The Role Website Quality, Credit Card, Sales Promotion On Online Impulse Buying Behavior

: The usage of online marketplace in Indonesia increases due to Covid-19 pandemic and its supporting environment such as payment systems. This investigation was conducted to determine the effect of Website Quality on Online Impulsive Buying Behavior moderated by Sales Promotion and Credit Card Usage in Indonesian marketplace. This study uses quantitative methods with causal analysis. In this research, data was collected through online questionnaires and 275 respondents who used the marketplace website responded. This research uses PLS-SEM data analysis technique. The results of this study showed that three out of five hypotheses are accepted. This study shows that Website Quality, Credit Card Use, and Sales Promotion have positive significant effect on Online Impulse Buying Behavior. However, the result of this study also revealed interesting findings, that there is not enough evidence to support moderation effect of Credit Card use and Sales Promotion in the relationship between web quality and Online Impulse Buying Behavior. Abstract:


INTRODUCTION
when compared to the first quarter of 2020 (Dianka, 2021). The number of site visits to Shopee reached 127.400,000 in the first quarter of 2021 (Dianka, 2021). Retailers create a website that serve as a platform for potential customers to engage and transact, making buying more convenient (Widagdo and Roz, 2021). Previous research has found that website features or design can impact impulsive buying behavior by functioning as an environmental signal (Wiranata and Hananto, 2020).
In research by (Ittaqullah et al., 2020), said that there are two factors that influence impulse buying or unplanned purchases, External elements like as discounts, displays, and promotions, as well as internal consumer characteristics such as lifestyle, shopping addiction, resource availability, and so on. Shopee itself is known for its Marketing Strategy because it can attract the attention of the Indonesian people to continue transacting on the platform (Christina, 2019). With a variety of promotional activities and campaigns carried out by Shopee in 2017 Shopee won the award for "The Best Marketing in Campaign" which was organized by Marketing Magazine (Dinisari, 2017). Shopee also recorded an increase in transactions in several other product categories, such as souvenirs and clothing (Ginee, 2021).
Another factor is Credit card use, that allows for rapid purchases and acts as a trigger for spending behavior from a psychological standpoint (Akram et al., 2018). This element is backed by the increased knowledge to stay at home to prevent the spread of the corona virus, as well as the increased usage of the marketplace, where credit cards are one of the favored payment tools (Maulana, 2021). Filianingsih Hendarta, Assistant Governor and Head of the BI Payment System Policy Department, said that, based on April 2021 statistics, nominal credit card transactions had climbed by 25.5 percent year on year (Putri, 2021). It is interesting to see how this payment method role change among other growing payment method offered by many financial technology nowadays. (Akram et al., 2018) revealed that the quality of a web page might drive online impulsive purchasing behavior. The study also discovered that sales advertising is a significant component in increasing OIBB since it encourages consumers to buy things spontaneously and instantly. OIBB and sales promotion have a strong and meaningful relationship. Credit card use allows consumers to make immediate purchases. As a result, the author uses the Shopee website as the object. Even though there has been researches that explores the influence of web quality on online impulse buying behavior, but to our knowledge, there is still limited research that explores the influence of web quality on online impulsive buying behavior especially moderated by sales promotion and credit card use in growing country like Indonesia. THEORETICAL REVIEW Impulse Buying. According to (Nurcaya and Rastini, 2018), Impulse Buying Behavior define as phenomena and a developing trend of purchasing behavior that occurs in the market and happens to be a key point that underpins marketing activity. It's also type of purchase that was not previously planned by the customer, where this circumstance is purposefully produced by the retail entrepreneur to survive in the business (Nurcaya and Rastini, 2018). Information and prediction of consumer reactions to changes in actions (situational variables), as well as understanding of the influence of demographic and quality of their website, failure of which may result in consumer loss. This statement in line with what (Wells et al., 2011), has claimed that a well-designed interface increases the likelihood of impulsive purchases. Empirically, this result is consistent with the study was conducted by (Akram et al., 2018), It concludes that web quality has a favorable and significant influence on online impulse buying behavior.
A study that conducts by (Turkyilmaz, 2015), reveals that the three characteristics of Web Quality, which include ease of use, usefulness, and entertainment, have a favorable and substantial influence on online impulsive purchasing behavior. It was determined that a high-quality website drives people to make accidental purchases (Fitri, 2018). But in research conduct by (Wiranata and Hananto, 2020), state that web quality does not influence online impulse buying since the analysis found that website quality was not found significant toward online impulse buying. This demonstrates that website quality alone was insufficient to drive online impulsive purchase. According to the explanation, the hypothesis in this study is as follows: H1: Website Quality has positive significant effect on Online Impulse Buying Behavior.
According to (Akram et al., 2018), the objective is to have a direct and immediate influence on the purchase habits of customers. Usually use by Marketers as develop customer marketing tactics for specific product categories to influence customers' impulsive purchases (Alimpic et al., 2020). In line with previous study by (Akram et al., 2018), states that what consumer gains from sales promotions which can be classified as hedonic (amusement, value, and discovery) or utilitarian (cost savings) (convenience and money saving). The result of research by (Fitri, 2018), It has been demonstrated that sales promotion tools, notably price reductions, free samples, buy one get one free, and loyalty programs, have a positive influence on consumer impulsive purchase behavior. These are the most used promotional methods to attract impulse buying behavior. From the explanation, the hypothesis in this research describe as follows: H2: Website Quality has positive significant effect on Online Impulse Buying Behavior.
In the previous study conduct by (Chen and Yao, 2018), It has been proposed that sales promotions cause customers to buy on impulse because they promote pleasant feelings and shift the consumer's standard of reference pricing, which affects their readiness to make a purchase. According to (Lo et al., 2016), indicates that website quality components were given as a hygiene element (i.e., the essential criterion to keep customers from quitting an online store, but not sufficient to stimulate impulsive spending), but sales promotion was highlighted as a driver of online impulse buying. From that it can be conclude that the presence of sales promotion would amplify the favorable effect already generated by strong website quality qualities, resulting in increased customer impulsive buying behavior (Wiranata and Hananto, 2020). Based on the explanation, the hypothesis in this research describe as follows: H3: Sales Promotion has positive significant moderate the relationship between website quality and Online Impulse Buying Behavior.
There are several aspects of internet purchasing that make it appealing and pleasurable for customers (Akram et al., 2018). This, along with the availability of non-cash payment systems in our current day, such as credit cards, simplifies materialistic purchases (Fendy Cuandra, 2021). Credit cards, from a psychological aspect, are a trigger for spending behavior (Akram et al., 2018). According to the findings of a study conducted by (Akram et al., 2018), credit card utilization has a favorable and substantial effect on online impulsive purchase. This result also found in (Badgaiyan and Verma, 2015), research show the fact that credit cards were discovered to have a considerable impact on impulsive purchasing behavior sums up the key contribution of the empirical investigation. However, these findings contradict the findings of the research conduct by (Yuniarti et al., 2021), which shows that the payment method using a credit card is included in the negative category or has no effect on online impulse buying behavior. Based on the explanation, the hypothesis in this research describe as follows: H4: Credit Card has positive significant effect on Online Impulse Buying Behavior.
Digital payments utilized as a route for transferring goods and services are referred to as cashless payments, such as transfers or using a credit card by simply displaying the credit card number on the e-commerce site (Dewi and Mahyuni, 2020). When using credit cards, make shopping more easy, and when contrasted to cash transactions, they increase consumer spending (Akram et al., 2018). This shows that Credit card use has significant effect on Impulsive Buying (Pradhan et al., 2018). In the research conduct by (Kumbhar, 2011), shows result that 28 percent of Credit Card User in e-payment were satisfy in the service quality dimensions namely ease to use, informativeness and etc. This reveals the existence of a relationship between Web quality and online impulsive purchasing behavior, which is moderated by credit card use. Similarly, in the research conducted by (Akram et al., 2018), states that in investigates the moderating behavior of credit card usage to determine whether or whether the existence of credit card also effects the OIBB. As well as website quality, it is discovered that credit card use has a positive and substantial influence on online impulsive purchasing (Akram et al., 2018). Based on the explanation, the hypothesis in this research describe as follows: H5: Credit card use has positive significant moderate the relationship between website quality and online impulse buying behavior.

METHODS
Research Characteristics and measurement. In this research the analytical method that used is the quantitative method and the type of purpose of the research used are descriptive and causal research. The Likert scale were used in this research. According to (Indrawati, 2015), The Likert scale is an example of an interval scale since it assesses how strongly respondents agree or disagree with things using a five-point scale. The five-point scale anchored by the following anchor are one for Strongly disagree, two for Disagree, three for Neither agree nor disagree, four for Agree and five for Strongly Agree measured by (Sekaran and Bougie, 2016).
Population, Sample, and Sampling Method. The population in this research is Indonesian citizen who have been shop in Shopee website by using credit card as the payment method. However, the population of Shopee website user who have shopped using credit card is unknown. Since the population is unknown, the formula that used in this research to find the sample is Cochran formula. Which the formula will be described as follows: In this study, the hypothesis testing used is a one-tailed test. So, the z used is 1.650. the used is 0.500 while the formula to find is is (1-) therefore the used is 0.500. the confidence level used is 95 percent where the value of significant level is 0.05 and the error rate is 0.050. then, the findings of the computation of the number of research sample as follows: Based on the calculation results that the total sample that needed in this research amounted 272.25 and it was rounded up to 273 respondents were needed in this research. As a result, it is feasible to suppose that the sample size in this study is as many as 275 respondents to minimize errors and enhance the outcomes of this investigation. In this study, the non-probability sampling that will be used is purposive sampling. Four criteria have been determined to be respondents, such as: participants are Indonesia Shopee user, participants are female and male with the range age 16 to 54 years old, participants have used Shopee website for at least three times, and participants used credit card in transaction.

Data Collection method.
In this research, primary data sources can be obtained from questionnaires related to this research and will be distributed in Indonesia through social media using online google form. The questionnaire items raised by the author are from (Li et al., 2017), (Firdausy and Fernanda, 2021), (Badgaiyan and Verma, 2015), (Wells et al., 2011), (Akram et al., 2018), (Christino et al., 2019), (Clark et al., 2013), , (Pradhan et al., 2018), (Zhang et al., 2021). Table 1 shows the operationalization variable that the author uses in this study.

Data Analysis Techniques
Descriptive Analysis. In this study the research use Likert scale interval that analyzes how strongly respondents agree or disagree with the five scale choices contained in the question. Therefore, to determine the interval group in which each evaluation of the respondents answer results will be updated as a percentage as follows (Riduwan, 2012): (1) The cumulative value of each statement, is the answer of all respondents, is the overall value of each statement. (2) Percentage is calculated by multiplying an item's cumulative value by its frequency value by one hundred percent. (3) In this study the number of respondents is 275with the largest measurement scale value is five and the smallest measurement scale is one. Therefore, the calculation will be: The largest cumulative number = 275 x 5 = 1375 and the smallest cumulative number = 275 x 1 = 275, the largest percentage value is 100 percent and the smallest percentage value = (275:1375) x 100 percent = 20 percent; Value range = 100 percent -20 percent = 80 percent. If the measurement scale's range value is divided by five, the percentage interval value is 16 percent, and the scoring requirements are shown in Table 2.

Structural Equation Modeling.
Modelling that suits in this study is variance-based matrix structural equation modelling (VB-SEM) with PLS is used as statistical analysis. The gathered data will be processed using the SmartPLS 3.0 software. According to (Indrawati, 2017), to process data using PLS, two steps must be taken: assessment of the measurement model or outer model, and evaluation of the structural model or inner model.

Respondents Characteristics.
During the data collecting process the author managed to collect as many there hundred one respondents. Since only 275 respondents were valid to fulfill the criteria on this research this research. Then, data processed will be 275 respondents. In the data collection period starting from January 2022 and ending in February 2022. In Table 3, Table 4, and Table 5 where will be explained the characteristics of respondents based on gender, age, and occupation. As shown in the Table 3, in this research the involvement of Female respondent was 245 out of 275 respondents or in percentage equivalent to 89.09 percent respondent who have shopped in Shopee Website at least three times by using Credit Card Use as payment method. As shown in Table 4, the most respondent contribution in this research came from age 25 until 34 years old with 116 out of 275 respondents or in percentage same with 42.18 percent on this research. This shows that Descriptive Analysis Result. Descriptive analysis in this research provides to explain the variable in research. Where the independent variable in this research is Web Quality and followed by the dependent variable is Online Impulsive Buying Behavior and the moderating variable in this research is Sales Promotion and Credit Card Use. The explanation of each variable is listed as Table 5, Table 6, Table 7, and Table 8.
Web Quality. The result of descriptive analysis on the item variable Web Quality indicates that the variable Web Quality categorized as Good. According to (Akram et al., 2018), state that five website features have a substantial impact on user perceptions regarding retail websites namely product information, convenience of use, currency, trust, and entertainment. In Web Quality variable the highest is 76.070 percent, which explain that respondents agree that Shopee website is easy to operate. Sales Promotion. The result of descriptive analysis on the item variable Sales Promotion in Table 6 indicates that the variable Sales Promotion categorized as Good. According to (Rahmadhani, 2015), Sales promotion is an activity carried out by sales team to persuade specific target market segments to purchase products through activities other than personal selling, advertising, guerrilla marketing, and public relations that stimulate consumer and dealer purchase effectiveness. The highest item of sales promotion is 73.09 percent where it explains that respondents agree that Shopee provides attractive discounts.

Credit Card Use.
As shown in the table 7, result of descriptive analysis on the item of variable Credit Card Use were categorized as Good Enough. Credit cards are a type of plastic money that is becoming increasingly popular among customers (Akram et al., 2018). By using credit cards makes shopping more convenient and painless (Akram et al., 2018). The highest score among the other items which is 65.02 percent where it describes that the respondents agree with using a credit card as a payment method when making transactions on the Shopee Website.  According to (Fernanda, 2019), impulse buying, also known as unplanned purchasing, is a type of shopping behavior in which a person does not plan ahead of time what they will buy. In the Online Impulse Buying Behavior variable item with the highest percentage is 73.53 percent. Where it indicates that respondents agree that purchasing products on the Shopee website is of their own volition.

SEM-PLS Analysis
Assessment of the Measurement Model . The findings of the outer model testing in this study are shown in Figure 2 above. SmartPLS 3.2.9 software was used for outer model testing in this study. The purpose of this outer model testing was to validate the validity and reliability of the data collection technology by employing all of the core data (Indrawati, 2017).

Figure 2. Measurement Model
Convergent Validity. Convergent validity is used to measure the level of accuracy of an item or set of items in a variable against what you want to measure (Indrawati, 2017). In testing convergent validity, it can be use from two indicators namely, Loading Factor (LF) and Average Variance Extracted (AVE). The characteristics of the loading factor value is the value must be more than 0.700, it might imply that an item in a factor has excellent convergent validity or valid (Indrawati, 2017). As in table 9 below, all items have a value above 0.7 which indicates that all items in this study are valid. In using the AVE value, a variable can be said to have convergent validity if the AVE value is above 0.500 (Indrawati, 2017). In this research, all the variable constructs show an AVE score greater than 0.500.  (Indrawati, 2017).  (Indrawati, 2017), indicate that a variable has discriminant validity if the value of the variable's square root AVE is larger than the correlation between the variables in the model. There are criteria for discriminant validity, including the requirement that the cross-loading value which also indicates the size of the correlation between each variable and the items, is bigger than the correlation value with other variables (Indrawati, 2017). This shows that the variable meets the criteria of discriminant validity. In Table 11 below shows the value of cross loading in this study that indicates valid of discriminant validity.
Reliability Test. The amount of confidence, dependability, consistency, or stability of a measurement's results is referred to as its reliability. One of the qualities of an excellent measuring tool is its dependability (Indrawati, 2015). Reliability is used to calculate how much the indicator variable grows as the latent variable grows (Indrawati, 2017). Cronbach's Alpha (CA) and Composite Reliability (CR) were used to assess reliability, with research criteria CA and CR having values more than or equal to 0.700 (Indrawati, 2017). In this study, all variables have value of Cronbach Alpha (CA) and Composite Reliability (CR) score greater than 0.070. That means reliability in this study has good criteria where the score of Cronbach's Alpha (CA) and Composite Reliability (CR) is greater than 0.070.

Assessment of the Structural
Model. The Inner model measurement to assess the influence of one latent variable on another latent variable (Indrawati, 2017). As shown in Figure 3 depicts an illustration of the inner model used in this study and ZX1 can be explain as moderating variable where sales promotion moderate the relationship between web quality and online impulse buying behavior and XZ2 explain as Credit Card Use moderate the relationship between web quality and online impulse buying behavior. There are two outputs in the assessment of the structural model testing by using Partial Least Square (PLS): 2 for latent variables and parameter coefficients and t-statistics. In addition, t-values in study were calculated using PLS bootstrapping. The t-value test is used to determine if the independent variable has a significant influence on the dependent variable (Indrawati, 2017). This research used 5 percent significant level, which means that if the t-value greater than 1.65 and p-value is smaller than 0.050 means that there is a significant influence between the independent variable and dependent variable. Testing the value of R 2 is used to measure the level of variation of changes in the dependent variable on the independent variable. The higher the R-square value, the better the model can predict the research object. The criteria for R-Square values are 0.670, 0.330 and 0.190 which indicate that the model is "Good", "Moderate" and "Weak" (Indrawati, 2017).

DISCUSSION
Based on the result that has been found, there are five findings in this research. First Website Quality (WQ) has the strongest positive effects towards Online Impulsive Buying Behavior (OIB). Second finding is that Sales Promotion (SP) positively affect OIB. Third finding is that Credit Card Use (CCU) has positive effect on OIB. Additionally, we found the fourth finding that SP doesn't have moderating effect on the relationship between WQ and OIB. The last finding is that CCU doesn't act as moderator for the relationship between WQ and OIB.
First finding is in line with the results of research conducted by (Akram et al., 2018), (Firdausy and Fernanda, 2021), (Widagdo and Roz, 2021). Web Quality is a factor that is the main influence for OIB found in this study. The quality of the web site such as in visual appearance, convenience and security in general will affect the behavior of buying back, whether impulsive or not. Website visitors who have good quality will be more likely to make a purchase. For example, a website that is difficult to use to find product information or make a purchase will prevent potential buyers from making a purchase because they feel they do not have sufficient control to complete the process. Perceived control is an important factor that affect behavior.
Second finding is in line with previous research conduct by (Akram et al., 2018), (Mutanafisa and Retnaningsih, 2021), and (Firdausy, 2021). Promotion is an important trigger in impulse buying. Impulse buying is generally not done in a planned manner. The existence of attractive discounts for a limited time will make potential buyers who previously had no interest to make a purchase. This behavior is because they think that they need to take advantage of these attractive promotional opportunities. Indonesian marketplace is a highly competitive environment where each marketplace needs to offer competitive advantage such as lowering cost by giving discount or other benefit related to customer cost. This situation also could be one possible explanation on this research finding.
Third finding on the effect of CCU on OIB, supports previous research by (Putra, 2015), (Akram et al., 2018), (Badgaiyan and Verma, 2015), and (Khare, 2013). High use of credit cards is a strong indicator of financial capability and adequate experience of making electronic payments. Both will trigger online purchases. Buyers with sufficient financial ability and experience will generally view the buying process through the marketplace as having a lower risk and encourage them to make purchases, including impulse purchases.
There are two interesting findings from this research related to the moderation effect of CCU and SP. Despite of expected results, sales promotion and Credit Card Use don't significantly moderate the relationship between web quality and Online Impulse Buying Behavior. The results of this study on Sales Promotion moderation effect have similarities with the results of research by (Akram et al., 2018). One of possible explanation according to (Akram et al., 2018), state that sales promotion has a negative effect on sales due to purchase acceleration and stockpiling, as well as a negative impact on brand image and reference price.
Credit Card Use also found not to have significant moderating effect. This result support (Engeln and Joseph, 2012) in (Abrar et al., 2017). Possible explanation offered is that it could be due to Online shoppers are anxious that their payment card information will be compromised or that they will be overcharged by sellers. Financial risk also refers to the possibility of not receiving the lowest feasible price for the chosen goods from a certain online business. In this research also have similarities respondent where it dominates by woman and age group twenty six until thirty years old.
The findings in this research indicate that Shopee marketplace needs to improve WQ, SP and CCU to increase OIB. Nevertheless, they need to focus on WQ, since this variable is the strongest factor that affect OIB. One of the main indicators that Shopee needs to focus indicated in Table 6. Based on that table, the less perceived quality indicator is CM2 which related to website usability. Shopee needs to improve the process and website interface to gain better perceived quality on this indicator.

CONCLUSION
This research found that web quality has positive significant influence on Online Impulse Buying Behavior. It also found that Sales Promotion and Credit Card use has positive significant influence on Online Impulse Buying Behavior. Sales promotion and Credit Card Use do not significantly moderate the relationship between web quality and Online Impulse Buying Behavior. Despite the unexpected results of the moderation effect, this model proven to be useful to explain the nature of impulsive buying in a marketplace affected by Web Quality, Sales Promotion and Credit Card Use.
Shopee can improve further in terms of web quality, sales promotion, and their credit card payment method at Shopee Website especially in the statement that has lowest percentage of response. One of them that Shopee can provide various types of cash back offers to users with the freedom to use cashback on all products. Because with that the user will be more motivated to use cashback in making shopping transactions in large quantities. And Shopee also can provide offers that can attract the attention of credit card users to use credit cards when shopping at Shopee, namely by providing discounts on buying through credit cards or on large shopping. That way users will feel more compelled to use a credit card because users find it more profitable and easier to use a credit card with the benefits provided by e-commerce itself.