THE EFFECT OF BRAND IMAGE AND PRODUCT QUALITY ON REPURCHASE INTENTION MEDIATED BY CONSUMER SATISFACTION STUDY AT UNIQLO IN SOLO

: The purpose of this study was to examine and analyze the effect of brand image and product quality on repurchase intention mediated by consumer satisfaction at Uniqlo in the city of Solo. The data/sample collection method used a questionnaire, with a sample of 200 consumers who had purchased Uniqlo products at least once. Data analysis in this study used PLS-SEM analysis. Based on the data and research results, it is known that there is a positive but not significant effect between brand image on repurchase intention. Meanwhile, product quality has a positive and significant effect on repurchase intention. Brand image and product quality have a positive and significant effect on the mediating variable of consumer satisfaction. And consumer satisfaction has a positive and significant effect on repurchase intention. The results showed that the mediating variable of consumer satisfaction had a positive and significant effect in mediating the relationship between brand image and product quality on repurchase intention. The results of the mediation effect test of the independent variable brand image on the dependent variable of repurchase intention through the mediation variable of consumer satisfaction shows the original coefficient value of 0.373, it shows that the direction of the influence is positive while the value of the t statistic is 4.742 ≥ 1.96 so that the consumer satisfaction variable mediates the effect of brand image on the dependent variable of repurchase intention. while statistic


Introduction
The fashion industry is currently growing rapidly throughout the world, including in Indonesia. This is indicated by the many local and international fashion brands that were born and developed because there is still a good market open in Indonesia. In the midst of the current booming fashion industry, several fashion brands with foreign brands have started to enter Indonesia through their own distributors and branch stores. This condition will certainly trigger competition between brands both with local brands and with foreign brands (Mariadi and Aima, 2014).
One of the fashion brands that is developing in Indonesia is Uniqlo. Uniqlo is a Japanese company brand that is engaged in fashion and produces various types of casual clothing. Uniqlo was founded in 1984 and is a premium fashion brand that is quite well known in Indonesia, especially among young people. Products marketed by Uniqlo in Indonesia include clothes, tshirts, pants, jackets, sweaters, and so on. In its business, Uniqlo targets the market share of middle-upper consumers, where the majority of the enthusiasts are young people. Uniqlo already has 46 stores spread across Indonesia, and one of them is in the city of Solo.
In marketing, Uniqlo emphasizes on introducing their brand as a clothing brand with good quality with the concept of Life Wear. The concept is a clothing concept that is constantly updated, lighter, provides warmth, with better designs, and is comfortable for people's lives. With the Life Wear concept, Uniqlo hopes to become a strong and big brand in the minds and hearts of consumers, and which in the end will make them interested in buying products from the Uniqlo brand. To place a brand image based on the Life Wear concept in the city of Solo, Uniqlo must Figure 1. Research Framework (this research framework is inspired by research conducted by Razak et al., 2016)

Method
This type of research is quantitative research because the data in this study are in the form of numbers and analysis using statistics and the research was conducted in the city of Solo. This study uses variables that include brand image, product quality, consumer satisfaction, and repurchase intention. Figure 1. shows a research framework inspired by Razak et al. (2016) research.
Based on temporary allegations, Brand Image has an effect on Repurchase Intention (Huang el al., 2019;Tunjungsari et al., 2020;Pradini, 2012). Brand Image has an effect on Consumer Satisfaction (Paramananda et al., 2018;Setyowati and Wiyadi, 2016;Cuong and Long, 2020). Product Quality has an effect on Repurchase Intention (Ika and Pebrianto, 2020;Ristiono, 2013;Devi and Sugiharto, 2017). Product Quality has an effect on Consumer Satisfaction (Rydin et al., 2018;North et al., 2003;Solomon and Rabolt, 2009) Consumer Satisfaction has an effect on Repurchase Intention (Kitapci et al, 2014;Ilyas et al, 2020;Mensah et al, 2018). Brand Image has an effect on Repurchase Intention mediated by Consumer Satisfaction (Paramananda et al, 2018;Mensah et al, 2018). Product Quality has an effect on Repurchase Intention mediated by Consumer Satisfaction (Devi and Sugiharto, 2017;Kitapci et al, 2014) Then the hypothesis in this study: H1: Brand image has a significant effect on repurchase intention. H2: Brand image has a significant effect on consumer satisfaction. H3: Product quality has a significant effect on repurchase intention. H4: Product quality has a significant effect on consumer satisfaction. H5: Consumer satisfaction has a significant effect on repurchase intention. H6: Brand image has a significant effect on repurchase intentions mediated by consumer satisfaction. H7: Product quality has a significant effect on repurchase intentions mediated by consumer satisfaction. In this study, the determination of the sample used 10 times the number of indicators, namely 20 indicators, so that the number of indicators obtained was 200 samples. The type of sample used by the researcher is a non-probability sampling method and the sampling technique is purposive convenience sampling. This technique was chosen because the information is needed to answer research questions and achieve the objectives of research in a particular sample group (Ferdinand 2005). The sample criteria in this study were several customers who had purchased products at Uniqlo at least once. The sample criteria are as follows: (1) consumers who make purchases at Uniqlo (2) consumers are decision makers to purchase clothes at Uniqlo (3) consumers who have made purchases at least 2-3 times.
Data collection in this study used a questionnaire by distributing questionnaires and accompanied by alternative answers using a Likert scale. The next testing stage is the Convergent Validity, Discriminant Validity, Average Variance Extracted (AVE) test, and the Reliability Test. Convergent Validity is the value of a latent variable loading factor that uses its indicators with an expected value of > 0.7. According to Chin, (1998) the initial stage in the study of 0.5 -0.6 is considered sufficient. The loading factor limit used in this study is 0.5. Discriminant Validity is the value of the cross loading factor used to determine whether a variable has an adequate discriminant, meaning by comparing the value of the intended loading variable. In other words, the value must be greater than the loading variable value.
In the Average Variance Extracted test, there are two types of validity in the PLS-SEM test, namely convergent and discriminant. Convergent validity itself has the meaning of a set of indicators that will represent one of the latent variables, and will also be the basis of the latent variable. Convergent validity can also be illustrated by means of the unidimensionality shown by using the value of the extracted mean variance (AVE). In this study, the expected AVE is > 0.5.
The reliability test is a measure related to an internal consistency of the indicators of a construct that shows the degree of each indicator indicating a general latent construct. That is, how these specific things will help each other in terms of explaining general phenomena. Hulland et al, (1996) and Hair et al, (1998) explain to assess the reliability of the accepted value limit is 0.07. In the research of Hair et al, (1998) and Sekaran, (1999) explaining the reliability of internal consistency used for the test results of each construct with a Cronbach's alpha coefficient that meets the requirements of the rules of thumb, namely  0.7. This can be said to be reliable if the composite reliability value is 0.7.

Results And Discussion
Test Outer Model 1. Convergent Validity Test In the convergent validity test that has been carried out, the results show that all indicators in this study there is 1 indicator that does not meet the limit > 0.7, so it will be eliminated. Thus, all indicators have been declared to meet the criteria of convergent validity.

Discriminant Validity Test
The results of the research conducted showed that the Composite Reliability value was above 0.5. Therefore, the conclusion that can be drawn is that all indicators are said to be reliable and reliable for use in research. And with a composite reliability value above 0.5, it shows that the construct in the model is estimated to meet the discriminant validity criteria.

Average Variance Extracted (AVE)
The results of the AVE test that have been carried out show that the overall value shows a value > 0.5. These results are in line with the researchers' expectations.

Reliability test/reliability test
The results obtained in the Reliability Test are that the value of Cronbach's Alpha > 0.7, so that it can be concluded that the study has fulfilled the rules of thumb.

Q 2 value
The value of Q2 has a value equivalent to the coefficient of total determination in the path analysis. The researcher uses the formula (Ghozali, 2016) namely the Stone-Geisser Q Square Test to determine the value of Q2, where the formula is as follows: Based on the results of these calculations, the value of Q2 has a value of 0.96 where the vulnerability is 0 < 0.95 < 1 where the closer to the value 1 the model is said to be better. The conclusion obtained by the model is said to be very good.

Test of Influence and Significance
a) The effect of brand image on repurchase intention. Table 3 shows the positive results of the original sample coefficient of 0.072. This shows that the direction of the influence of brand image on repurchase intention is positive. Furthermore, the P Value of 0.384 is greater than the level of significance, namely 5% / 0.05. The conclusion is that there is a positive influence but there is no significant effect of brand image on repurchase intention. b) The effect of brand image on consumer satisfaction. Table 3 shows a positive result on the original sample coefficient of 0.637. This shows that the direction of the influence of brand image with consumer satisfaction is positive. Furthermore, the P Value of 0.000 is smaller than the level of significance, namely 5% / 0.05. The conclusion that can be drawn is that there is a positive and significant influence of brand image on consumer satisfaction. c) The effect of product quality on repurchase intention. Table 3 shows a positive result on the original sample coefficient of 0.280. This shows that the direction of the influence of product quality with repurchase intention is positive. Furthermore, the P Value of 0.000 is smaller than the level of significance, namely 5% / 0.05. The conclusion that can be drawn is that there is a positive and significant influence of product quality on repurchase intention. d) The effect of product quality on consumer satisfaction. Table 3 shows a positive result on the original sample coefficient of 0.322. This shows that the direction of the influence of product quality with consumer satisfaction is positive. Furthermore, the P Value of 0.000 is smaller than the level of significance, namely 5% / 0.05. The conclusion that can be drawn is that there is a positive and significant effect of product quality on consumer satisfaction. e) The effect of consumer satisfaction on repurchase intention. Table 3 shows positive results on the original sample coefficient of 0.586. This shows that the direction of the influence of consumer satisfaction with repurchase intention is positive. Furthermore, the P Value of 0.000 is smaller than the level of significance, namely 5% / 0.05. The conclusion that can be drawn is that there is a positive and significant influence on consumer satisfaction on repurchase intention. f) The effect of brand image on repurchase intention mediated by consumer satisfaction. Table 3 shows a positive result on the original sample coefficient of 0.373. This shows that the direction of the influence of brand image on repurchase intention through the mediation of consumer satisfaction is positive. Furthermore, the P Value of 0.000 is smaller than the level of significance, namely 5% / 0.05. The conclusion is that there is a positive and significant influence of brand image on repurchase intention through the mediating variable of consumer satisfaction. g) The effect of product quality on repurchase intention mediated by consumer satisfaction. Table 3 shows a positive result on the original sample coefficient of 0.189. This shows that the direction of the influence of product quality on repurchase intention through the mediation of consumer satisfaction is positive. Furthermore, the P Value of 0.000 is smaller than the level of significance, namely 5% / 0.05. The conclusion is that there is a positive and significant effect of product quality on repurchase intention through the mediating variable of consumer satisfaction.

Testing Direct Effects-Indirect Effects
From the test results, there is a direct effect of exogenous variables on endogenous variables, namely Brand Image on Repurchase Intention of 0.072, which means that the effect of exogenous and endogenous variables is positive and significant. Product Quality Variable on Repurchase Intention of 0.322, which means that the effect of exogenous and endogenous variables is positive and significant.
The indirect or indirect effect through the mediating variable Brand Image on Repurchase Intention is 0.373, which means that the effect of exogenous and endogenous variables is positive and significant. Product Quality variable on Repurchase Intention through the mediation variable of consumer satisfaction is 0.189, which means that the influence of exogenous and endogenous variables is positive and significant.  (2004), said GoFsmall = 0.1, GoFmedium = 0.25 and GoFlarge = 0.38. Based on the results of the above calculation, the GOF value in the study was 0.746 or 74.9%, this shows the level of feasibility of the high/large model.