The role of brand love in mediating the effect of intensive distribution and social media promotion on brand loyalty and e-WOM

This study aims to examine and explain the role of brand love in mediating the effect of intensive distribution and promotion of social media on brand loyalty and e-WOM. The population of this research are customers of minimarket franchise businesses in Bali. The sample size used was 200 people with purposive sampling method. The analytical technique used is Path Analysis using SEM-PLS. The results showed that the intensive distribution had a positive and significant effect on brand loyalty and brand love. Brand love is able to mediate the effect of intensive distribution on customer loyalty, brand and e-WOM and brand love are able to mediate the effect of social media promotion on brand loyalty and e-WOM. Therefore, it is important for the minimarket franchise business to pay more attention to intensive distribution and vigorously promote social media through various choices of the latest social media platforms.


Introduction
Brand loyalty is an attitude shown by customers to a particular brand even though in making a purchasing decision for a product, there is an offer from another brand (Akoglu & Ozbek, 2021;Santos & Schlesinger, 2021;Safeer et al., 2021;Nyamekye et al., 2021;Guan et al., 2021). Brand loyalty is very important for companies because with high brand loyalty, it means that the level of consumer loyalty to the brand is very high and has an impact on increasing product sales and ultimately improving company performance. Therefore, every company strives to continue to increase loyalty to its brand. One of the businesses, namely the minimarket franchise business in Bali, even though there is still a COVID-19 pandemic, the phenomenon is always making a lot of efforts to increase its brand loyalty. Many previous researchers have conducted research on brand loyalty, including: Amoako et al. (2021); Amegbe et al. (2021); and Kwon et al. (2021). In addition to brand loyalty, the minimarket franchise business in Bali also hopes that communication about the minimarket franchise brand will be positive. Today's word of mouth communication through electronic media is important and gets attention because this positive information can increase consumer intention to buy the product and have an impact on sales turnover. Marketing communication through online media is known as e-WOM. A positive E-WOM can also indicate that the company's marketing performance is improving. So, brand loyalty and e-WOM are marketing performances to be achieved by companies including companies engaged in the minimarket franchise business in Bali. The results of previous studies indicate that brand loyalty and e-WOM can be built through increased distribution intensive, and there are other researchers who state that brand loyalty and e-WOM can be increased by promoting social media; and there are other researchers who state that brand loyalty and e-WOM can be built through consumer satisfaction (Ong et al., 2017;Espinosa et al., 2018;Kataria and Saini, 2020), through brand image (Soni, 2021) and through brand love (Le, 2021;Ghorbanzadeh & Rahehagh, 2021). Among the determinants of brand loyalty, according to the results of a preliminary survey conducted on 30 minimarket franchise business customers in Bali, it was shown that the intensive distribution variable greatly determines brand loyalty and e-WOM. When viewed from previous research, there have been several researchers including: examining the effect of intensive distribution on brand loyalty, (Yo et al., 2000;Al-Dmour et al., 2013;Effany et al., 2018). The results of research by Langga et al. (2021) showed that intensive distribution had a positive and significant effect on brand loyalty. Similar results were also shown by the results of previous studies, namely Amini et al. (2012), that the higher the intensive distribution, the brand loyalty will increase, but there are still researchers who show the opposite result, namely the intensive distribution has no significant effect on brand loyalty (Kazemi et al., 2013). From the research results of several researchers, it can be said that there is still a research gap regarding the effect of intensive distribution on brand loyalty. Likewise, intensive distribution is able to build positive e-WOM (Langga et al., 2021), but there are also other researchers, namely Kim and Hyun (2011), who state that intensive distribution is not able to build positive WOM.
In addition to intensive distribution, there are other variables that determine brand loyalty and e-WOM, namely social media promotion. Social media promotion has indeed become a trend of promotional media used today because of its relatively low costs and it is able to reach wider market share (Sheng et al., 2020). Social media promotion used by companies is indeed able to improve business performance (Yasa et al., 2020;Yasa et al., 2021); increase brand loyalty (Ismail, 2017;Kwon et al., 2021) and e-WOM (Bu et al., 2021). Research conducted by Machado et al. (2020) and Hazzam (2021) also found consistent results, namely through social media promotions, such as using the Instagram or Facebook platforms to increase brand loyalty; Likewise, there are those who use Facebook to market their products so that brand loyalty also increases. In addition, social media promotion is also able to build positive e-WOM (Ramanathan et al., 2017). Of the previous studies on the effect of social media promotion on brand loyalty and e-WOM, there are still some that show that social media promotion is not able to increase brand loyalty (Huang et al., 2014) and increase positive e-WOM (Kazemi et al., 2013). Because of the four research gaps, it is necessary to add a mediating variable brand love. The reason for using brand love as a mediating variable is because: if intensive distribution is carried out and increased by businesspeople, it can increase brand love. Consumers feel increasingly happy with the brand because they feel that they get a lot of convenience to buy the product. This is revealed in the results of research from Dornyei (2020), which states that intensive distribution can increase brand love. Likewise, by promoting social media, consumers get a lot of information they need before making a buying decision. This also makes his heart even more in love with the brand. Results like this were revealed in research conducted by Kim and Kim (2018); Raji et al. (2019); Amaro et al. (2020);and Hamzah et al. (2021), which states that social media promotion can increase brand love. Furthermore, brand love makes consumers loyal to the company's brand. (Huang, 2017;Fernandes & Inverneiro, 2021;Santos & Schlesinger, 2021) and increasing brand love can build positive e-WOM (Loureiro et al. (2017). The more love for one brand makes customers more loyal and e-WOM is more positive. Based on the background of the existing problems, this study develops a new conceptual model based on existing gap research by adding a mediating variable brand love, so the purpose of this study is to examine and explain the role of brand love in mediating the effect of intensive distribution and social media promotion on brand loyalty and e-WOM.

Intensive Distribution
Intensive distribution is a company's way to make it easy for consumers to get the products they buy. Intensive distribution reflects the company's marketing strategy that leads to using an intensive distribution strategy, namely, to create as many distribution sites as possible (Yu et al., 2014;Abushaikha, 2018;Lin et al., 2020). Intensive distribution measurement refers to the measurements made by Langga et al. (2021), where the distribution of these products is spread everywhere, easy to obtain, spreads to the outskirts of the city, and provides services that do not know time.

Social Media Promotion
According to Yasa et al. (2020), social media promotion is a promotion that is very widely adopted by companies today because this promotion requires low costs and can be adopted by all types of businesses from large-scale companies to smallscale companies. Social media promotions that are usually used can take the form or platform, including: Instagram, WhatsApp, and Facebook (Jose, 2018;Bellaaj, 2021). That is the social media platform that is usually adopted by businesses in the retail sector.

Brand Love
According to Khan et al. (2021) and Safeer et al. (2021a), brand love is indicated by the feelings of consumers who feel happy about a brand. The brand is already in his heart so that every time he makes a purchase, other brands are out of the question and customers will always feel comfortable to shop at that specific place . Brand love from consumers can build higher levels of brand loyalty and more positive WOM .

Brand Loyalty
Brand loyalty shows the commitment of a consumer not to switch to other brands (Kim and Lee, 2020. These consumers always remember the brand and decide to keep buying the brand of a product and keep buying at the brand of a place to shop. Brand loyalty has a positive impact on the company, because if consumers are always loyal to certain brands, then the sales turnover of the product or business will certainly increase and provide higher profits (Rubio et al., 2015;Nyamekye et al., 2021;Le, 2021). behavior of customers who do not want to switch to other brands; always increase future purchases; and always make the brand the first-choice brand in making future purchasing decisions (Ramanathan et al., 2017).

E-WOM
e-WOM is communication by word of mouth of consumers through online media (Li et al., 2021). E-WOM is a place for consumers to provide comments about the brand of a product or the brand of a place to shop. Comments from consumers are usually information for potential consumers before making a buying decision. If comments from previous consumers are positive, this is certainly good information influencing the choice of future consumers (Mainolfi & Vergura, 2021). Therefore, every business always tries to create conditions so that its customers are willing to give good and positive reviews about their products or place of business so that they also have a positive impact on other target markets (Langga et al., 2021).

Research Concept Framework
Existing businesspeople, including minimarket franchise businesses, want the brand loyalty of their business brands to be stronger. The steps taken are to increase brand loyalty and positive e-WOM. Realizing this goal by building higher brand love. Likewise, to have high brand love, the way is to build intensive distribution and carry out social media promotions.

Intensive distribution
Brand loyalty

Brand Love
Social media promotion e-WOM

Research Hypothesis
Based on the existing conceptual framework, the research hypotheses that can be formulated are as follows.

Research method
This study uses a quantitative approach, which analyzes and examines the causal relationship between intensive distribution variables, social media promotion, brand love, brand loyalty, and e-WOM. This research was conducted on customers of minimarket franchises in Bali. Questionnaires in the form of google forms were distributed to several WhatsApp groups to obtain the required data. The data collected by as many as 30 respondents were tested for validity and reliability, and the results are presented in Table 1. Table 1 shows that all variables are valid because the correlation value is above 0.30 and reliable because Cronbach's Alpha value is above 0.6. Furthermore, data collection was continued by distributing questionnaires in google form format to several WhatApp groups and collected as many as 200 respondents. This number has exceeded the minimum targeted sample size of 100 respondents (5 × 20 indicators). A sample of 200 respondents, then analyzed using analytical tools, namely: Path Analysis with the SEM-PLS approach.

Characteristics of respondents
Profiles of 200 respondents are presented in general with several characteristics including gender, age, last education, occupation and monthly income. The characteristics of the respondents in this study can be described as follows. There are more female respondents than male respondents, namely 138 women (69.0%) and 62 male respondents (31.0%). The age range of 18-28 years dominated filling out the questionnaire by 177 people (88.5%). Respondents with high school education dominate, as many as 143 people (71.5%). Respondents in this study were dominated by respondents with jobs as college students and students (other groups), as many as 141 people (70.5%). The grouping of respondents based on monthly income obtained data that the dominant respondents were those with income in the income range, Rp. 2 -5 million, as many as 174 people (87.0%) of the total 200 respondents. Characteristics of respondents are presented in Fig. 2

PLS SEM Analysis Results
This study uses a two-stage approach to measuring the model before it is used for hypothesis testing, aiming to verify the validity and reliability of a research model. First, by analyzing the convergent validity, then by analyzing the discriminant validity.

Convergent Validity
The outer model test is carried out to ensure the research indicators are suitable for use as their role in measuring research variables, so to see whether a model is valid to be the basis for research, there are three criteria that must be met, namely: (1) all loading indicators must be above 0.65 (2) Composite Reliability (CR) must be above 0.8, and (3) Average Variance Extracted (AVE) for each construct must exceed 0.5. Based on Table 2, it shows that all outer loading indicators have values above 0.65 with a range between 0.808 to 0.941 meaning they are at the recommendation limit, then the Composite Reliability (CR) value is in the range between 0.869 to 0.956, all of which are above 0.8, meaning that all constructs formed has good consistency as a research model, the third is the Average Variance Extracted (AVE) value where all of them are valued above 0.5, namely with a range from 0.688 to 0.871 so that it can be concluded that the research model in this study has good validity.

Discriminant Validity
To evaluate discriminant validity, a research model is suggested to ensure that the root value of the Average Variance Extracted (√AVE) of a latent variable must be larger. Discriminant validity is considered good if the root value of AVE (√AVE) in Table 3 is greater than 0.5. The research model proposed in this study can be considered good, where the smallest AVE value is 0.829

Inner Model test
Structural models focus on hypothesized relationships or pathways between latent variables. The results of the inner model test can be seen in Fig. 2.

Coefficient of Determination (R 2 )
In this study, bootstrap will be carried out which will produce two measurements of the structural model, namely: the value of t (t-test) and R2 which will be interpreted the same as multiple regression analysis in general. The predictive power of a research model can be seen by looking at the R2 value generated by the bootstrapping process, in Table 4. The R 2 value for each exogenous variable contained in the model will be presented.

Fig. 3. Structural Model
The structural model was evaluated using R-square for the dependent construct and t-test and the significance of the coefficients of structural path parameters. Based on Table 3, it can be explained that the highest R 2 value is found in the brand loyalty variable of 0.524 which means that as much as 52.4% of the brand loyalty variable can be explained by the constructs contained in the model, namely intensive distribution, social media promotion, and brand love, while the lowest value is found the e-WOM variable with a value of 0.470 which means that 47.0% of the e-WOM variable can be explained by the constructs that affect these variables, namely intensive distribution, social media promotion, and brand love. From the examination of the R2 value, it can be concluded that in general the predictive ability of this research model is quite good, seen from all variables that have an R2 value of almost and above 50%

Hypothesis testing Direct Effect Test
The significance of the estimated parameters provides very useful information about the relationship between the research variables. The basis used in testing the hypothesis is the value contained in the output path coefficients which is presented in Table 4. Hypothesis testing is done by using t-statistics and looking at the p-value. If the p-value 0.05 then the hypothesis is accepted. Based on Table 5, it can be explained that the intensive distribution of brand loyalty has a t-statistic value of 2.808 with a p-value of 0.005 0.05, so the hypothesis is accepted. This means that the higher the intensive distribution, the higher the brand loyalty to the minimarket franchise business brand in Bali. The intensive distribution of e-WOM has a t-statistic value of 2.765 with a p-value of 0.006 0.05, so the hypothesis is accepted. This means that the higher the intensive distribution, the higher the positive e-WOM. The intensive distribution of brand love has a t-statistic value of 2,469 with a p-value of 0.014 0.05, so the hypothesis is accepted. This means that the higher the intensive distribution, the higher the brand love for minimarket franchise brands in Bali. Furthermore, social media promotion also affects brand loyalty with a t-statistic value of 2.132 with a p-value of 0.033 0.05, so the hypothesis is accepted; Social media promotion also affects e-WOM with a tstatistic value of 2.293 with a p-value of 0.022 0.05, so the hypothesis is accepted. Likewise, social media promotion has an effect on brand love with a t-statistic value of 8.634 and with a p-value of 0.000 0.005, the hypothesis is accepted. Furthermore, brand love has an effect on brand loyalty with a t-statistic value of 6371 with a p-value of 0.000 0.05, then the hypothesis is also accepted. Brand love has an effect on e-WOM with a t-statistic value of 5.735 with a p-value of 0.000 <0.05, then the hypothesis is accepted .

Indirect Effect Test (Mediation Test)
Test the mediating role of brand love on the effect of intensive distribution and social media promotion on brand loyalty and e-WOM by examining the indirect effects which are the output of Smart PLS as presented in Table 5. Based on Table 5, it can be explained that the t-statistic value is greater than the value of t-table (2.227 > 1.96), then brand love significantly mediates the intensive distribution of brand loyalty. Likewise, it can be explained that the t-statistic value is greater than the t-table value (2.203> 1.96), then brand love significantly mediates the intensive distribution of e-WOM. Furthermore, brand love also significantly mediates the effect of social media promotion on brand loyalty with a t-table value of 5.155 > 1.96; and brand love significantly mediates the effect of social media promotion on e-WOM with a t-statistic value of 4.662 > 1.96.

Effect of Intensive Distribution on Brand Loyalty
Based on the analysis of the effect of intensive distribution on brand loyalty, the beta coefficient value is 0.174 with a significance level of 0.005 0.05, which means Ho is rejected and H1 is accepted. These results mean, the intensive distribution variable has a positive and significant effect on brand loyalty. This means that, the higher the intensive distribution of the minimarket franchise business in Bali, which is shown by the easy to find minimarket franchise business, it is everywhere, reaches the outskirts of the city, the opening hours are longer, then it can increase brand loyalty. The results of this study simultaneously strengthen the results of previous research conducted by Al-Dmour et al (2013), which states that intensive distribution has an important role in building brand loyalty. Similar results have also been obtained previously by Lin et al. (2020), which states that intensive distribution makes customers more loyal. The results of this study are also strengthened by the results of research from Langga et al. (2021), who found that the intensive distribution variable had a positive and significant effect on the brand loyalty variable. It can be concluded that intensive distribution has a positive and significant influence on brand loyalty of minimarket franchise businesses in Bali.

Effect of Intensive Distribution on e-WOM
Based on the results of the analysis of the effect of intensive distribution on e-WOM, the beta coefficient value is 0.195 with a significance level of 0.006 0.05, which means Ho is rejected and H1 is accepted. These results mean that the intensive distribution variable has a positive and significant effect on e-WOM. So, the higher the intensive distribution, which is shown by the convenience of minimarket franchises, they are everywhere, even in the outskirts of the city, the opening hours are longer, the brand loyalty will increase.
The results of this study at the same time strengthen the results of previous studies conducted by Yu et al. (2014) which states that the intensive distribution variable has a positive and significant effect on e-WOM. The results of this study are strengthened by the research of Abushaikha (2018), which found the results of the intensive distribution variable having a positive and significant effect on e-WOM so that it can be concluded that the intensive distribution has a positive and significant effect on e-WOM, with the higher the intensive distribution, the e-WOM. Positive WOM about minimarket franchise businesses (such as Alfamart, Indomaret, and Circle K) in Bali is increasing.

Effect of Intensive Distribution on Brand Love
Based on the analysis of the effect of intensive distribution on brand love, the beta coefficient value is 0.243 with a significance level of 0.014 0.05, which means Ho is rejected and H1 is accepted. These results mean, intensive distribution has a positive and significant effect on brand love. This means, the higher the intensive distribution indicated by the variable indicator, the minimarket franchise business is easy to find, everywhere, there are up to the outskirts of the city, the opening hours are longer, then it can increase customer brand love for minimarket franchise brands in Bali. The results of this study at the same time strengthen the results of previous research conducted by Amaro et al. (2020) regarding the role of intensive distribution in increasing brand love. These results are also strengthened by the results of research from Hamzah et al. (2021) found that intensive distribution had a positive and significant effect on brand love. It can be concluded that, to increase brand love for minimarket franchise brands in Bali, it is necessary to increase the distribution so that it is more intensive.

The Effect of Social Media Promotion on Brand Loyalty
Based on the analysis of the effect of social media promotion on brand loyalty, the beta coefficient value is 0.167 with a significance level of 0.033 0.05, which means Ho is rejected and H1 is accepted. These results mean, social media promotion has a positive and significant effect on brand loyalty. This means that the higher the social media promotions carried out, which is indicated by the use of Instagram, WhatsApp, and Facebook social media, the higher the brand loyalty in the minimarket franchise business in Bali. The results of this study at the same time strengthen the results of previous research conducted by Huang (2017) about social media promotion being able to increase brand loyalty. Furthermore, Fernandes and Inverneiro (2021); and Santos and Schlesinger (2021), researching the effect of social media promotion on brand loyalty also found the same results, namely social media promotion had a positive and significant effect on brand loyalty. It can be concluded that social media promotion has a positive and significant influence on brand loyalty, this means that the higher the promotion of social media, the brand loyalty of minimarket franchise brands in Bali will increase.

Effect of Social Media Promotion on e-WOM
Based on the analysis of the effect of social media promotion on e-WOM, the beta coefficient value is 0.161 with a significance level of 0.022 0.05, which means Ho is rejected and H1 is accepted. These results mean that social media promotion has a positive and significant effect on e-WOM. This means, the higher the promotion of social media, which is indicated by the use of social media Instagram, WhatsApp, and Facebook, the more positive e-WOM is about the minimarket franchise business in Bali. The results of this study at the same time strengthen the results of previous research conducted by Ramanathan et al. (2017), regarding the promotion of social media can increase positive e-WOM. Furthermore, Jose (2018); Bellaaj (2021), researching the effect of social media promotion on e-WOM also found the same results, namely social media promotion had a positive and significant effect on e-WOM. So it can be said that social media promotion has a positive and significant influence on e-WOM regarding minimarket franchise businesses in Bali.

The Effect of Social Media Promotion on Brand Love
Based on the analysis of the effect of social media promotion on brand love, the beta coefficient value is 0.565 with a significance level of 0.000 0.05, which means Ho is rejected and H1 is accepted. These results mean that social media promotion has a positive and significant effect on brand love. This means, the higher the promotion of social media, which is indicated by the use of social media Instagram, WhatsApp, and Facebook, the higher the brand love for minimarket franchise brands in Bali. The results of this study at the same time strengthen the results of previous research conducted by Raji et al. (2019); and Amaro et al. (2020), about social media promotion that can increase brand love for a product. Furthermore, Hamzah et al. (2021), researching the effect of social media promotion on brand love also found the same results, namely social media promotion had a positive and significant effect on brand love. Therefore, it can be concluded that social media promotion has a positive and significant influence on brand love from minimarket franchise businesses (Alfamart, Indomaret, and Cyrcle K) in Bali which is getting higher as well.

The Effect of Brand Love on Brand Loyalty
Based on the analysis of the influence of brand love on brand loyalty, the beta coefficient value is 0.506 with a significance level of 0.000 0.05, which means Ho is rejected and H1 is accepted. These results mean that brand love has a positive and significant effect on brand loyalty. This means, the higher the brand love shown by feeling happy with the brand, feeling comfortable, and the brand is in the heart, being the first choice, really loving the brand, then being able to increase customer brand loyalty to keep shopping at minimarket franchise businesses in Bali. The results of this study simultaneously strengthen the results of previous research conducted by Huang (2017) about the effect of brand love on increasing brand loyalty. Furthermore, there are other researchers, namely, Fernandes and Inverneiro (2021) who study about brand love being able to make brand loyalty higher; and Santos and Schlesinger (2021), found that brand love has a positive effect on brand loyalty. So it can be said that brand love has a positive and significant effect on brand loyalty The Effect of Brand Love on e-WOM Based on the analysis of the influence of brand love on e-WOM, the beta coefficient value is 0.455 with a significance level of 0.000 0.05, which means Ho is rejected and H1 is accepted. These results mean that brand love has a positive and significant effect on e-WOM. This means, the higher the brand love, which is indicated by feeling happy with the brand, feeling comfortable, and the brand being in the heart, being the first choice, really loving the brand, then being able to increase positive e-WOM. The results of this study at the same time strengthen the results of previous research conducted by Loureiro et al. (2017), regarding the effect of brand love on increasing positive e-WOM. Furthermore, there are other researchers, namely Li et al. (2021) studied about brand love being able to make e-WOM more positive; and Mainolfi and Vergura (2021), found that brand love has a positive effect on e-WOM, so it can be said that brand love has a positive and significant effect on e-WOM regarding minimarket franchise businesses in Bali.

The Role of Brand Love in Mediating the Effect of Intensive Distribution on Brand Loyalty
Based on the analysis of the influence of brand love mediating the effect of intensive distribution on brand loyalty, the beta coefficient value is 0.123 with a significance level of 0.026 0.05, which means Ho is rejected and H1 is accepted. These results mean that brand love can significantly mediate the effect of intensive distribution on brand loyalty. This means, with brand love, the effect of intensive distribution on brand loyalty is increasing.

The Role of Brand Love in Mediating the Effect of Intensive Distribution on e-WOM
Based on the analysis of the influence of brand love mediating the effect of intensive distribution on e-WOM, the beta coefficient value is 0.110 with a significance level of 0.028 0.05, which means Ho is rejected and H1 is accepted. These results mean that brand love is able to partially and significantly mediate the effect of intensive distribution on e-WOM. This means, with brand love, the effect of intensive distribution on e-WOM is increasing.

The Role of Brand Love in Mediating the Effect of Social Media Promotion on Brand Loyalty
Based on the analysis of the influence of brand love mediating the effect of social media promotion on brand loyalty, the beta coefficient value is 0.286 with a significance level of 0.000 0.05, which means Ho is rejected and H1 is accepted. These results mean that brand love is able to significantly mediate the effect of social media promotion on brand loyalty. This means, with brand love, the effect of social media promotion on brand loyalty is increasing.

The Role of Brand Love in Mediating the Effect of Social Media Promotion on e-WOM
Based on the analysis of the influence of brand love mediating the effect of social media promotion on e-WOM, the beta coefficient value is 0.257 with a significance level of 0.000 0.05, which means Ho is rejected and H1 is accepted. These results mean that brand love is able to significantly mediate the effect of social media promotion on e-WOM. This means, with brand love, the influence of social media promotion on e-WOM is increasing.

Conclusions, Implications, and Limitations of the Research
The conclusion from the results of this study is that brand love is able to act as a mediator on the effect of intensive distribution on brand loyalty; and mediating the effect of intensive distribution on e-WOM, as well as mediating the effect of social media promotion on brand loyalty; and mediating the effect of social media promotion on e-WOM.
This study develops an integrated model by linking several relationships between intensive distribution variables, social media promotion, brand love with brand loyalty and e-WOM. Likewise for minimarket franchise businesses, especially in Bali, to always increase their intensive distribution and always carry out social media promotions according to the development of existing social media platforms so that brand love, brand loyalty, and positive e-WOM are increasing.