The role of brand engagement in self-concept toward brand love and brand advocacy

This research examines the impact of relevance between brand engagement in self-concept, brand love, and confidence toward brand advocacy and moderation by involvement and subjective knowledge. The population of this research is a generation of millennials in Indonesia with an age range of 18 to 28 who have purchased and used Converse shoes with a minimum of 6 (six) months. Research samples were taken 220 respondents with purposive sampling techniques and primary data obtained from online questionnaires using LimeSurvey's website. The data obtained is tested using Smart PLS 3.2.8 software. The results show that brand engagement in self-concept and brand love positively affects brand advocacy; brand love positively affects brand engagement in self-concept, and confidence positively affects brand love.


Background
The rapid development of today will indeed always trigger new trends in the present.These developments can emerge from all sorts of ideas, ranging from technology to how people look or fashion.Nowadays, style serves as a body cover and accessories alone and has become a necessity for communication to convey personal identity.This requirement is based on consumers' desire to increase prestige according to the symbol imposed (Nia and Lynne Zaichkowsky, 2000).Fashion products are not merely clothing or accessories worn, one of the fashion products that are rapidly growing and quickly follow the trend of the development era is shoe products.
First, shoes were created only limited to foot protectors and comfort.But nowadays, shoes provide significant benefits to support the wearer's appearance, trend, and fashion.There are many market demands and improvements in its use, making the designers and shoe companies constantly strive to present the shoes of the finest quality and design.Outside the country, in the country itself, but shoes have also become a necessity and a must-have item to support the appearance.This can be seen by the proliferation of shoe outlets that offer various brands.The majority of millennials in Indonesia have not only one pair of shoes.It can be said to be natural because it is so essential for shoes for appearance.Shoes are now crucial to everyone's lifestyle, including the Indonesian millennials.As good as any outfit worn, there would be something lacking if no shoes were accompanying.
Since the relationship can occur between the subject (customers, students, etc.) with objects (brands, firm, etc.) if it is reviewed in terms of concept, most of Indonesia's millennial customers connect the brand with oneself and tend to connect favorite brands as part of the consumer self-concept especially millennials in Indonesia.This relationship with self-concept is appropriate in learning how consumers have a difference in attitudes and behavior activities of consumers while they connect concepts themselves to their favorite brands, including Converse.Significantly since the fashions are constantly changing and fashion enthusiasts tend to get heavily involved with their favorite fashion brands to show themselves, it's called Brand Engagement in Self-Concept.The confidence and confidence of consumers, especially millennials in a fashion brand, will make customers, especially millennials, make a purchase, and customers will feel satisfied with the product being used.It can affect the brand love of a fashion product.
Brand Love is an emotional passion for a brand's customer satisfaction/consumer (customer satisfaction) (Carroll and Ahuvia, 2006).The love that occurs in a fashion brand can affect brand advocacy and a positive word of mouth for other consumers, especially the Indonesian millennials who love interaction and discussion about fashion brands with other fellow millennials.Brand love can belong to various product categories, fashion products, indulgent products, and other products.Also, the development of technology on one side can help the love of the brand by supporting a variety of consumer needs that make consumers can feel a sort of satisfaction that can ultimately lead to a feeling of love for the brand, and social media and website trading support this on the favorite fashion brand Millennials Indonesia that creates indirect brand advocacy on other millennials.
One of the shoe products that are widely used by generations of millennials in Indonesia is Converse.This shoe brand originated in the United States.With Converse brand shoes in Indonesia that can be assessed quite an extensive existence, it is proven by the presence of four (4) production and distribution of shoes with Converse brand that operates in areas such as Bogor, Purwakarta, Tangerang, and Sukabumi.
Converse shoe brands have high sales value and top fans on social networking sites, and the majority are millennials.This concept focuses on brand engagement in the self-concept of Converse brand users and, if associated with Brand love and brand advocacy, will produce stunning results.Suppose a customer has a feeling or emotion of love and can be more than just liking a Converse brand.
In that case, it can be said the loyalty level of a brand and a positive word of mouth (brand advocacy) toward the Converse brand obtained will be greater (Carroll and Ahuvia, 2006).

Literature review
Millennials are a unique group that is young, affluent, educated, tech-savvy, and manifest positive social behavior.They are confident and more fashion-oriented.Millennials search and learn about the new fashion trends and related information through magazines and online media (Geraci and Nagy, 2004).They are the primary group participating in online shopping.Hence marketers find innovative digital media activities to engage them.
Involvement between consumers and brands has become a very effective tool so that it can improve customer satisfaction and loyalty, especially on millennials since relationships can occur between the subject (customers, students, etc.) with objects (brands, firms, etc.).If it is reviewed in terms of concept, most consumers of Millennials connect the brand with themselves.Millennials tend to consider brands that are always worn as part of themself (Hollebeek, 2011).

Brand Engagement in Self-Concept
The relationship between brand engagement in self-concept and brand loyalty proves that consumers always involve a brand in their self-concept by buying and advising a brand to another consumer or another potential consumer (brand advocacy) (Sprott et al., 2009).This research stated WOM is positive (brand advocacy) due to brand engagement in self-concept because brand engagement in self-concept indirectly determines consumer behavior and purchasing products that include providing recommendations, commenting and love about the brand on social media (Hollebeek et al., 2014).The relationship between BESC and brand loyalty, as evidence that customers who unite a brand in their self-concept by making a purchase and providing advice regarding a brand that is imposed on other prospective consumers including through social media or known as brand advocacy (Sprott et al., 2009).The study finds it appropriate to confine the outcome of engagement to positive brand advocacy through WOM: H1. Brand engagement in self-concept positively affects brand advocacy.

Brand Love
Brand Love is characterized by a passion for a brand, a brand bond, a brand-building evaluation, a positive feeling of feedback on a brand, and a feeling of love for the brand (Carroll and Ahuvia, 2006).
Brand love relates to self-concept (Albert et al., 2008).Consumers show a love of objects that require significant energy and time (Ahuvia, 2005).And fashion clothing brand is a brand that offers an image and a consumer personality (Khare, 2014).
Consumers use a longer time and more energy to shop for fashion products because they communicate their personal and image.On the other hand, the positive influence of Brand Love is also brand engagement in the self-concept is a result of a brand that is beyond satisfaction, such as liking, commenting, and giving recommendations.Suggestions include recommendations and positive talks about the brand, and positive WOM about the brand can be said to be advocating (Du et al., 2007).There is a tremendously positive relationship between the love of a brand and the Word of Mouth positively (Carroll and Ahuvia, 2006).Hence, we can hypothesize that brand love positively influences brand advocacy.We also seek to understand consumer behavioral outcomes (like WOM) through their engagement with the brands: H2. Brand love positively affects brand engagement in self-concept.H3. Brand love positively affects brand advocacy.

Self-Confidence
Studies show that confidence is a fundamental element of purchase intent, brand loyalty, commitment, and critical components of the relationship between consumers and brand Love (Kim and Jones, 2009).Self-confidence is the willingness of consumers to believe in the brand's expectations.Thus, trusted brands can generate more purchases, increase market share and loyalty, and love of the brand (Ruparelia et al., 2010).Consumers who believe in beloved brands are never disappointed and have permanent dissatisfaction when using the brand (Albert et al., 2008).The hypotheses that can be formulated and proposed are: H4.Self-confidence positively affects brand love.

Brand Advocacy as an Outcome
Brand advocacy is an act of trying out new products from the same brand, talking about the brand, and being willing to receive any minor mistakes from the brand (Du et al., 2007).Consumers are willing to rate the brand and recommend it to others and act as defenders of the brand.The role of advocacy is not only recurring purchases.This study considers Word Of Mouth as positive as the essential dimension of advocacy, which is a form of behavioral relationship demonstrating.Brand advocacy has been stated to have a direct effect on repurchase.When consumers are satisfied with a product and brand, it is expected to recommend these products and brands to other friends and intend to make a purchase again.So too negative word of mouth can happen to unsatisfied consumers.An example is when consumers feel unhappy or unsatisfied for a reason, it will result in complaints and the dissemination of negative news (Ladhari, 2007).Therefore, a company should be able to keep its customers from spreading negative word of mouth.Consumers who have a good product experience will reflect positive word of mouth so that brand advocacy can be conveyed well (Kitapci et al., 2014).

Involvement as a Moderator
Involvement has a potential for brand engagement on consumers (Hollebeek et al., 2014).In other contexts, participation in consumer engagement as a moderator is essential, especially when consumers follow various online-based communities related to brands to share their product-related experiences and product engagement as a potential predictor of consumer behavior in search of the benefits of the brand (Aksoy et al., 2013a).Consumers can often say that spending a lot of attention and ways to make decisions on the condition of involvement is relatively high, which will eventually pose a lot of attention to the self-concept of the information of a brand (Xue, 2008).
In the same context, a study provides support for the relationship between BESC and consumer involvement (Flynn and Goldsmith, 1999).Looking at increasing interest rates in consumer-brand literature to incorporate involvement in understanding consumer-brand relationships, this research intends to learn the role of involvement in enhancing the relationship between brand love and brand engagement in self-concept.Looking at the increasing level of interest in the consumer brand literature for including involvement construct to understand the consumer-brand relationship, the present study intends to study the role of the participation in enhancing the relationship between brand love and BESC: H5.Involvement acts moderately the relationship between brand love and brand engagement in selfconcept, i.e., the higher the level of the participation, the higher will be the effect of brand love on brand engagement in self-concept.

Subjective Knowledge as a Moderator
There is a significant moderation influence of product class knowledge on the matter, and the WOM deduced from the previous findings that an exceptionally subjective form of expertise creates a motivation response to the various traits of the brand as well as what fashion products concerning the WOM and the innovators explain that a more innovative inclined skating will talk more positive than negative (Sohn and John, 2005).Slides that follow the sustainability of a brand can potentially advocate positively about the fashion brand, and the same will be altered by their subjective knowledge.This happens more when the brand is a fashion outfit because there is a sought-after by a colleague of the innovator who will be high (Lelo de Larrea et al., 2019).Therefore, we hypothesize that the level of subjective knowledge would alter the direction and strength of the relationship between BESC and brand advocacy.All six hypotheses are shown in Figure 1: H6.Subjective brand knowledge acts the relationship between brand engagement in self-concept and brand advocacy moderately, i.e., the higher the level of expertise, the higher will be the effect of BESC on brand advocacy.

Method, sample, and data analysis
The research conducted is research with a type of quantitative study aimed at testing causality relationships researching samples and a variety of specific populations with hypothesis testing.The approach used in this research with the deductive method as a research approach is generally then concluded with unique properties (Burney and Saleem, 2008).The population in this study is the consumer who uses and buys Converse products.
The data collection method is a survey with the spread of questionnaires as research instruments.The survey is selected in the form of a questionnaire, and the selected sample has been chosen, which contains the following statements into data and processed in this research.In addition to using these methods, researchers plan to use questionnaires deployed through a LimeSurvey-based web network to save time, wider reach, and the charging time of respondents that the system can know.The distribution of the questionnaire is done on various social media such as WhatsApp, Line, Instagram, Facebook, and mixed other social media.The survey method is efficient and has a high level of generability.Besides, researchers can collect a variety of quantitative data for various types of research questions (Sekaran and Bougie, 2017).

Demographics and Sample Size
The method used in this research is purposive sampling with a non-random sampling technique.Purposive sampling is a non-random sampling technique that requires researchers to perform sampling and sampling by assigning a distinctive characteristic or criteria that is appropriate to the purpose of the study to provide an answer to the issues researched and can provide relevant responses to this research (Sekaran and Bougie, 2017).
The target respondent is the consumer of millennials in Indonesia with criteria of 18 years to 28 years and has been using Converse shoes for men.Six months.The determination of the appropriate number of samples ranged from the scope of 100-200 respondents (Hair, 2017).The example used in this study is 220 respondents by conducting questionnaire data collection method in the form of a list of questions to respondents.The limitation of this research is that researchers make restrictions according to respondents' criteria, namely the age of respondents aged 18 years -28 years and have been using the Converse brand shoes for men.Six months of usage.
The results of a descriptive analysis of respondents showed that the accumulated 220 respondents were millennials for Converse shoe users in Indonesia who contributed to the study.The majority of male-sex respondents with a percentage of 51%, the majority of 21-24-year-old respondents with a 68% percentage, the majority of respondents domiciled in Central Java and DIY with a ratio of 53%, the majority of the job the respondent studied was a student/I with a ratio of 72%.The majority of respondents used Converse shoes for more than three (3) years, with a percentage of 42%.

Results
The data analysis techniques in this study used the analysis method of SEM-PLS (Partial Least Square Structural Equation Modeling).SEM-PLS is a method to estimate the model's path (complex) with variable latent and its relationship.SEM-PLS allows authors to calculate a very complex model, especially when the prediction is the goal of the analysis (Hair, 2017).The validity test aims to measure the good or legitimate presence of a questionnaire.A questionnaire is good if the question indicator on the questionnaire can reveal something that will be measured by the questionnaire (Hair, 2017).The reliability test aims to determine the level of consistency in various instruments that measure the concept and stability of a measuring instrument in measuring a symptom or occurrence.The reliability of measurement reflects whether the measurement can be error-free, resulting in consistent measurements of different conditions (Sekaran and Bougie, 2017).
The table above shows the results of the validity test in 220 respondents.Of the total of 30 question items (indicators), There are three (3) Items of Invalid questions (indicators) that are indicators BLOV1, BLOV3, and BADV1 because they have an outer loadings value of < 0.7 and subsequent retesting by excluding the indicators that are declared to be invalid.The result of the loading factor of 27 items has been extracted well, and overall with an outer loadings value of > 0.7, therefore the variable used can be deduced as a valid variable and can perform reliability testing.All of the research variables are reliable if reviewed from Cronbach's alpha and composite reliability > 0.70 and AVE value > 0.50.This indicates that the processed data is valid and reliable so that this data can be used for the next testing phase.

Discriminant Validity (HTMT) Measurement
Discriminant Validity (HTMT) is the extent to which construction is entirely different from the other construction (Hair, 2017).According to the table, it shows that the value of linear Validity in subjective knowledge (SKNO) and involvement (INVL) is 0.204.Self-confidence (CONF) and brand love are generating a value of 0.893.The relationship values of other variables are among them.Discriminant validity test results can be concluded that the entire research variable is declared correlated between variables as acceptable values if reviewed in a test result indicating a value of less than one (< 1).In other words, all of these discriminant validity outcomes are supported and can be used for subsequent testing phases.The role of goodness-of-fit measurement is to assess the quality of the model estimated by comparing the actual value of the dependent variable to the value predicted by the estimated model.The criteria of the GoF value are 0.1 (GoF small), 0.25 (GoF moderate), and 0.36 (GoF Large) (Hair, 2017).If the value of GOF > 0.36, then considered the suitability model is good.The value of the Goodness of Fit (GoF) calculation can be seen as the result with the formula:  = √  × √  .The GoF value obtained in the above calculation is 0,596, which belongs to the large category (GoF Large).From those results, the model of this study can be stated already has good goodness of fit and the research model already has good value and the ability to explain empirical data.

Discussions
Based on data analysis results, H1 is significantly supporting the positive effect of BESC on brand advocacy; it is supported by test results indicating p-values = 0.007, t-statistics = 2.727, and the original sample = 0.232.This affects following previous research by Hollebeek et al. 2014, that WOM is positive (brand advocacy) as a result of BESC and indirectly determines consumer behavior beyond purchase in terms of recommending, commenting, liking about brands on social media.
H2 is also accepted, proving that brand love positively affects BESC.It is supported by the test results indicating the p-values = 0.000, t-statistics = 8.223, and the original sample = 0.539.This relationship follows previous studies stating that consumers spend more time and energy choosing fashion clothes because they communicate their personality and image (Du et al., 2007).There is a direct positive effect of brand love on brand advocacy since H3 is also significant at p-values = 0.000, t-statistics = 4.210, and original sample = 0.343.It is supported by a significant positive relationship between the love of the brand and the positive WOM (Carroll and Ahuvia, 2006).
H4 self-confidence is positively related to brand love.It is supported by the test results indicating the value of p-values = 0.000, t-statistics = 26.113,and the original samples = 0.767.This relationship follows previous research, which suggests that confidence is the fundamental element of the intent of purchase, brand loyalty, commitment, and key components of the relationship between consumers and brand Love (Kim and Jones, 2009).H5 involvement negatively moderates the relationship between the love of brand engagement in self-concept.It is supported by test results indicating p-values = 0.162, t-statistics = 1.400, and original sample value = 0.074.Involvement is when consumers engage in online brand communities to share experiences related to their loved brands (Aksoy et al., 2013), in which case it is not present on millennials in Indonesia that very rarely found the Converse community can be present in each city and represent the involvement of the consumer millennials in Indonesia.
H6 subjective knowledge positively moderates the relationship between brand engagement in self-concept and brand advocacy, it is supported by the results of an interaction moderation test that shows the p-values = 0.005 value, t-statistics = 2.843, and the original sample = 0.145.This relationship corresponds to previous research, which suggests that there is a significant moderation influence of the subjective knowledge consumers forming their motivational responses to the features of product information concerning WOM and the innovative consumers tend to talk more often positively than the negative about the product (Sohn and John, 2005).Indirect effect analysis shows that self-confidence has indirect effects, which are significant with brand advocacy through a mediator of brand love.The same also shows that confidence has indirect effects that are significant with the brand engagement in self-concept through the mediator of brand love.A significant p-value of both is 0.000 (< 0.05).The indirect effect test results are concluded in the following table:

Conclusion, Limitation, and Future Research Directions
Based on the results of research conducted can explain the phenomenon of consumer behavior of millennials of Converse shoe users in Indonesia.The results of a descriptive analysis of respondents showed that accumulated with 220 respondents.The majority of male-sex respondents with a percentage of 51%, the age of the majority respondent aged 21-24 years with a 68% percentage, the majority of respondents domiciled in Central Java and DIY with a ratio of 53%, the majority of job respondents were students with a ratio of 72%, and the majority of respondents used Converse shoes for more than three (3) years with a percentage of 42%.
Brand engagement in self-concept and brand love positively affects brand advocacy, indicating that the consumer behavior of millennials in Indonesia who love a brand, especially Converse, will directly or indirectly advocate for brand use towards other prospective consumers.Brand love positively affects brand engagement in self-concept.The consumer behavior of millennials in Indonesia who loves a brand, especially the brand of Converse shoes, has a close relationship or arrangement with the brand on their personal.
Confidence positively affects brand love; it shows that the trust of consumers of millennials who use Converse products indirectly love the brand used with confidence when using the product.
Involvement does not moderate the relationship between brand love and brand engagement in selfconcept, involvement, i.e., when consumers are involved in the online brand community to share experiences related to their loved brands (Aksoy et al., 2013), in this case not applicable to millennials in Indonesia who are very rarely various Converse communities can exist in each city and represent the involvement of millennials in Indonesia.
Subjective knowledge moderates the relationship between brand engagement in self-concept and brand advocacy.It is based on consumer knowledge of fashion products and converse products used so that expertise can strengthen the ties between BESC and brand advocacy, which is the higher the level of knowledge, the higher the influence of brand engagement in self-concept to brand advocacy.
A significant positive relationship indicates that there is a causality relationship between the variables studied and positive.It can be concluded that the study has five (5) hypotheses of significant relationships during research, and one (1) rejected the hypothesis that is about the involvement that does not moderate the relationship between brand love and brand engagement and the self-concept.
Further research is expected to be able to examine respondents at a more comprehensive age range, and the job can be more varied, so it can be known whether the segment difference will affect the outcome of the research.Further research is expected to further add to and extend to other variables included in the brand engagement criteria in self-concept, Brand love, and brand advocacy on personal consumer products related to fashion and expansion of segments of various generations besides expected subsequent research can use other methods other than quantitative methods, such as qualitative and observation methods.Future research is expected to research with the development of research objects, both fashion and non-fashion products.
Fashion is, multiplying significantly in the national and international shoe industry; this is evidenced by the growing number of brand new shoes that are present in the market both online and present with the opening of new outlets.The results of this research are expected to be a reference for shoe companies or other fashion products always to do the development of a new brand and shoe products that prioritize the quality, the company's reputation so that consumers can positively love and arise emotions to the brand and the product.Shoe companies and other fashion products can establish and present brands of brand communities in every city in Indonesia so that the consumer involvement of millennials can increase.The results of this research are expected to be a reference to academics and people who have an interest in marketing that there is a relationship between confidence, Brand love, brand engagement in self-concept, involvement, and subjective knowledge on consumer behavior to brand advocacy a fashion brand, especially Converse shoes.

Figure
Figure 1.Conceptual Model

Table 1 .
Validity test result

Table 2 .
Validity test results II

Table 3 .
Reliability test result

Table 5 .
Goodness of Fit Results

Table 6 .
Direct Effect, Indirect Effect, and Total Effect

Table 7 .
Evidence for Mediation