DIGITAL COMMUNICATION: BRIDGE TO REPURCHASE INTENTION

ABSTRACT


INTRODUCTION
Milo is a Nestlé dairy product, invented in 1934 by Thomas Mayne in Sydney, Australia. It is offered in many countries worldwide, including Indonesia, and is renowned as a milk brand by Indonesians, even though it sells other items as well. According to the Top Brand Index (TBI) statistics (2021), it was ranked fifth in the ready-to-drink liquid milk category. Furthermore, in the milk powder category, it was picked third out of four milk brands in the same year. The research suggests that customers are more familiar with and like Milo's powdered milk versions than other Milo products. The data then became the foundation for this study. The company's success is reflected in the many loyal consumers who continually purchase its products. The more customers intend to repurchase Milo products, the more revenue the firm earns. To maintain customer loyalty and generate strong repurchase intentions, Nestlé must be able to meet the demands and desires of its customers.
The idea of brand familiarity refers to the accumulation of similar experiences that customers have with brands (Alba & Hutchinson, 1987;Huang, 2017). The frequency of service-related experience-bound products accumulated measures customers' brand familiarity. Consumers with experience with a product or brand are more familiar with them (Ha & Perks, 2005). Familiarity, sometimes used interchangeably with the term "knowledge," is a broader concept encompassing consumer experience, prior knowledge, perceptions, and beliefs. Brand familiarity refers to people's familiarity or intimacy with a brand through direct or indirect interactions. e-WOM is described as positive and negative comments or talks about specific items, services, or companies that may be accessible by anybody online (Hennig-Thurau et al., 2004;Kajtazi & Zeqiri, 2020). Oral, written, and electronic communication amongst communities about the quality and experience of ordering or utilizing a product or service is called electronic word of mouth (Litvin et al., 2008). Electronic word-of-mouth (e-WOM) refers to consumergenerated communication about products or services shared through digital platforms and is typically intended for other consumers (Rahman et al., 2019). This communication can include reviews, ratings, recommendations, and other forms of online communication.
Brand love is characterized by customers' strong emotional attachment to a specific brand. It is a deep connection beyond just liking or preferring the brand and involves feelings of admiration and devotion (Carroll & Ahuvia, 2006;Fetscherin & Raghabendra P., 2021). It is defined as the emotional degree and arousal owned by existing consumers who are happy with that brand (Cho & Hwang, 2020). In addition, brand love refers to consumers' intense emotional attachment to a brand that satisfies them. This attachment is marked by intense admiration and devotion toward the brand (Batra et al., 2012). It is concluded that brand love generates customer desire, loyalty, and advocacy to convert them into advocates or influencers for a brand.
Brand attitude is a consumer's overall opinion of a brand based on their feelings and perceptions, shaped by their experiences with the brand and resulting in positive or negative responses (Grace & O'Cass, 2005). Brand attitude is essential because it is the main factor influencing a consumer's decision to choose a particular brand (Low & Lamb, 2000;Sharma et al., 2022). Customers tend to prefer and base their purchasing decisions on a brand they are satisfied with, leading to actual purchases of that brand.
Consumer attitudes toward brands are a well-established topic in marketing and branding that has been thoroughly researched over the years. Consumer attitudes toward brands, or simply brand attitudes, are a well-established topic in marketing and branding that has been thoroughly researched over the years (Keller, 2020). Brand attitude is a summary judgment that influences behavior and considers various aspects of a brand. It may impact how a consumer perceives the brand (Singh & Spears, 2013). It is established that Brand Attitude is a consumer appraisal of the brand and response to the brand. All elements, such as ideas, values, beliefs, and perceptions, influence a person's attitude, making it critical to deciding one's attitude.
Repurchase intention refers to the likelihood that an individual will continue to purchase products or services from a particular company in the future (Wong & Yu, 2013). The individual decides to make repeat purchases of a particular service from the same company (Hellier et al., 2003). Essentially, repurchase intention is the desire of customers to repurchase products or services in the future.
Online communication, such as reviews on websites and apps, social media posts, and forum discussions, can significantly impact costumer's purchasing decisions for Milo products. Therefore, this research focuses on the factors that influence repurchase intention for the Milo brand, intending to understand how to increase customer loyalty and repurchase intentions. The researchers hope to shed more light on digital communication's importance in modern marketing, especially for food and beverage marketing.
Hypotheses used in the study were: 1) H-1: Brand Familiarity has a significant effect on Repurchase Intention; 2) H-2: e-WOM has a significant effect on Repurchase Intention; 3) H-3: Brand Love has a significant effect on Repurchase Intention; 4) H-4: Brand Attitude has a significant effect on Repurchase Intention; 5) H-5: Brand Familiarity has a significant effect on Brand Attitude; 6) H-6: E-WOM has a significant effect on Brand Attitude; and 7) H-7: Brand Love has a significant effect on Brand Attitude.

METHOD
This study used a sample to represent the larger population being studied. The Structural Equation Modeling (SEM) analysis sample size should be between 100 and 200 respondents (Sparkman et al., 1979). To meet this requirement, the number of research indicators was multiplied by 5 to 10. The survey method was employed, with respondents who had consumed Milo products being interviewed and asked to complete questionnaires. The primary data collected from these respondents were randomly selected using a random sampling method.
The questionnaire used in this study included items rated using the Likert scale. This scale measures how strongly respondents agree or disagree with a statement, with the following options: 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, and 5 = strongly agree (Sekaran & Bougie, 2016).
The outlier data were removed before analyzing the collected data. Then, the validity and reliability of the measuring instrument were checked. Construct validity was assessed using the average variance extracted (AVE) value, which should be greater than 0.3 (Sugiyono, 2019). The reliability of the test was measured using Composite Reliability and Cronbach's Alpha, with values above 0.7 indicating adequate internal consistency (Leguina, 2015). Finally, the data was verified and analyzed using the partial least squares method.
The validity test assesses the consistency between the data collected and the characteristics or qualities of the studied item (Sugiyono, 2019). The correlation test is utilized, and the validity is a metric that reveals an instrument's reliability and validity. The instrument is valid if the correlation coefficient between the indicators and the total indicators is greater than or equal to 0.30 (r 0.30). It has a star, meaning that two stars are acceptable or valid with an error rate of 1% or 99% truth, and 1 star is still acceptable with an error rate of 5% or 95%.
The model's appropriateness was evaluated using various goodness-of-fit criteria to determine whether it is acceptable or should be rejected. Several suitability indices and associated cut-off values are used to make this determination, including 1) A low chi-square statistic for the tested model is considered excellent or adequate if the p-value is more significant than 0.10. 2) The Root Mean Square Error of Approximation (RMSEA) measures the predicted goodness of fit if the model is estimated in the population. A value of 0.08 is considered acceptable. 3) The Root Mean Square Residual (RMR) and Standardized Root Mean Residual (SMRM) are used as input to covariances. However, the residuals are expressed in terms of covariances, which makes them difficult to interpret because the scale of the indicators influences them. A value of RMR ≤ 0.10 is considered acceptable. 4) The Goodness of Fit Index (GFI) is a non-statistical metric with a range of 0 (poor fit) to 10 (excellent fit). Higher scores indicate a better fit. 5) The Adjusted Goodness of Fit Index (AGFI) should be at least 0.90 for an acceptable model. 6) The minimum sample Discrepancy Function divided by its degrees of freedom (CMIN/DF) is a relative chi-square statistic divided by the degrees of freedom. A relative score of less than 2.0 or 3.0 indicates an appropriate fit between the model and the data. 7) The Tucker-Lewis Index (TLI) is a cumulative index that compares a model under test to a baseline model, with a value of 0.95 being the suggested reference for adopting a model and a value close to 1 indicating a very excellent fit. 8) The Comparative Fit Index (CFI) has a range of 0-1, with values closer to 1 indicating the best fit (a perfect fit).
Hypothesis testing uses statistical techniques to evaluate a proposition to determine whether it is statistically significant. Before testing a hypothesis, it is necessary to identify it, a statement still open to question. The data is then collected, and statistical tests are performed to determine whether the claims made in the hypothesis are supported. Based on the results of these tests, the hypothesis can be accepted (if the data does not provide evidence to contradict it) or rejected (if the data provides evidence to reject it).

RESULTS AND DISCUSSION
In the data collection stage, the researchers distributed 220 offline questionnaires, which were distributed using the Accidental Random Sampling technique, namely by giving questionnaires to those who had purchased or consumed Milo in Samarinda City. After that, a selection of 147 data was carried out by randomly selecting samples from the available population, which was feasible for use in the following research stage. The loading factor numbers indicate that all of the measuring instruments used in this study can adequately represent the variables they are designed to measure, as they meet the standards used in the analysis.  According to the data above, the fit model test results show that all tests are marginally fit or close to fit. The term "marginal fit" refers to how well the data fits the current model. However, it is not unsuitable, but it is nearly or entirely fit. Three of the six relationships were significant based on their probability values (p-values) below 0.05. These relationships were between e-WOM to brand attitude, brand love to brand attitude, and brand attitude to repurchase intention. The remaining three relationships were deemed insignificant because their p-values were more outstanding than 0.05.

CONCLUSION
Brand Familiarity had no significant effect on brand attitude among Milo consumers in Samarinda City. The result suggests that consumers' familiarity with the Milo brand does not impact their experiences with the brand. e-WOM significantly affected brand attitude among Milo consumers in Samarinda City. The results of this study indicate that electronic word of mouth through social media can influence consumers' attitudes toward the Milo brand.
Brand Love significantly affected brand attitude among Milo consumers in Samarinda City. The result suggests consumers love the Milo brand because it offers various size options to suit their needs, has a delicious and healthy taste and content, and provides above-average satisfaction that leads to a pure enjoyment experience for consumers and a positive belief that Milo can fulfill their nutritional needs. The results of this study show that love for the brand can influence consumers' attitudes toward Milo.
Brand Attitude significantly affected repurchase intention for Milo consumers in Samarinda City. The result suggests that consumers' positive attitudes towards the Milo brand led to a desire to repurchase Milo products. Therefore, consumers' attitudes towards Milo are reflected in their intention to continue consuming its products. The results of this study show that consumer attitudes toward a brand can influence their intention to repurchase it.
Brand Love did not significantly affect repurchase intention for Milo consumers in Samarinda City. The result suggests that while consumers may like Milo, they do not necessarily plan to repurchase it, possibly because, according to the respondents in this study, their average age is 21-30 years. They do not view Milo products as their primary choice for beverages. The results of this study show that love for the brand does not influence the repurchase intention of Milo consumers. e-WOM had no significant effect on repurchase intention for Milo consumers in Samarinda City. The result suggests that information about where to buy Milo on social media does not influence consumers to repurchase the brand. The results of this study show that electronic word of mouth does not affect repurchase intentions among Milo consumers.
The research is mainly limited to one brand, despite of the brand's various products. Therefore, future research can take the current study as an example of brand research and conduct a deeper study regarding other brand's marketing strategies and potentials.