THE INFLUENCE OF SYSTEM QUALITY, INFORMATION QUALITY, AND SERVICE QUALITY ON USER SATISFACTION THROUGH USE IN THE BLIBLI MARKETPLACE

: This study aims to determine what factors influence the success of the Blibli application using the Delone & McLean model. The variables used in this research are system quality, information quality, service quality, user activity, and user satisfaction. The population in this study are marketplace users who have used and shopped at the Blibli application, with a sample size of 190 respondents. The data collection technique used a questionnaire, and the data analysis technique used was the SmartPLS analysis tool. Based on the results of the study, of the ten hypotheses, seven hypotheses are accepted, namely information quality and service quality have a positive effect on user activity, system quality, service quality, and user activity have a positive effect on user satisfaction, information quality and service quality have a positive effect on user satisfaction through user activity. The contribution to this research resulted in a model for measuring the effectiveness of the Blibli application system that companies can use to develop applications in the future to increase user satisfaction with the Blibli application.


INTRODUCTION
The Marketplace Marketplace is one of the digital advances experienced by society today.A marketplace is a place to make it easier for sellers and buyers to buy and sell online through internet media (Apriadi and Saputra, 2017).The existence of a marketplace can help business actors to sell their products to interested consumers.On the other hand, consumers can search for the goods they need with various product choices such as brands, colours, sizes, and others just by using cellphones and other internet tools.Marketplace in Indonesia is experiencing development from year to year.Shopping transactions through the Marketplace Marketplace have increased since 2019, which reached Rp. 265.07 trillion.Although 2020 decreased to Rp. 253 trillion, the transaction value increased by 33.2% to Rp. 337 trillion in 2021 (Aseng and Pandeirot, 2022).
According to data from DataIndonesia.id on September 22, 2022, transactions through the MarketplaceMarketplace in 2022 experienced a significant increase to Rp. 842.3 trillion.The findings from We are Social show that the ratio of marketplace users in Indonesia reached 88.1% and named Indonesia the country that uses the MarketplaceMarketplace the most (Artasya and Nuri, 2022).Marketplace has several advantages, so it is in great demand by online buying and selling actors.The advantages of marketplaces are that they are more effective in terms of time and more efficient in terms of cost (Kasmi and Candra, 2017).Marketplaces that have been present in Indonesia include Tokopedia, Shopee, Lazada, Bukalapak, Orami, Blibli, and others.
Blibli is one of Indonesia's growing marketplaces, based on data from the I-2022 quartile price.Co.id Blibli ranks sixth as the Marketplace Marketplace with the highest number of visitors in Indonesia.Blibli's monthly visitors reached 16.3 million in the I-2022 quartile.Based on sources from Tribunnewswiki.com on April 11, 2022, Blibli partners offer electronic products, fashion needs, home appliances, etc.The Blibli application not only sells products from cooperating partners, but Blibli also sells its products.In addition, the Blibli application also provides service features such as Blibli Mitra.This service feature is for micro-businesses that are running small businesses.Furthermore, this feature makes Blibli Pay later easier for buyers and sellers because buyers can pay for goods based on a predetermined period.Blibli Express Service, this feature can be called Blibli's internal courier.Blibli Mart, this feature offers shopping categories and provides household products.Blibli also provides 24-Hour Customer Care services.Customer Care services to assist users in consulting about application products and services.Although Blibli has provided good and complete services to its consumers, consumers still feel many complaints.Based on sources from Mediakonsumen.com on December 23, 2022, Blibli Application user complaints such as unclear Blibli delivery and poor Customer Service service, orders cancelled for no apparent system reason, and the goods sent did not match those ordered.
System Quality is very important for User Satisfaction.According to Rakhmadian et al. (2017), if consumers feel that the system is easy to use and does not need much effort, it will impact User Satisfaction.System Quality can be measured from several aspects, namely ease of Use, access speed, and system security (Natalia and Br Ginting, 2018).According to Hernita et al. (2020), System Quality measurement focuses on the results of interactions between users and systems.In addition to System Quality, Information Quality significantly affects purchasing decisions.The information presented in the online shop includes information related to the products sold on the online shop.Reliable information is useful for assessing the quality and usefulness of the products sold.
To satisfy customers in online shopping, product information must be up-to-date.This can help consumers purchase (Ayuningtiyas and Gunawan, 2018).(Iskandar and Nasution, 2019).Service Quality is something that every company pays attention to.Quality Service Quality will meet consumer expectations and impact User Satisfaction (Cesariana et al., 2022).If consumers receive appropriate service or exceed expectations, then consumers will feel satisfied (Sigit and Soliha, 2017).According to Herlistyani et al. (2012) in Armanto (2018), there are efforts to give and receive information from each other in the service process.If the information provided and received by consumers provides good benefits, then consumer User Satisfaction will increase.Services can be known by comparing consumer expectations with the services consumers receive.If the service received or felt by consumers follows expectations, then Service Quality can be called good and high-quality service.However, if the service consumers receive is not as expected, the service is referred to as poor and lowquality Service Quality.Service is essential to attract customer interest and will affect sales (Pane et al., 2018).Purwati et al. (2021) found that System Quality positively affects User Satisfaction.This research was supported by Hidayatullah et al. (2020).This contradicts the research of Mangun Buana and Wirawati (2018)

RESEARCH METHODS
From the hypothesis above, the research model can be described as follows: The research design used in this study is casual associative research which is useful for analyzing the relationships between one variable and another or how one variable affects another.This study aims to determine the direct influence of System Quality, Information Quality and Service Quality on Use, to find out the direct influence of System Quality, Information Quality, Service Quality, and Use on User Satisfaction and to find out the indirect influence of System Quality, Information Quality and Service Quality on User Satisfaction through Use on the Blibli Marketplace.The research shorthand used in this study is a quantitative shortcoming in this study using primary data.Primary data is data that has never been processed before by others.Data collection in this study used a survey with a questionnaire.The survey method is a way or method of collecting data by sharing questions with respondents by distributing questionnaires.The distribution of questionnaires using google forms through social media such as WhatsApp, Instagram, Facebook, and telegram.
In this study, to collect data, a questionnaire was carried out.In this study, the preparation of a questionnaire using a Likert scale.The Likert scale is a research scale used to measure attitudes and opinions.This study used four levels of the Likert scale, namely Strongly Disagree (STS), Disagree (TS), Agree (S), and Strongly Agree (SS).This study used five variables: System Quality, The population in this study were marketplace users who had used and shopped on the Blibli application.According to Hair et al. (2019), the number of samples must have five times the number of questions analyzed to get a tangible result.There were 38 questions in the questionnaire in this study, so the minimum number of samples required was 190 respondents.This research will use the purposive sampling method.According to Sugiyono (2018), purposive sampling aims to obtain samples that match the researchers' criteria.Criteria predetermined criteria are.(1) have used and shopped on the Blibli application more than two times in the last year, (2) have been Active users of the Blibli application in the last year, and (3) are in the Jabodetabek area.
In the research conducted using the Smart PLS analysis tool wherein analyzing the data, there are several stages, namely, the first stage is the outer model where the validity test is carried out (1) Convergent Validity Test with a standard loading factor value of > 0.60 is the ideal size (Hair et al., 2019) (Ghozali, 2016).There is a criterion of the t-statistical test (Ghozali, 2016), namely (1) if the value on the construct of the significance of the t-test is > 0.05, then Ho is accepted, and Ha is rejected so that it can be interpreted that there is no influence between variables, (2) If the constructed value of the significance of the t-test is < 0.05 then Ho is unacceptable, and Ha is acceptable so that it can be interpreted as an influence between variables.

Characteristics of Respondents
Based on the results of the questionnaires distributed online using Google Forms, 190 respondents were collected.The respondents who use and shop the most through the Blibli application are women, with 147 people (77.4%) and the most aged 17 -22 years, which is 157 people (82.6%).Furthermore, the most respondents with unmarried status were 170 (89.5%) and domiciled in the Tangerang area, which was 100 people (52.6%).Then, most respondents had the status of high school / vocational high school students, namely 157 people (82.6%) and working as students/students, which amounted to 147 people (77.4%).Moreover, most of them have used the Blibli application for > 1 year, which is 96 people (50.5%).Most respondents have shopped through the bible application 1 -2 times a month, which 97 people (51.1%).Moreover, the most purchased products by respondents were accessories and fashion products, totalling 93 people (48.9%).Based on the results, it can be concluded that many respondents are women aged 17-22 years.They are generation Z who like online shopping, are always connected to the virtual world and can do everything using the sophistication of existing technology.(Results can be seen in Appendix 6).

Outer Model Validity Test
This study tested a validity test that is useful for knowing a measurement of a statement against the questionnaire used in this study which aims to measure the indicators and variables studied.The statements that will be tested in this study The invalid indicators are SQL1 with a constructed value of 0.518, SQL3 with a constructed value of 0.486, SQL5 with a construction value of 0.590, IQ1 with a construction value of 0.660, IQ4 with a construction value of 0.611, IQ5 with a construction value of 0.655, IQ9 with a construction value of 0.652, and U1 with a value of 0.599 the indicator must be omitted.(Results can be seen in Appendix 6).

Average Variance Extracted (AVE)
The AVE results can prove the latent variable values' ability to represent the original data score.The greater the AVE value, the higher the ability to explain the value of indicators that measure latent variables.The standard provision of the AVE value always used is 0.50, where the AVE value is at least 0.50 proving the convergent validity value in this study has a good value.

Table 1. Average Variance Extracted (AVE)
Source: Data processed by author, 2023 The Average Variance Extracted (AVE) value in the table above shows that all latent variables have an AVE value above the minimum criteria, which is 0.5, so it can be concluded that the AVE value in this study is good.

Discriminant Validity Cross Loading
Discriminant validity is a measurement model where reflective indicators are assessed from the results of cross-loading tests on the measurement of a construct.Suppose the correlation in one construct of a measurement item is greater than the dimension of another construct.In that case, it can be concluded that the result can prove that a latent construct can predict the dimensions of a block better than the dimensions of another block.(Results can be seen in Appendix 6).

Reliability Test
The reliability test in this study aims to find out whether it is true that a measuring instrument has consistency (can be used at a later stage).The reliability test results use the measurement provisions of Composite Reliability and Cronbach's Alpha.

Inner Model
The next stage is the Inner model.By determining the R-square value, if the R-Square value is at 0.75, 0.50, 0.25, it can be concluded that the inner model is strong, moderate and weak.Furthermore, for hypothesis testing, T-Statistical standards were used with values of >1.96 and ρ-Value <0.05 and the next stage determined the fixed model with an SRMR value of <0.1 from the results of the study obtained an Rfair value on the Use variable of 57.8% this can be said that the magnitude of the influence of System Quality, Information Quality, and Service Quality on Use by 57.8% and the remaining 42.2% is influenced by other variables such as ease of Use (Ease Of Use), Fullfillment, and Privacy.The R-square value on User Satisfaction is 72.7%.Based on these results, it can be interpreted that the magnitude of the influence of System Quality, Information Quality, Service Quality, and Use on User Satisfaction is 72.7%, and the remaining 27.3% is influenced by other variables that were not studied in this study such as Product Quality, Price, and Brand Image.(Results can be seen in Appendix 6).

Model Fix
A fixed model is a value that shows the degree of conformity for the whole in a model that can be calculated with the residual value of a predicted model by comparing it with the actual data.System Quality does not affect Use regarding the T value of 1.510 < 1.96 and the significant value ρ -Value 0.250 > 0.05.This shows that System Quality is not a variable that determines user activity (Use) on the Blibli application.However, Information Quality positively affects Use regarding the T value of 3.194 > 1.96 and the significant value ρ -Value 0.001< 0.05.This shows that the higher the Information Quality provided to Blibli application users, the higher the user activity (Use) on the Blibli application.The results further stated that Service Quality also positively affected Use regarding significant values of ρ -Value 0.000 < 0.05 and statistical T values of 4.362 > 1.96.This shows that the higher the Service Quality obtained by Blibli application users, the higher the user activity (Use) on the Blibli application.
System Quality positively affects User Satisfaction regarding the significant value of ρ -Value 0.033 < 0.05 and the T value of statistics 2.133 > 1.96.This shows that the higher the System Quality in the Blibli application, the more User Satisfaction will increase using the Blibli application.However, Information Quality does not affect User Satisfaction regarding significant values ρ -Value 0.154 > 0.05 T and statistical values of 1.427 < 1.96.This shows that Information Quality is not a variable that determines User Satisfaction when they use the Blibli application.In contrast, Service Quality has a positive effect on User Satisfaction seen from the significant value of ρ -Value 0.000 < 0.05 and T value of Statistics 3.510 > 1.96.This shows that the higher the System Quality in the Blibli application, the more User Satisfaction will increase using the Blibli application.The results further stated that Use positively affected User Satisfaction regarding the significant value of ρ -Value 0.002 < 0.05 and the statistical T value of 3.111 > 1.96.This shows that the higher the Use of the Blibli application, the more User Satisfaction will also increase in using the Blibli application.System Quality positively affects User Satisfaction regarding the significant value of ρ -Value 0.033 < 0.05 and the T value of statistics 2.133 > 1.96.This shows that the higher the System Quality in the Blibli application, the more User Satisfaction will increase using the Blibli application.However, Information Quality does not affect User Satisfaction regarding significant values ρ -Value 0.154 > 0.05 T and statistical values of 1.427 < 1.96.This shows that Information Quality is not a variable that determines User Satisfaction when they use the Blibli application.At the same time, Service Quality has a positive effect on User Satisfaction seen from the significant value of ρ -Value 0.000 < 0.05 and T value of Statistics 3.510 > 1.96.This shows that the higher the System Quality in the Blibli application, the more User Satisfaction will increase using the Blibli application.The results further stated that Use positively affected User Satisfaction regarding the significant value of ρ -Value 0.002 < 0.05 and the statistical T value of 3.111 > 1.96.This shows that the higher the Use of the Blibli application, the more User Satisfaction will also increase in using the Blibli application.

Hypothesis
Hypothesis Statement

H1
System Quality has a direct positive effect on the Use

H7 accepted
Source: Data processed by author, 2023 Based on the analysis results from the table above, it can be concluded that out of the seven hypotheses, five are declared accepted, and two hypotheses are declared rejected.The 5 accepted hypotheses are H2, H3, H4, H6 and H7, while the rejected hypotheses are, H1 and H5 are declared rejected.

Indirect Effect
The next stage is to find out the indirect influence between System Quality, Information Quality, and Service Quality on User Satisfaction through Use, which can see in the following Indirect Effect table:

CONCLUSION
Based on the study's results, System Quality does not affect Use, which means that it can be said that System Quality is not a factor consumers consider when using the Blibli application.While Information Quality positively affects Use, the higher the Information Quality provided to users, the higher the user activity (Use) on the Blibli application.Service Quality positively affects Use.This shows that the higher the Service Quality obtained by Blibli application users, the higher the user activity (Use) on the Blibli application.
System Quality has a positive influence on User Satisfaction.This shows that the higher the System Quality in the Blibli application, the more User Satisfaction will increase using the Blibli application.However, Information Quality does not influence User Satisfaction, which shows that Information Quality is not a variable that determines User Satisfaction when using the Blibli application.Meanwhile, Service Quality has a positive influence on Figure 1.Research Models Source: author-processed (2023) are 38 statements.The validity test in this study is as follows: Convergent Validity Outer Loading (Loading Factor) Of all the results that have been analyzed in the Loading Factor table, a questionnaire indicator totalling 38 statements, there are 30 valid indicators, and eight indicators declared invalid.Source: Data processed by author, 2023 Figure 2. Outor Model The Influence of System Quality, Information Quality, and Service Quality on User Satisfaction Through Use in The Blibli Marketplace Page 190 Asian Journal of Engineering, Social and Health Volume 2, No. 3 March 2023 Information Quality, Service Quality, User Satisfaction and Use.System Quality is measured by reliability, East Of Use, Response Time, and Visual Display (DeLone and McLean, 2003).

,
Average Variance Extracted (AVE) with a Cut-off value of AVE > 0.50.continued (2) The Discriminant Validity Test, judging from the cross-loading value where a value in the standard relation of the construct with measurement items is greater than other construct sizes, can show the results of latent constructs predicting the size on the block better than other block sizes.Meanwhile, the reliability test will use Cronbach alpha and composite reliability measures, with a reliable value standard of > 0.70 (Ghozali, 2016).The next stage is the Inner model by determining the value of the R-square.If the value of the R-Square is at the value of 0.75, 0.50, 0.25, it can be concluded that the inner model is strong, moderate and weak.T-Statistics is a value that is useful to see at the significance level to hypothesis testing by looking for T-Statistics values through the bootstrapping procedure on Smart PLS.Hypothesis testing starts by calculating the path coefficient, then compared with the T-statistical value > T of the table.A hypothesis is acceptable if a statistic T value can be greater than the table T of 1.96 (α 5%) so that it can be interpreted that if a value of T statistic in each hypothesis can be greater than T, the table can show that the hypothesis is accepted or proven

Table 2 .
Reliability Test

Table 4 .
The following are the results of his research: Hypothesis Test1

Table 6 .
Indirect Effect3 The Influence of System Quality, Information Quality, and Service Quality on User Satisfaction Through Use in The Blibli Marketplace Page 202 Asian Journal of Engineering, Social and Health Volume 2, No. 3 March 2023 User Satisfaction.This shows that the higher the System Quality in the Blibli application, the more User Satisfaction will increase using the Blibli application.Furthermore, Use has a positive effect on User Satisfaction, which, if Use is increased, will also increase users' satisfaction with the Blibli application.System Quality does not have an indirect influence on User Satisfaction through Use.Use cannot mediate the System Quality variable.Meanwhile, Information Quality and Service Quality indirectly influence User Satisfaction through Use.If Information Quality and Service Quality are improved, it will increase Use and impact increasing User Satisfaction.This research does not prove the direct influence of System Quality on Use and Information Quality on User Satisfaction in the Blibli application.The results of this study also state that Service Quality has the most influence in increasing user activity (Use) and User Satisfaction, so that this can be said that Service Quality in the Blibli application must be further developed and improved in the future in order further to increase user activity (Use) and User Satisfaction.