The Impact of Perceived Usefulness and Perceived Ease-of-Use toward Repurchase Intention in the Indonesian E-Commerce Industry

This research was conducted to understand the impact of perceived usefulness and perceived ease-of-use on repurchase intention in the Indonesian e-commerce industry. A total of 400 respondents participated in this research, in which, collected data were further analyzed using PLS-SEM method. SmartPLS 3.2.7 software was used to analyze and generate the results required for the concluding this research. Based on the results of the data analysis, it could be concluded that both perceived usefulness and perceived ease-of-use had a positive impact on Indonesian consumers’ repurchase intention in the e-commerce industry, both directly and indirectly through trust. Based on these result, author would like to suggest that companies practicing their business in the Indonesian E-Commerce Industry to make upgrade their system to be more simpler yet understandable regarding the website or the apps that the company develop in order to reduce the level of confusion between the users, while at the same time, companies should also create a website or system in which people will perceive that such an app or e-commerce website would be beneficial in helping them in enhancing their daily activities, thus pushing them to learn and use the system. To author’s best knowledge, this is one of a few studies which conduct an assessment regarding the relationship between variables in the e-commerce industry using PLS-SEM approach. Keywords— E-Commerce Industry, Perceived Ease-of-Use, Perceived usefulness, Repurchase Intention, Trust Abstrak Penelitian ini dilakukan untuk mengetahui pengaruh perceived usefulness dan perceived ease-of-use terhadap repurchase intention pelanggan pada industri e-commerce di Indonesia. Penelitian ini menggunakan metode survey, dimana, sebanyak 440 responden berpartisipasi dengan mengisi kuesioner yang diberikan. Setelah mengumpulkan seluruh data dari responden, data tersebut kemudian diolah dengan menggunakan metode PLSSEM. Software SmartPLS versi 3.2.7 digunakan untuk menganalisis dan menemukan hasil akhir dari pengolahan data ini akan digunakan untuk menyimpulkan hasil dari penelitian ini. Berdasarkan hasil analisis data, peneliti dapat menyimpulkan bahwa kedua variabel independen, yaitu perceived usefulness dan perceived ease-of-use berpengaruh positif terhadap repurchase intention, baik secara langsung maupun secara tidak langsung melalui trust. Berdasarkan hasil ini, peneliti ingin menyarankan kepada seluruh perusahaan yang bergerak di industri e-commerce di Indonesia untuk dapat menigkatkan kualitas sistemnya, dimana, system yang diimplementasikan oleh perusahaan haruslah mudah untuk dipelajari dan dimengerti oleh masyarakat dengan tujuan agar masyarakat tidak bingung di dalam menggunakan system yang dikelola oleh perusahaan. Selain itu, peneliti juga ingin menyarankan kepada perusahaan untuk dapat menciptakan suatu sistem e-commerce yang dimana sistem tersebut akan dipersepsikan sebagai suatu sistem yang akan berguna bagi masyarakat di dalam membantu mereka untuk melakukan aktivitas sehari-hari, sehingga, masayarakat akan belajar dan menggunakan sistem yang dikelola oleh perusahaan e-commerce tersebut. Berdasarkan hasil studi yang dilakukan oleh peneliti, penelitian ini merupakan satu dari sedikit penelitian yang bertujuan untuk mengetahui hubungan antar variabel melalui metode PLS-SEM pada industri e-commerce di Indonesia. Kata kunci— Industri E-Commerce, Perceived ease-of-use, Perceived usefulness, Repurchase Intention, Trust Wilson Jurnal Manajemen Indonesia (19(3), 241249, 2019)


I. INTRODUCTION
In the current era fueled and marked by globalization and rapid development of ICT, many business practitioners have decided to push and develop their business to the next level that they've never imagined before. In the other words, it could be stated that globalization had transform the current business activities from physical business into digital business. For example, in the past, physical stores were often acting as criteria which should be fulfilled in order for a person or organization to establish a business. However, the rapid development of the internet had diminished that criterion, since, people of business practitioners who were involved in the business practices could sell and promote products/services through digital platform over the network (Wilson & Keni, 2018;Wilson, 2018). Such a practice or phenomenon leads the existence of a new business trend called "E-Commerce." The existence and development of the e-commerce business not only had changed the way people are doing business, but also open a chance to several people who would like to kickstart their business, which they couldn't do in the past since they don't have the resources needed to establish a business. Moreover, in regards with the development of the e-commerce industry, not only that many individuals of new companies decide to establish a business through the e-commerce platform, but also many existing companies which decide to integrate or expand their business into the e-commerce platform. For example, a lot of companies who used to sell their products/services via physical stores had also opted to sell the same products or services through the e-commerce platform. Many companies had even developed their own mobile version of their e-commerce website in form of application, in which, people could directly accessing the company's e-commerce page through the application, rather than typing the URL manually on the website.
However, some problems did exist in terms of the development of the e-commerce trends. One of the problems was that many customers still believe that e-commerce is a new technology equipped with so many distinctive features and system which render them difficult to learn. Such a perception was normally possessed by many customers grouped or classified as "Baby Boomers", in which, customers in this group were often categorized as individuals who were slow in terms of their learning and adaptation of a new technology (Kotler & Keller, 2015). Therefore, it is important for companies to be able to always develop a system or technology for their respective e-commerce platform which was easy to be learned by customers (perceived ease-of-use). Furthermore, companies also should be able to convince customers that making some transactions through the ecommerce platform will bring them more simplicity, advantages, and practicality in terms of their buying behavior and activities compared to making a transaction through physical stores (perceived usefulness). When companies were able to develop their e-commerce platform to meet these two objectives, then it could lead customers to make another purchasing activity toward the website. Based on the description above, this research was conducted in order to better understand the impact of perceived ease of use and perceived usefulness toward customers' repurchase intention through trust.

A. Perceived Usefulness
Perceived usefulness could be defined as "the overall assessment and perception on the part of mobile users of the usability of 3G phones" (Amin et al., 2014). Furthermore, Davis et al. (1989) defined perceived usefulness as "the degree to which a person believes that using a particular system would enhance his or her job performance". Therefore, it could be concluded that in this research, perceived usefulness could be defined as customers' judgment and perception on whether the new system that they used will bring them some addedvalue compared to the old system.

B. Perceived Ease-of-Use
Perceived Ease-of-Use could be defined as "users perception of whether performing a particular technical task would require a mental effort on his or her part" (Ajzen & Fishbein, 1980;Rouibah et al., 2011;Amin et al., 2014). Furthermore, Davis (1989) defined perceived ease-of-use as "the degree to which a person believes that using a particular technology would be effortless." Therefore, in this research, perceived ease-of-use could be defined as customers judgement on the amount of effort that customers should spends in learning and using the new technology or system.

C. Trust
Trust could be defined as "individual willingness to depend based on the beliefs in ability, benevolence, and integrity" (Gefen et al., 2003). Moreover, trust could also be defined as customers' willingness to be engaged in future transactions, in which, they're reflecting their confidence in the reliability of the seller's behavior, which will lead to a perception of reduced risk of further have another transactional activities toward the same seller or 243 company (Steenkamp et al., 1998;Japutra, Keni & Bang (2015); Wilson & Keni, 2018;Wilson, 2018). In this research, trust could be defined as customers' perception regarding the extent of risk that might happen during the transactional activities between two or more parties.

D. Repurchase Intention
Repurchase intention could be defined as "individual's judgment about buying a service again, the decision to engage in future activity with a service provider and what form this activity will take" (Chen & Chen, 2017;Wilson andKeni, 2018, Wilson, 2018). Moreover, Moslehpur et al. (2017) defined repurchase intention as "consumers' consideration to purchase again a particular product." In this research, repurchase intention was defined as customers' intention in making another purchasing activities in the future time.

A. Perceived The Impact of Perceived Usefulness Toward Repurchase Intention
Previous research by Wen et al. (2011) shown that perceived usefulness have a positive Impact on repurchase intention. Furthermore, another research by Purnami & Nurcaya (2015) and Chen (2012) also found that perceived usefulness have a positive impact on repurchase intention. Therefore, the first hypothesis was posited as follows. Based on the literature above, author would like to posit the following hypothesis: H1: Perceived Usefulness Had a Positive Impact On Repurchase Intention.

B. Perceived The Impact Of Perceived Ease-Of-Use On Repurchase Intention
Previous research by Aren et al. (2013) had found that perceived ease-of-use had a positive impact on repurchase intention. Furthermore, previous study by Chen (2012) also found that perceived ease-of-use could positively affecting customers' intention on making another buying activity through the same stores. Based on the literature above, author would like to posit the following hypothesis:

C. The Impact Of Trust On Repurchase Intention
Previous research by Oroh et al. (2015) found that trust had a positive impact toward repurchase intention. Furthermore, another research by Aren et al. (2013) also found that trust could influence customer's intention to make another purchase decision from the same company. Further research by Setyorini & Nugraha (2016) and Wilson & Keni (2018) also confirmed that in the e-commerce sector, trust had a positive impact on repurchase intention. Based on the literature above, author would like to posit the following hypothesis:

D. The Impact Of Perceived Usefulness And Perceived Ease-Of-Use Toward Repurchase Intention Through
Trust Previous research by Setyorini & Nugraha (2016) had found that perceived usefulness had a positive and significant influence on customers' repurchase intention through trust. Furthermore, Chiu et al. (2009) also found that customers' trust will disappear when customer started to think that the new system of technology bring them no advantages toward the old method or system, which could alter their intention to re-buy or re-use the system of technology. Based on the literature above, author would like to posit the following hypotheses: This research utilize survey method, in which questionnaire was chosen as the instrument to collect all of the data needed for this research. There are two parts in this questionnaire, in which the first part contains the questions regarding the demographics of the respondents. In the other hand, the second part of this questionnaire contains a total of 16 items which represent all 4 variables analyzed in this research. From the total of 16 items, 3 items represent variable perceived ease-of-use, 3 items represent variable perceived usefulness, 5 items represent variable trust, and 5 items represent variable repurchase intention. A total of 3 indicators representing variables perceived ease-of-use and 3 indicators representing variable perceived usefulness were adapted from Amin et al., (2014), while items representing trust were adapted from Zhou et al., Fang et al., (2014). Items representing consumers repurchase intention was adapted from Zhou et al., Fang et al., (2014). All items had been modified and edited to ensure it's conformity with the topic discussed in this research. Questionnaires were both physically and electronically distributed to all respondents participated in this research.
Furthermore, Interval scale with 5-Point Likert Scale method was also used in this research, in which each respondent was asked to give their respond regarding each statement presented on the second part of this questionnaire, which was consisted of "1" to reflect their strong disagreement toward the statement, "2" to reflect their disagreement toward the statement, "3" to reflect that they are neither agree nor disagree toward the statement, "4" to reflect their agreement toward the statement, and "5" to reflect their strong agreement toward the statement.
Before research was conducted, a pretest research will be conducted beforehand. Malhotra (2007), Sekaran & Bougie (2016), and Wilson (2018) had stated that a questionnaire should and required to be pretested adequately before it could be used on the actual survey. A total of 50 respondents who were drawn from the same population and share the same characteristics as the respondents participated in the pretest. After all of the data were collected, validity and reliability test were conducted in order to ensure that all of the data are valid and reliable, and the questionnaire could be used in the research. After conducting both tests, it was revealed that all items have been deemed as valid and reliable, thus enabling the questionnaire to be used in the research.
Regarding the amount of sample size determined in this research, Hair et al. (2010) stated that a total between 150 to 400 sample size were the right number for a research or study using both structural equation modeling and partial least squares-structural equation modeling approach (PLS-SEM). Moreover, Beavers, Iwata & Lerman (2013) and Wilson (2018) had also stated that for a multivariate research, a total of sample size between 150 to 300 respondents had been deemed appropriate. Furthermore, previous research by Wilson, Theodorus & Tan (2018) and Wilson & Keni (2018) also underlined that for a research, the total number of sample size determined by researchers should exceed 200. Based on the theories stated above, a total of 400 sample size was determined and set for this research, in which this number was twice the amount of sample size suggested by researchers stated above.
A total of 400 respondents participated in this research, in which, respondents who were chosen are those who have board any domestic airlines at least twice a year for the past two years. Furthermore, the data used in this research were collected from respondents who lived or originated from Jakarta. Regarding the sampling method, this research used purposive sampling in order to ensure the objectivity of the research since all respondents participated in this research should fulfill the criteria set beforehand, in which they should have made a purchasing activity on e-commerce websites in the past eight months. The response rate in this research is 100% since all 400 questionnaires were filled and returned. After further assessment was made regarding the response given in the questionnaires, all data generated from the questionnaires were deemed valid and usable. Therefore, data collected from a total of 400 questionnaires were used and processed in this research.
A total of 400 respondents participated in this research, in which, respondents who were chosen are those who have made a purchasing activity on e-commerce websites in the past 2 years. Furthermore, the data used in this research were collected from respondents who live in Indonesia. This research implemented purposive sampling method in order to ensure the objectivity of the research since all respondents participated in this research should fulfill the criteria set beforehand, in which they should have made a purchasing activity on ecommerce websites in the past 2 years. Most of the respondents are male (64.5%), and most of them have made at least 3 purchasing activities toward e-commerce websites each month (61.75%). After conducting the demographics analysis on the respondents, data were then analyzed using partial least square-structural equation modeling (PLS-SEM) using SmartPLS 3.2.7 software in order to determine the validity and reliability of the model, and to determine the relationship between variables.

V. RESULTS AND DISCUSSIONS
All of the data from the questionnaires were processed using Partial Least Squares-Structural Equation Modeling (PLS-SEM) method by using SmartPLS 3.2.7 software. Using PLS-SEM method, the outer model and inner model measurement needed to be conducted to determine the relationship between variables and to test the hypotheses proposed in this research.
Before testing the hypotheses in this research by conducting the inner model analysis, the outer model analysis need to be assessed beforehand in order to confirm the validity and the reliability of the model. In the outer model analysis, several criteria needed to be fulfilled in order for a model to be classified as valid and reliable, and in order for the inner model analysis to be conducted. Several criteria that required to be fulfilled are : The value of AVE should be above 0.5 and the value of AVE of each latent construct should be higher than the construct's highest squared correlation with any other latent construct (fornell-larcker criterion) as suggested by Hair, Ringle & Sarstedt (2011), Wilson, Theodorus & Tan (2018), and Wilson (2018), the factor loadings value of each indicators should exceed 0.5 as suggested by Barclay and Thompson (1995) and Chin (1998), and the value of both composite reliability and cronbach's alpha should be greater than 0.7 as suggested by Ebert (2010) and Peterson (1994). Only after the model was confirmed to have passed the outer model analysis (which indicate that the model had deemed valid and reliable) that the inner model analysis could be further conducted in order to find out and assess the relationships between variables, the mediating effect of mediating variables, and to test whether or not hypotheses were supported.
First of all, Outer model analysis was conducted afterwards in order to determine and confirm the validity and the reliability of the data and model. In conducting the outer model analysis, several criteria required or needed to be fulfilled in order for a model to be classified as valid and reliable. Those criteria are : The value of AVE and composite reliability of each variable should be greater than 0.5 and 0.7 respectively (Henseler, Ringle & Sinkovics 2009;Ebert, 2010), the factor loadings value of each items should exceeded 0.50 as suggested by Barclay, Higgins & Thompson (1995) and Chin (1998), and the value of AVE of each latent construct should be higher than the construct's highest squared correlation with any other latent construct (fornell-larcker criterion) as suggested by Hair, Ringle & Sarstedt (2011). Based on the results of the outer model analysis, it was revealed that the model had fulfilled all criteria set or described in the outer model analysis. The factor loadings value of each item had exceeded the cut-off value of 0.50, while the AVE and the composite reliability of all variables had exceeded the minimum acceptance value of 0.50 and 0.70 respectively. Furthermore, the value of AVE of each latent construct had been higher than the construct's highest squared correlation with any other latent construct. Based on these results, it could be concluded that the data and the model had fulfilled all criteria set in the outer model analysis. Therefore, the inner model analysis could be further conducted on the data and the model. After completing the outer model analysis, the inner model analysis was further conducted on the data and the model in order to determine the relationship between variables. Furthermore, mediating analysis was also conducted in this research in order to determine whether variable customer satisfaction didn't mediate, partially mediated, or fully mediated the relationship between website design quality and repurchase intention, and between service quality and repurchase intention. Moreover, data generated from the inner model analysis will be used as a benchmark in terms of whether the hypotheses proposed or posited in this research were supported or rejected. By implementing 5% of significance level, the cut-off value of 1.96 was set in this analysis in order to determine whether the hypotheses were supported or rejected. If the t-value is greater than 1.96, then it could be concluded that the hypothesis was supported. However, a t-value lower than 1.96 indicated that the hypothesis was rejected (Wilson, 2018). Based on the results generated on the inner model analysis, it was revealed that the relationship between variables were significant and positive. Furthermore, based on the results of the r-square value generated in this analysis, it was revealed that variable trust had an r-square value of 0.839. Based on these results, it could be concluded that variable trust was explained by both variable perceived usefulness and perceived ease-of-use by 61.7%, while 38.3% of variable trust was explained by the other variables not included in this research. Additionally, variable repurchase intention had an r-square value of 0.551. Based on these results, it could be concluded that variable trust was explained by variable perceived usefulness, perceived ease-of-use, and trust by 55.1%, while 44.9% of variable repurchase intention was explained by the other variables not included in this research. Furthermore, path coefficient analysis was conducted not only in order to know the relationship between variables, but also to know whether the hypotheses proposed in this research were supported or rejected. Based on the results generated in the path coefficient analysis, it was revealed that the t-value of all hypotheses had exceeded the cut-off value of 1.96. The first hypothesis proposes that perceived usefulness positively affect 247 trust. With a t-value of 2.030 which was well above the cut-off value of 1.96 (2.030 > 1.96), it could be concluded that the first hypothesis was supported in this research. The second hypothesis propose that perceived ease-of-use positively affect trust. With a t-value of 2.038 which was well above the cut-off value of 1.96 (2.038 > 1.96), it could be concluded that the second hypothesis was also supported in this research. The third hypothesis proposes that trust positively affect repurchase intention. With a t-value of 2.941 which was well above the cut-off value of 1.96 (2.941 > 1.96), it could be concluded that the third hypothesis was also supported in this research.
Other than testing the hypotheses proposed in this research, the mediation effect of trust was also analyzed and assessed in this research, in which a t-value above 1.96 in the indirect effect suggest that the variable mediated the impact of independent variable toward dependent variable. Based on the assessment on the indirect effect of the mediating variables, it could be concluded that trust mediated the impact of perceived ease of use and perceived usefulness toward repurchase intention in a positive manner, since the t-value of both relationships had well exceeded the cut-off value of 1.96. Moreover, it could also be stated that the mediation effect that exist on this research were partial mediation since perceived ease of use and perceived usefulness had a direct and positive impact on repurchase intention. Therefore, the fourth and fifth hypotheses were supported. Based on these results, it could be concluded that all hypotheses in this research were supported.

VI. CONCLUSION
Based on the results stated and described in the previous section, author would like to conclude that both perceived usefulness and perceived ease-of-use had a positive impact on consumers' repurchase intention, both directly and indirectly toward trust in the Indonesian e-commerce industry. Furthermore, based on the results of the data analysis, it could also be concluded that perceived ease-of-use played a more important role in determining or affecting consumers' repurchase intention compared to perceived usefulness, as this was reflected by the fact that perceived ease-of-use had a higher t-value compared to perceived usefulness.

VII. MANAGERIAL AND PRACTICAL IMPLICATIONS
The finding of this research had increased people's and company's understanding regarding the importance of perceived ease-of-use and perceived usefulness toward consumers' repurchase intention, either directly or indirectly through trust in the Indonesian e-commerce market. Furthermore, this research also provide an insight for managers who were involved in the e-commerce market that establishing trust in consumers' mind and instill perception on customers' mind that the system installed in the e-commerce system was easy to be learned, and buying from the e-commerce website will be more practical and have some advantages compared to buying the products/services from the traditional market. Therefore, customers' willingness to buy or even re-buy the product/service from the e-commerce website could be increased.

VIII. THEORETICAL IMPLICATIONS
Based on the results obtained in this research, it could be concluded that both perceived ease of use and perceived usefulness had a positive impact on Indonesian consumers' repurchase intention in the e-commerce industry, either directly or indirectly through trust. This conclusion was drawn based on the findings that perceived ease of use and perceived usefulness had a direct and positive effect toward repurchase intention, and had an indirect and positive effect on repurchase intention through trust. Furthermore, it was also concluded that trust had a positive impact on consumers' repurchase intention. The result of this research had shown that in the Indonesian e-commerce market, customers' perception on whether e-commerce could provided them some advantages in terms of their buying activities compared to traditional commerce, and customers' ability to learn the system installed in the e-commerce website play a positive role in affecting consumers' intention to re-visit and re-purchase the product or service from the same company or website. Previous research by Wen et al. (2011) and Aren et al. (2013) also shown that both variables, which are perceived ease of use and perceived usefulness had a positive impact on customers' repurchase intention and loyalty.

IX. LIMITATIONS AND SUGGESTIONS
Several limitations existed in this research. First of all, this research was conducted in the Indonesian ecommerce market, which means that the result might not be generalizable to the other countries or continents, as 248 customers' buying behavior, attitude, and perceptions are differ across countries. Therefore, it was suggested that further research needed to be conducted in the other countries in order to increase the generalizability of the results. Moreover, future research in the other industries were strongly suggested to be conducted in order to expand the scope of this research.