The effect of product quality and price discounts on word of mouth and purchasing decisions at Esteh Indonesia

: This study aims to determine the effect between product quality and price discounts on word of mouth (WOM) and purchasing decisions at Esteh Indonesia. The subjects of this survey were all Esteh Indonesia consumers. In this study, 250 respondents were involved, selected by accidental sampling method, which is a sampling technique based on the coincidence or availability of consumers in a location. Data was collected through the distribution of questionnaires. Data analysis was conducted using Structural Equation Modeling (SEM) using Smart-PLS software. The results showed that product quality has a significant impact on WOM, but is not proven to have a significant influence on purchasing decisions. In addition, price discounts are also proven to influence and have a significant impact on WOM and purchasing decisions. WOM is proven to have a significant impact on purchasing decisions, which is influenced by credibility, personal relevance, emotional connection, group influence, and relatively low costs. Because of this, WOM is one of the effective marketing strategies in influencing consumer behavior.


Introduction
The development of the beverage business in Indonesia is currently growing rapidly as seen from many new beverage businesses popping up.One of the popular businesses in Indonesia today is Esteh Indonesia.Esteh Indonesia is one of the franchise beverage businesses that has the largest number outlets in Indonesia.
In running a business, it is important for companies to understand the aspects that make consumers enamored so that they buy a product that is being sold.There are various aspects that make consumers attracted to a product including quality and price discount offers.Both of these factors can be considered by word of mouth (WOM) as a mediator, which has the potential to strengthen consumer desires in making purchasing decisions on a product.
One of the company's strategies for marketing is by means of word of mouth (WOM).Wom was a policy related conversation between one individual and another.Word of Mouth (WOM) was made by companies for free to their loyal customers.WOM can also be defined as one of the important roles in purchasing decisions which serves to convince consumers that the desired product matches the criteria being sought (Wiratama et al., 2022; https://economics.pubmedia.id/index.php/jbhiDabhilkar et al., 2011).There are several benefits of the wom strategy, namely being able to provide the best service for consumers, asking consumers to provide reviews, and utilizing social media (Miraza et al., 2021).Therefore, WOM can establish a positive impression of a product, so this can influence purchasing decisions.Thus, WOM is not just a form of communication, but it can significantly affect consumer behavior.
What is meant by purchasing decisions is the stage of consumers in making a decision to buy a product after considering relevant information to assess whether the product is feasible or not (Saputra et al., 2022).Purchasing decisions can be said to be the main key to consumer behavior, because consumers take action after consuming a product.Purchasing decisions can also be interpreted as a step that allows consumers to make choices according to their needs, expectations, and desires so that when the item has been purchased, there will be a feeling of satisfaction or dissatisfaction.
Based on Kotler and Armstrong (2008), the definition of product quality is such an important capabilities when performing direct marketings so that it may influence the implementation of a product.Product quality refers to the superiority of a product in fulfilling customer expectations and needs.Product quality is also reviewed from the fulfillment of standards, specifications and competitiveness of products on the market.Good product quality indicates that the product is designed, manufactured and delivered with purchasing decisions and excellence in mind.Therefore, it is very necessary for companies to maintain the quality of products from food or beverages so that consumers are more satisfied with the products served (Taufik et al., 2022;Amin et al., 2022).Based on research by Ababil et al., (2019) that there is no influence between purchasing decisions and product quality.The same thing is also explained based on Astaki and Purnami's (2019) research that wom can mediate purchasing decisions with product quality.
Price discount is a price decrease given to buyers when they buy a product, either directly in a store or through an online platform (Syarifah Roviqoh & Supriati, 2022).The purpose of price discounts is to entice buyers to make large purchases, and this can affect consumer purchasing patterns and sales volume.The availability of a price discount makes consumers more likely to purchase the product, especially if the discount is perceived as attractive.Therefore, price discounts play an important role in consumer decisions to buy a product.The results of research conducted by Yildrim and Aydin (2012) explain that price discounts have a significant and positive effect on purchasing decisions.Therefore, it was concluded that price discounts are created when a consumer receives an information about a huge amount of price discounts, and it has potential for a product being purchased by a customer.As a result, the purpose of this research is to determine product quality and price discounts on wom and purchasing decisions. https://economics.pubmedia.id/index.php/jbhi

Sample and procedure
The population of this study were all consumers of Esteh Indonesia.This research was conducted in 2024.The 250 research participants were selected using the non-probability accidental sampling method, which is a sampling technique based on special considerations.Accidental sampling refers to a sampling technique that is done by chance or based on the availability of consumers in a location.Data were collected through the use of a questionnaire designed with questions based on indicators related to the research variables (Taherdoost, 2022;Haitao, 2022).A sample test was conducted on 30 Esteh Indonesia consumers to test the validity and reliability of the research variables.

Measure
Primary data was collected through distributing questionnaires to consumers who purchased Esteh Indonesia products using a platform, namely google form.This research evaluates each question using a Likert scale with a range of 1 to 5 which can be interpreted as: (1) strongly disagree, (2) disagree, (3) neutral, (4) agree, and (5) strongly agree.The variables measured involve product quality, price discounts, word of mouth (wom) and purchasing decisions.

Data analysis procedures
This research uses a quantitative approach where the Structural Equation Modeling (SEM) method with the Partial Least Square (PLS) approach is used to test the hypothesis.The analysis tool applied is SmartPLS version 3.0.PLS-SEM was used to evaluate the hypothesis in this study.According to Hair et al. (2014), PLS is a multivariate data analysis technique that is not bound by distribution assumptions.With SEM, the relationship between latent variables can be measured.

Descriptive Analysis Results
Based on the demographic data of the respondents recorded in the table, there were 250 respondents who took part in this study.The majority of respondents, 51.6%, are female, while the remaining 48.4% are male.In the age of group, respondents aged 15-25 years dominate with the highest percentage reaching 35.2%.Meanwhile, the S1 education level shows the highest percentage at 50.4%.The occupation as a civil servant has the largest share with a percentage reaching 18.8%.

Assessment of measurement model
Once the data is inputted to the construct model in SmartPLS, it is run to measure its validity and reliability.This process can be repeated several times until all indicator factor loading values reach the validity limit of 0.70 (Hair et al., 2017).Indicators with a loading factor under 0.70 need to be removed to improve the validity and reliability of the model.The results of SmartPLS calculations that pass the validity and reliability requirements can be seen in the following figures and tables:  According to the results of data analysis using SmartPLS listed in Table 2, it can be seen that all indicators on each variable in this research have a loading factor value that exceeds 0.70, indicating qualified convergent validity.Discriminant validity is done by observing the cross-loading value between measurement constructs.The cross-loading https://economics.pubmedia.id/index.php/jbhivalue shows how much the cross-loading value is correlated between each construct with its own indicators and indicators from other constructs.A measurement model is considered to have good discriminant validity if the correlation between the construct and its indicators is higher than the correlation with indicators from other constructs.After the data analysis process using SmartPLS 3.0, the cross-loading results can be seen in Table 3.The cross-loading results recorded in Table 3 show that the correlation between each construct and their indicators is higher than the correlation with other constructs.Therefore, it can be concluded that all the constructs or latent variables have shown sufficient discriminant validity, where indicators in each construct block are superior to indicators in other blocks.(Sekaran & Bougie, 2019).Based on the table above, the highest values were found for the purchasing decision variable (0.957), the wom variable (0.953), the product quality variable (0.923), and the price discount variable for staying (0.888).This shows that each question indicator has a higher cross-loading value on the latent construct compared to its construct compared to indicators on other latent constructs.As a result, the conclusion is that the discriminant validity test is valid.Besides evaluating the existence of significant relationships between variables, the authors are also suggested to use a size effect such as f-squared to assess how large the effect between variables is (Kwong & Wong, 2013).The standard f-squared measure is 0.02 for small effects, 0.15 for medium effects, and 0.35 for large effects.Values less than 0.02 can be considered insignificant or can be ignored (Sarstedt et al., 2017).
Based on Table 6, the size of the effect of the Price Discount variable on Purchasing Decisions and WOM of 0.047 and 0.220 indicates a significant effect at a small to medium level.For the Word Of Mouth (WOM) variable on Purchasing Decisions is 0.579 which indicates a large influence.The amount of f2 of the Product Quality variable on WOM is 0.247, indicating a moderate influence.Meanwhile, there are variables that have had no effect, namely Product Quality on Purchasing Decisions of 0.007.From these f-squared values, it can be concluded that the influence of the Word Of Mouth (WOM) variable on Purchasing Decisions is the greatest influence observed in this analysis.
https://economics.pubmedia.id/index.php/jbhiEvaluation of the accuracy of the prediction method can be done using the Q-square value.When the Q2 value is greater than 0, it signifies that the exogenous variables have significant predictive relevance to the endogenous construct variables.Otherwise, when the Q2 value is smaller than 0, it indicates that the exogenous variables have lower predictive relevance to the endogenous construct variables.Once a model is considered to have good predictive ability if Q2 is greater than 0, while a model is considered to have poor predictive ability if Q2 is less than 0. Table 7 shows that purchasing decisions and word of mouth (WOM) are included in variables with Q2 greater than 0. Testing predictive accuracy can be done using the Q2 value.

Discussion
The results of this hypothesis testing are based on the analysis of parameter coefficients in the inner (structural) model and t-statistics.The assessment of whether a hypothesis can be accepted or rejected is based on the significance value between constructs, the t-statistic, and the p-value.Hypothesis testing in this research was carried out using SmartPLS (Partial Least Square) 3.0 software, and these values can be found in the results of the bootstrapping process.The rule of thumb used in this study is that the t-statistic should be greater than 1.96, with a p-value of less than 0.05 (5%) at the significance level, as well as the presence of a positive beta coefficient.The results of hypothesis testing are presented in Table 8.
The first hypothesis tests whether word of mouth has a positive impact on purchasing decisions.The analysis results show that the coefficient for word of mouth (WOM) on purchasing decisions is 0.673, with a t-statistic value of 9.843, or >1.96, and a p-value of less than 0.05 (0.000).Because of that, the first hypothesis is accepted.This finding confirms that word of mouth (WOM) significantly influences purchasing decisions with a positive effect.
This second hypothesis examined if the product quality has a positive effect on word of mouth (WOM).The analysis results show that the beta coefficient for product quality on WOM is 0.479, with a t-statistic value of 6.884, and a P-value of 0.000.Based on the results, the t-statistic is considered significant because the value is higher than 1.96, and the P-value is smaller than 0.05, so that the second hypothesis can be accepted.This result indicates that product quality significantly influences word of mouth (WOM) with a positive impact.
The third hypothesis tests whether product quality has a positive impact on purchasing decisions.The analysis results show that the beta coefficient for product quality on purchasing decisions is 0.077, with a t-statistic value of 1.022, and a P-value of 0.307.From these results, the t-statistic is considered insignificant because the value does not exceed 1.96, and the P-value is greater than 0.05, so the third hypothesis cannot be accepted.This is supported by previous research which states that product quality has no effect on purchasing decisions (Ababil et al., 2019) .Therefore, it has been proven that product quality does not have a significant impact on purchasing decisions because consumers prioritize products that can meet their needs and desires, not just focus on product quality alone.Consumers also look for products that have prominent benefits and advantages.
The fourth hypothesis tests whether price discounts have a positive impact on word of mouth (WOM).The analysis results show that the beta coefficient of price discount on WOM is 0.452, with a t-statistic value of 6.566, and a P-value of 0.000.Based on these results, the t-statistic is significant because the value exceeds 1.96, and the P-value is less than 0.05, so the fourth hypothesis can be accepted.This result confirms that price discount significantly influences word of mouth (WOM) with a positive effect.
The fifth hypothesis tests whether price discounts have a positive effect on purchasing decisions.The analysis results show that the beta coefficient for price discount on purchasing decisions is 0.204, with a t-statistic value of 2.793, and a P-value of 0.005.Based on these results, the t-statistic is considered significant because the value is above 1.96, and the P-value is less than 0.05, so the fifth hypothesis can be accepted.This finding indicates that price discounts significantly influence purchasing decisions with a positive impact.

Figure 1
Figure 1 Structural Model https://economics.pubmedia.id/index.php/jbhiConclusion The conclusion should be written in Palatino Linotype 12 font size.The conclusion should only address the research objectives.It should narrate how your research advances the field from the current state of knowledge.Without a clear conclusion, reviewers and readers will struggle to evaluate the work and determine its suitability for publication in a journal.Avoid repeating the Abstract or simply listing experimental results.Provide a clear scientific justification for your work and demonstrate potential applications and extensions.This conclusion should be articulated in one paragraph.You should also suggest future experiments and/or indicate ongoing experiments.

Table 1
Demographic Profile of Respondents

Table 2
Convergent Measurement Items, Reliability and Validity.
(Hair et al., 2017)n Test is used to determine how much effect on the independent variable has on the dependent variable by measuring the coefficient of determination.Based on information from Table5, it can be concluded that 86.2% of purchasing decision variables can be explained by WOM variables, price perceptions, and price discounts, while the remaining 13.8% is influenced by other factors not included in the model.Meanwhile, about 82.4% of the variation in Word of Mouth (WOM) can be explained, while the other 17.6% is influenced by factors not explained in this study.The criteria for the R-square value, which is described as substantial, moderate, and weak with values of 0.75, 0.50, or 0.25, indicate that the model has moderate strength(Hair et al., 2017). https://economics.pubmedia.id/index.php/jbhi