814 ***

SEaccomnewthin <--- SE 1.103 .053 20.924 ***

gs

SElearnnewthing <--- SE .915 .040 22.696 ***

s

BRReflect <--- BR 1.000

BRTobeme <--- BR .958 .053 17.957 ***

BRConnection <--- BR .990 .050 19.824 ***

BRMeaning <--- BR .962 .048 19.841 ***

BLexactlyexpect <--- BL 1.000

BLsalenot <--- BL 1.006 .088 11.465 ***

BLfavorite <--- BL .767 .085 9.035 ***

BLrcm <--- BL .741 .084 8.803 ***

Figure 13: Regression analysis results in CFA

According to the CFA results, all of the observed variables had P-values of 0.000, which
indicates that the observed variable is meaningful and can display the latent variable's
characteristics. The standardized weights all are within the permitted limits (> = 0.5).

Standardized Regression Weights: (Group number 1 - Default model)

 Estimate

SElargerper <--- SE .884

SEincreknowl <--- SE .897

BRPerformance <--- BR .853

BRSociety <--- BR .848

BLtrust <--- BL .598

BLhaveused <--- BL .579

SEexpandsense <--- SE .919

SEgreaterawe <--- SE .893

SEaccomnewthings <--- SE .834

SElearnnewthings <--- SE .863

BRReflect <--- BR .871

BRTobeme <--- BR .815

BRConnection <--- BR .860

BRMeaning <--- BR .860

BLexactlyexpect <--- BL .785

BLsalenot <--- BL .728

BLfavorite <--- BL .571

BLrcm <--- BL .557

Figure 14: Standardized Regression analysis results in CFA

The findings indicate that the variables have varied influences. First, "SEexpandsense" with
confidence level of 0.897, "SEincreknowl" with confidence level of 0.897, and "SEgreateawe"
with confidence level of 0.893 have the most influence. "BLrcm" with a confidence level of 0.557
is the least result. This outcome will be put to use in the AMOS software to create the data for
confirmatory factor analysis (CFA).
 Figure 15: Important factors to evaluate the model fit (CMIN/DF, GFI, CFI, TLI)

With p=000, this model's CMIN/DF value is 4.4695. Other indications (CFI = 0.889>0.8; GFI =
0.805>0.8) demonstrate that this model is consistent with the data. The RMSEA and PCLOSE
indices, however, failed to meet the requirements.

With a standard deviation under 0.05, the correlation coefficient between the components is less
than 1. Therefore, discriminant validity is achieved by the SE, BR, and BL measures. The
standardized weights all have statistical significance of 0.000 and are within the permitted limits
(> = 0.5). The variables utilized to calculate the aggregate and the total variance retrieved are
therefore >0.5, indicating that all of these components are reliable.

4. Structural equation modeling (SEM)

 Figure 16: Structural equation modeling diagram

Estimat S.E. C.R. P Label

e

BL <--- SE .403 .069 5.811 ***

BL <--- BR -.021 .074 -.281 .779

Figure 17: Regression Weight in SEM

Based on the SEM findings, it is obvious that the independent variable SE has a positive
effect on the dependent variable BL because its p-value is equal to 0.000 <0.5. Contrarily,
the independent variable BR's p-value is 0.779 > 0.5, indicating that it has no impact on the
dependent variable BL. Looking at the Estimate, we can explain that when SE goes up by 1,
BL goes up by 0.403.

Estimat
e

BL .147

Figure 18: Squared Multiple Correlations in SEM

The Squared Multiple Correlations of BL is 0.147, as shown in the table above, indicates that SE
has 14.7% effect on variation of BL. We do not take into account the extent of this variable's
influence because BR has no impact on BL.

From the findings above, we can conclude that:

● We reject H1: There is a positive relationship between brand relevance and brand
loyalty in the green cosmetics industry.
● We do not reject H2: There is a positive relationship between self-expansion and
brand loyalty in the green cosmetics industry.

Discussion
The research questions were to find out if there was a positive relationship between brand
relevance and self-expansion to brand loyalty. The results we discovered depend on the
Standardized Regression Weights table which shows that self-expansion has a positive
effect on brand loyalty, however, there is no influence of brand relevance on brand loyalty. It
can be seen that there are many green cosmetic brands that have nearly the same
categories or prices at some points so that brand relevance may not work at boosting brand
loyalty. People consider the brand “to be me” is the least choice when they use products of
that brand so that we can assume that brand relevance has almost no effects on brand
loyalty of the customers. On the other hand, customers feel that they are thankful for this
brand because the brands have improved several optimistic values to their sense of self and
followed by increasing the ability to accomplish new things when people use the green
cosmetic products. We can take on all findings and show that self-expansion theory has
great influence on brand loyalty in the green cosmetics industry.

Recommendation

From the above research and findings, we have come to some recommendations for the
green cosmetics industry to strengthen brand loyalty of customers. By finding out among the
self-expansion factors, brands help customers in expanding their sense of themselves
(SEexpandsense) and broaden their knowledge about everything (SEincreknowl) have the
most important effect on Brand Loyalty. It shows that all the green cosmetic brands should
focus more on the two factors to build up the relationship with customers. Target customers
of green cosmetic brands are young and middle age people, so their needs for exploring
themselves in particular and the world in general are quite big. Green cosmetic brands
should provide products that allow them to achieve certain achievements such as becoming
more beautiful or more confident will make customers feel valued and connected to the
brand.

Another recommendation is green cosmetics enterprises should still research more about
this factor because it may change in the future and people might change their behavior and
attitude when buying products. Our findings correspondingly suggest that customers
respond more favorably to brands that add positive qualities to their sense of self. From that
point, what a brand should do to engage the relationship with customers is to focus on this
main factor to reinforce brand loyalty.

Conclusion
Based on the result of this study, the positive effect of the self-expansion model on brand
loyalty is confirmed. This is an important conclusion that directly has an impact on the green
cosmetic industry. Through a self-expansion model, brands can have distinctive ways to
differentiate themselves from the others. The research design inevitably places constraints
on any investigation. It is also important to note that there were a lot more women than men
in our sample. We hypothesize that some males may use cosmetics less than females so
they choose not to complete our survey when we posted it on the Internet. To the extent that
there is a positive relationship between self-expansion and brand loyalty and there is no
connection between brand relevance and brand loyalty, understanding more and having a
deeper look into customers’ insight is always an urgent need for brands in this industry to
build more engagement and maintain brand loyalty for a long time.

References:
Aaker, D. A. (1991), Managing Brand Equity: Capitalizing on the Value of a Brand Name, NY:
The Free Press
Beauty Industry (11 June 2021) ‘What are green cosmetics? Why they are necessary for
skincare’, Ambuja,
https://www.ambujasolvex.com/blog/what-are-green-cosmetics-why-they-necessary-for-skinc
are/
Clement Thibaud (2020) ‘Brand Relevance: How to successfully connect with your
customer’, Brandingmag, accessed 29 June 2020.
https://www.brandingmag.com/2020/06/29/brand-relevance-how-to-successfully-connect-wit
h-your-customers/
Carol M Kopp (2022) ‘Brand Loyalty’, Investopedia, accessed 8 June 2022.
https://www.investopedia.com/terms/b/brand-loyalty.asp
Feng-Xiang J and Wei S (2011) ‘A Review of Self- Expansion Model’, Advance in
Psychological Science, 20 (1): 137-148.
Jacoby, J., and Chestnut, R. W. (1978), Brand Loyalty: Measurement and Management, NY:
John Wiley & Sons.