Content area
Purpose
In recent years, the ongoing development and rapid popularization of mobile devices, not only has brought new forms of lifestyle but also has boomed a wide array of service contents, allowing users to engage in entertainment, study and other daily activities, in addition to business transactions. To explore the factors that affect users' continued use of mobile app software, this paper aims to take into account the utilitarian value and privacy, as well as security factors and intends to propose an integrated research model based on the information systems success (ISS) model.
Design/methodology/approachThe data for this study were collected via an online survey questionnaire. A total of 190 valid responses were used to examine the hypotheses in the research model. The casual model was assessed using partial least squares techniques.
FindingsThe results indicated that the three quality factors (information quality, system quality and service quality) in the ISS model significantly affected satisfaction and utilitarian value. Private and security have significant negative association with utilitarian value. The results also showed that private and security did not significantly affect satisfaction. The findings also provide some directions for further research.
Practical implicationsThe results show that user's continued use of mobile app software depends upon the perceived utilitarian value and satisfaction. The quality characteristics are the three critical factors that would affect the perceived utilitarian value and satisfaction. Compared to information quality and service quality, system quality has a higher degree of positive impact, which indicates that users are more concerned about the immediate download ability and ease-of-use provided by the system. Furthermore, the research also finds that privacy and security concerns have a negative influence on the utilitarian value, showing that in the context of task-oriented IS, the perceived risk of privacy and security during the process of using mobile app software services, will negatively affect users’ perception of utilitarian value. The insights provided by this study can help mobile app developers optimize design and marketing strategies.
Originality/valueThis study provides a better understanding of how the factors in the theories influence the continuance usage intention of life style mobile applications.
Introduction
The popularity of the Internet and the continuous growth of mobile devices have led to the development of new commercial services. According to a research report of IDC (2018), an international research organization, the shipment of mobile devices has accumulated to 1.462 billion by 2018 and is estimated to accumulate to 1.654 billion by 2022. The rapid development and popularity of mobile devices result in the diversified development of mobile apps and also changes traditional business patterns and structures. This accelerates the advance of the mobile business, changes people's previous knowledge of convenience provided by e-commerce and enables people to get services or products free of device, location and time limitations. In addition, the market research institution eMarketer (2018) pointed out that the turnover of global e-commerce would increase to US$4.878tn in 2021 from US$2.842tn in 2018, equivalent to 17.5 per cent of total turnover in offline retail stores. As the mobile business emerges out, enterprises attach importance to the development of mobile apps progressively. The main advantage of mobile apps over Web pages lies in providing enterprises with more customized services and thus improving the degree of satisfaction. Many enterprises develop their apps to provide different customized functions for use by customers. However, not all mobile apps developed by enterprises will be used continuously by customers. This is mainly because quality characteristics provided by mobile apps, such as quality of system, information, and service, will have an influence on the satisfaction of customers towards the apps and thus affect their intention for continued use of apps.
Many companies develop mobile apps to attract more customers and bring more revenues. For example, Mcdonald's Newspaper App developed by Mcdonald's, Starbucks App by Starbucks and MOS App by MOS Burger seek to continuously understand their own customers and provide more convenient functions to get close to users' life. Moreover, today's customers will judge whether some action of using information technologies is successful through information consistency (Xu and Du, 2018). When customers feel confused about information, they will be less satisfied and thus reduce their intention to use such information (Martins et al., 2019; Stefanovic et al., 2016). Therefore, a product or service with no good information quality and benefits will make customers not willing to use it. In previous studies, academics explored whether the quality of mobile payment systems affected the satisfaction of customers by means of an information system success (ISS) model, and their research results verified that the quality indeed had an influence on the satisfaction of customers. Additionally, the issues of privacy and security were incorporated in the research, and such issues were also verified to have an influence on the satisfaction of customers and their intention to use (Gao et al., 2015). Some academics explored the influence of quality on the satisfaction of customers on C2C platforms, and their research results demonstrated that three items of quality in an ISS model had an influence on the satisfaction of customers (Chen et al., 2017). When users adopted a technological product or service, they will evaluate whether its use benefits have a positive effect. Some other academics explored the influence of smartwatches on customers' intention for continued use by means of utilitarian value, and their research results confirmed that smartphones had a positive effect (Hong et al., 2017). Therefore, this indicated that the quality of functions, interfaces, services etc. provided by mobile apps indeed has an influence on the satisfaction of customers, and as for current enterprises, it is important to meet demands of existing customers and acquire new customers.
In recent years, people have shown continuously increased acceptance of food ordering apps, in which Starbucks App is frequently used in Taiwan. Thus, in this research, by taking the Starbucks App, for example, the authors will explore customers' continued intention to use the app held by them. Viewing from an ISS model, we will discuss the continued intention to use the Starbucks App by building an integrated model incorporating utilitarian value and privacy and security.
Literature review
Mobile applications
According to the reports of market research institutions, the number of apps in the Google Play for the Google Android system has accumulated to about 2.1 million by September in 2018 from 1.6 million in 2015, and that in the App Store for the Apple iOS system has accumulated to about 2.45 million by July in 2018 from 1.4 million in 2015 (Statista, 2019a, 2019b, 2019c). From the above data, mobile apps are developed in a diversified manner. According to the statistic reports of Statista (2019a, 2019b, 2019c), in the growth-type ranking of the most popular apps in the Android system by June in 2018, the first is tool apps followed by communication and media apps. The Starbucks App to be discussed in this research belongs to the category of lifestyle, with its use coverage of 65.3 per cent. Accordingly, even if mobile apps grow year by year and people pay more attention to practicability for life, how to maintain old customers and attract new customers is very important to enterprises. Thus, in this research, the authors will explore the Starbucks App (belonging to lifestyle, with its use coverage of 65.3 per cent) and figure out why the factors that have an influence on customers' intention for continued use of the app are very important.
Information systems success model
The ISS model was proposed by DeLone and McLean (1992) and has been widely cited in literature of information technology. It is also used to predict interpretation and verification theories of individual behavior modes. Pitt et al. (1995) believed that information systems had evolved from closed systems to current open service platforms, and users of the information systems needed not only hardware equipment, but also information service to meet their demands. As a result, the service quality is increasingly important. DeLone and McLean (2003) updated the old IS success model with a new success model, considering IS qualitiy including information quality, system quality, service quality and degree of satisfaction as key indicators for successful IS prediction. The service quality refers to the output quality of IS evaluated, the system quality refers to the quality of IS evaluated, the service quality refers to the quality of service provided by IS evaluated, and the degree of satisfaction refers to the satisfaction appraisal of IS by users.
The ISS model has been used by many academics for exploration of personalized information systems (Aldholay et al., 2018; Aparicio et al., 2019; Chen et al., 2017; Gao et al., 2015; Huang et al., 2015; Xu and Du, 2018). Over the past years, based on an IS success model, many academics explored the influence and benefits of IS application by users, such as users' intention for continued use of mobile payment (Gao et al., 2015), customers' intention for re-purchase on a C2C (customer to customer) platform (Chen et al., 2017), users' loyalty to use e-libraries (Xu and Du, 2018), factors affecting users to use e-learning (Aparicio et al., 2019), actual behavior factors affecting users' e-learning (Aldholay et al., 2018), interests and benefits of using mobile libraries (Huang et al., 2015), and educational management IS system (Martins et al., 2019). All of these academics pointed out that users would be affected by information quality, system quality and service quality, which would, in turn, affect their personal satisfaction or intention to use.
As seen in the above-mentioned literature, the ISS model is regarded as an important theory that affects individual behaviors in their use of information systems. Indeed, several research models have been constructed by scholars who have tried to extend the concept of ISS model. Given that, this study will use the ISS model as the basic theoretical framework to explore the behaviors that affect users' adoption of mobile applications. However, factors affecting mobile apps use behaviors are not as straightforward as those of conventional IT. Chen and Chen (2010) suggested that users' adoption of IT should be explored from the perspective of quality, value, and satisfaction. By synthesizing various arguments raised by scholars, and based on the concept of “ISS characteristics” as the theoretical basis, this study examines users’ decision-making process (quality, satisfaction, and value). Furthermore, this study tries to interpret consumers’ idea about the values (such as utilitarian value) from a task-oriented perspective and to understand whether the three IS, quality factors affect customers’ assessment of the utilitarian value of the service.
Additionally, frequent incidents like leakage or other illegitimate misappropriation of customer information in recent years, have led to serious doubt of or unwillingness to use company services. Many scholars have pointed out that privacy and security risks are major factors that can negatively affect users' adoption or continued use of apps (Chopdar et al., 2018; Oghuma et al., 2016).
According to those, above and based on the ISS model by adding utilitarian value and privacy and security concerns, this study, therefore, establishes an integrated research model to explore the behaviors affecting customers' continued use of Starbucks App.
Privacy and security
Privacy and security refer to data held by individuals, such as credit card information, ID number, private information and privacy on mobile devices, all of which are within personal privacy and security. Many academics (Kleijnen et al., 2007; Yang et al., 2016) pointed out that users would be afraid of personal data leakage when they were transmitting data on the wireless Internet, and thus have risk concerns about privacy and security for using mobile networks. In other words, when users were transmitting data with mobile apps, security was the main concern, and they worried about security and privacy risks (such as personal privacy and data leakage, etc.) during usage.
A body of research in literature indicated that perceived risks included three risks, namely, security (e.g. leakage of personal data), privacy (e.g. loss of personal privacy) and cash flow (e.g. transaction and processing fees) (Karjaluoto et al., 2018). Academics believed that privacy and security should be incorporated in mobile business studies, for example by exploring whether sufficient security measures or other guarantees exist to ensure customers' own interests. When there were no sufficient security measures, customers would not use a product due to security concerns about its functions (Hajli, 2015; Johnson et al., 2018). Ozturk et al. (2016) found that perceived risks would have a negative effect on utilitarian value and hedonic value, thus affecting customers' intention for continued use of mobile technologies. Arpaci et al. (2015) also pointed out that privacy and security would negatively affect customers' personal use attitudes and thus their intention to use cloud-based services.
From the literature above, if users have concerns about privacy and security in the used information systems, these concerns will negatively affect users' intention to use the systems and thus reduce their use or continued use of the systems. This research will examine how the issues of privacy and security posed by mobile apps influence users' intention for continued use of mobile apps by considering characteristics in the scenario of mobile business.
Satisfaction
Oliver (1999) considered the satisfaction of customers as a cognition process in which the difference between their product expectation (before purchasing) and real product efficiency (after purchasing) constituted the satisfaction. Some academics (Bhattacherjee, 2001; Chen and Lin, 2015; Dong et al., 2014; Hsu et al., 2015) thought that the satisfaction of customers meant the degree to which they were satisfied by product quality or performance compared with prior expectations after they experienced products or services. This is to say, when customers used products or services, they would be satisfied if the perceived quality or performance of products or services met or exceeded prior expectations.
Many academics (Bhattacherjee, 2001; Hsu and Lin, 2016; Natarajan et al., 2017; Osatuyi and Qin, 2018) found in their studies related to information systems and e-commerce that, the satisfaction of customers played a considerably important role in retaining customers and promoting them to reuse products or services. Customers are the key to the ups and downs of web sites. As for service providers, users are not simply users, and they have turned into a helper to assist enterprises in their survival. It is therefore very important to understand factors having an influence on the satisfaction of customers.
Utilitarian value
Utilitarian value is an overall evaluation of customers in functional benefits and expenditures (Overby and Lee, 2006), and it is related to meeting a particular target (Kim et al., 2007). The utilitarian value is a functional concept, meaning users would complete some task-related targets with the help of mobile network service (Kim et al., 2007). Moreover, the utilitarian value is defined as benefits and costs (Lin and Lu, 2015) in consumers' functional evaluation after they use the overall product or service, for example, assessing whether social network service will increase people's social capability or improve their convenience. Ryu et al. (2010) argued that the utilitarian value would affect users' intention to use products or services; Hong et al. (2017) further pointed out that when users perceived a mobile technique (such as smart watches) was capable of satisfying them in terms of using the technique, they would find it useful; Karjaluoto et al. (2018) found that the intention of users to use the software would be affected when mobile financing software provided mobile payment or mobile banking functions that made customers feel their utilitarian value.
Research model and hypotheses
According to the literature reviewed above, the ISS, utilitarian value and satisfaction play important roles in affecting customers' behavioral intention to use information systems. Hence, this research will explore whether these factors have an influence on the behavioral intention to use mobile apps Figure 1.
System quality
System quality means the loading speed, ease of use, attractiveness, and correlation of information systems. In previous studies, web execution speed, mobile apps loading speed or ease of use of a product were used to test the system quality (Aldholay et al., 2018; Huang et al., 2015). For example, a Web page with no quick-display capabilities will cause customers to wait too long and not want to use the Web page. Some web pages are designed with excessively complex functions, and if customers are not clearly informed of these functions, their demands will not be met. By taking the Starbucks App for example, it has advantages of real-time image loading, ease of system use, and acceptable interface design, and customers can use services immediately through the app, free of time limitation.
Furthermore, in recent years, many studies have explored mobile payment service through system quality, in which the payment can be made with QR, NFC, bar code, etc. Some other academics explored the intention of customers to use mobile payment in China through system quality, and their research results showed that the system quality indeed had a positive effect on the intention and enhanced perceived value (Gao et al., 2015). Therefore, the following hypotheses are proposed:
System quality will positively affect utilitarian value.
System quality will positively affect user satisfaction.
Information quality
Information quality means the relevance, sufficiency, accuracy and timeliness of information provided by a system (Aldholay et al., 2018; Huang et al., 2015; Martins et al., 2019). In literature, the information quality was used to explore whether asymmetric or insufficient information contents provided by media, such as clothing information or images from online clothing stores, cause customers to get confused about products and thus avoid shopping. If the preview method of information obtained on mobile devices is improper or the information is incorrect, it will make customers not willing to preview products with the site or app and thus affect their satisfaction (Aldholay et al., 2018; Aparicio et al., 2019). In the case of the Starbucks App, the information quality means that product information, event information, membership information and mobile payment function are ensured to be correct and updated immediately. Additionally, in previous studies, the information quality was used to explore mobile payment service, in which the payment should be made correctly. Once no sufficient and immediately updated correct information was provided, the satisfaction and perceived value of customers to use mobile payment service would be affected (Gao et al., 2015). Hence, the following hypotheses are formed:
Information quality will positively affect utilitarian value.
Information quality will positively affect user satisfaction.
Service quality
Service quality means the reliability, responsiveness, assurance and personalization of system services (Huang et al., 2015; Martins et al., 2019). The service quality refers to guaranteed services, after-sales services, accurate and tangible services provided by enterprises for customers (Aldholay et al., 2018). General enterprises have the intention of after-sales service, i.e. an after-sales service center, with which customers will have its questions answered online or consult the enterprises by telephone. In other words, good service qualities provided by enterprises, such as interpersonal communication (customer service centers), product warranty, service stability and tangible service, are all within the scope of service quality (Aldholay et al., 2018; Huang et al., 2015). Personalized services are provided in the Starbucks App, such as electronic membership cards that contain membership bonus points and relevant information, and network push services that can inform customers of recent services, etc. That is to say, good and stable service quality will improve customers’ satisfaction and perceived value and thus ensure their intention to use the service. Therefore, we propose the following hypotheses:
Service quality will positively affect utilitarian value.
Service quality will positively affect user satisfaction.
Privacy and security concerns
Privacy and security concerns mean that customers worry about whether their privacy and payment information will be used improperly, and whether they will be subject to fraud risks or property losses due to improper personal information disclosure. Previous studies pointed out that customers would focus on safety protection of their information by enterprises to prevent property losses due to information disclosure when they use a product or service (Arpaci et al., 2015; Gao et al., 2015; Johnson et al., 2018). Ozturk et al. (2016) studied the usage of mobile room-booking software, and their research results showed that perceived risks had a major negative influence on utilitarian value and hedonic value. Karjaluoto et al. (2018) examined the factors of using mobile finance services, and their research results indicated that perceived risks had a major negative influence on utilitarian value.
Gao et al. (2015) explored mobile payment service, and their research results demonstrated that privacy and security had a negative influence on the satisfaction of users. This research suggested that the Starbucks App provided by Starbucks had the function of mobile payment with QR. In details, customers top up with a credit card online to deposit money to their Starbucks Cards and then make payments through a QR code or the entity card. However, with the rapid development of today's technologies, privacy rights and personal data leakage problems affect people's life significantly. Customers are afraid that disclosure of personal data or credit card information in their used service will bring them losses in terms of rights and money, and thus reduce their influence on enterprise brand satisfaction, decrease their perceived value and prevent their intention to use information system services. Thus, the following hypotheses are proposed:
Privacy and security concerns will negatively affect utilitarian value.
Privacy and security concerns will negatively affect user satisfaction.
Utilitarian value
Utilitarian value means purposeful, rational, and functional use of a product or service (Ryu et al., 2010). This research suggested that users would get utilitarian value when using mobile apps on smartphones that help to achieve some functional or utilitarian capabilities to meet the purpose of use.
The utilitarian value can also be defined as the perception of users gained by using services or products for a particular purpose (Chen and Fu, 2018). It also includes exploration of the use of information technologies with features such as ease of use and serviceability, indicating that the utilitarian value contains such ease of use and serviceability. Karjaluoto et al. (2018) showed that the utilitarian value would affect users' behavior to use information systems.
This research held that the app released by Starbucks was to meet the trend of mobile business, also combining with previous entity card payment functions and in-store information such as events, goods, individual store locations, etc. The main purpose was to consolidate its own customer group, and transfer previous offline promotions to online stores through the app and integrate them to the system, enabling customers to know immediately about the existing amount of their stored value cards, discount information, etc. From the above, to meet the trend of mobile business, enterprises launch customized services for their respective customer groups, to get close to the lives of customers. Thus, the utilitarian value indeed has an influence on the use intention of customers and indirectly affects the satisfaction of customers. Therefore, the following hypotheses are proposed:
Utilitarian value will positively affect continued usage intention.
Utilitarian value will positively affect user satisfaction.
Satisfaction
Satisfaction means that customers are pleased with their perception after using a product or service, i.e. showing a high degree of satisfaction. In other words, satisfaction will be achieved when external and internal motives are met. However, the degree of satisfaction has been widely used to explore the use of educational information systems (Martins et al., 2019), airline brand influence (Kim, 2015), mobile financing software services (Karjaluoto et al., 2018), mobile apps (Chen and Fu, 2018), e-commerce (Yoo and Park, 2016), purchase intention and re-purchase intention (Mencarelli and Lombart, 2017) and social commerce (Osatuyi and Qin, 2018). This research held that when app functions provided by Starbucks can satisfy customer groups, customers will continuously use the app. Therefore, the following hypothesis is proposed:
Satisfaction will positively affect continued usage intention.
Research method
We use SPSS 22.0 and SmartPLS 3.0 statistical software as the primary instruments for data analysis. At the first stage, the measurement model was analyzed to verify the common method variance, reliability, validity and collinearity analysis of the research tool. At the second stage, the path model analysis of various dimensions was conducted. Due to the small sample size, Bootstrapping was adopted based on Bollen and Stine’s (1992) recommendation to resample 500 samples and compute the significance of model coefficients.
Data collection and participant demographics
This study targeted the Starbucks App’s Taiwanese consumers. According to the Starbucks Taiwan office, the number of Starbucks App users has reached 1.5 million by 2018, accounting for 51 per cent of the total revenue, and was predicted to rise to 1.8 million in 2019. Encouraged by the rapid growth of mobile payment, the company was planning to integrate various payment instruments into two modes, e-ticket and credit card, to attract other Starbucks customers to use the Starbucks App. Garnering annual revenue of approximately $100m in Taiwan, Starbucks is currently the largest coffee chain store in Taiwan. It is for the aforementioned reasons that we chose Starbucks as the research object.
This study used online questionnaires to collect data. The questionnaires were issued to the randomly selected Starbucks users from February 15 to March 15, 2018, and were simultaneously posted on the Starbucks board of PTT, Taiwan's largest bulletin board system. To increase the recovery rate, free gifts (e.g. Starbucks e-gift cards) were offered to attract potential respondents. In addition, to avoid repetition, respondents’ identities were confirmed by tracking their e-mail and IP address.
A total of 190 valid questionnaires, with an 85.2 per cent response rate, were selected from the 223 returned questionnaires, excluding 33 that were done by consumers who did not use the Starbucks App, or were invalid and repeated questionnaires. The majority of research samples were female, accounting for 67.9 per cent; main age group was between 21-30 years old, accounting for 50 per cent. Those with university degrees occupied the highest percentage of 64.7 per cent, and with office workers accounted for 58.8 per cent of the occupation group. Detailed demographic data of the samples is shown in Table I.
Measurement development
Apart from the demographic variables of the items which belong to the categorical data, the scale measurements were divided into two types. The 5-point Likert Scale was adopted for measuring the variables, ranging from 1, representing highly disagree, to 5, representing highly agree. Sources of each variable measurement instrument were compiled as shown in Appendix.
Results data analysis and results
Tests of common method variance
To avoid correlation deviation resulted from human-induced covariation between independent and dependent variables caused by the nature of semantics, research situation, and the scale itself, the present study used the Harman’s single factor test method. According to previous literature (Podsakoff et al., 2003), if any single factor is capable of explaining more than 50 per cent of the variance, a significant common method variant may exist.
However, Harman’s single factor test has extracted a total of five factors with the characteristic of ≧1. The explained variation of the first factor was 37.266 per cent (at maximum) and was < 50 per cent, and the explained variation of the accumulative total 5 factors was 66.770 per cent. As the maximum explained variation according the result of Harman’s single factor test is lower than 50 per cent, it can be reasoned that this study has found no significant common method variation.
Tests of the measurement model
This study used the confirmatory factor analysis to test the measurement model. Regarding the reliability analysis, we used factor loading, Cronbach's α, and composite reliability (CR) to evaluate the internal consistency of the model. The results are shown in Table II. The factor loading of all dimensions was greater than 0.5 (Fornell and Larcker, 1981), and the Cronbach's α value of each dimension falls within the range of 0.74-0.92, which is greater than the 0.7 acceptable standard recommended by Nunnally (1978). All the composite reliability was > 0.8, higher than the 0.7 suggested by Fornell and Larcker (1981), indicating good reliability and stability of each item of the dimensions.
In the aspect of convergent validity, there were three standard model measurements suggested by Bagozzi and Yi (1988):
all factor loadings should be > 0.5 (Hair et al., 2009);
composite reliability should be > 0.7 (Fornell and Larcker, 1981); and
the average variance extraction (AVE) of each dimension should be > 0.5 (Hair et al., 2009).
As shown in Table II, the factor loadings of all the items exceed 0.7; the composite reliability of each dimension is at 0.83-0.94, and the average variation of each dimension is at 0.56-0.83. Therefore, it is evident that the measurement model of this study has exhibited the required convergent validity.
As for the discriminant validity, Fornell and Larcker (1981) have suggested that the square root of AVE for the dimension itself should be greater than its correlation coefficient between other dimensions. As shown in Table III, the AVE of each dimension is greater than the square root of the correlation coefficient between the dimensions. Consequently, the various dimensions in the measurement model are different from one another, which further verifies the discriminant validity of each construct. To sum up, the measurement pattern of this research model shows good reliability, convergent validity, and discriminant validity.
The problem of collinearity may arise when multiple regression models are analyzed. Multi-collinearity means that a complete or almost complete linear correlation may exist between two or more independent variables. Because using the method of least squares to calculate regression coefficient requires that “the independent variables are not in complete linear correlation”, complete linear correlation existing between the independent variables may results in the regression coefficient not being the only result, and also, no value of regression coefficient can be obtained. In most cases, a correlation may exist between the independent variables and the coefficient is at 0-1, which can lead to the decreased estimated value and stability of the parameters when the regression model is needed. Therefore, this study used the variation inflation factor (VIF) to conduct the test. The VIF is the inverse value of the tolerance. The greater the value of VIF is, the greater the possibility of the existence of collinearity would be, and vice versa.
In light of the standard of VIF value proposed by previous scholars, the VIF is ≦10. In addition, the results of collinearity analysis in this study, as shown in Table IV, indicates that there is no collinearity problem existing between the dimensions (Petter et al., 2007).
Tests of the structural model
In this study, Smart PLS 3.0 was used for path analysis, and Bootstrapping was used in resampling 500 times for t-values to verify the significance of the variable path of the proposed model. The analysis results are shown in Figure 2. In addition, both the direct and indirect effects of each dimension are shown in Tables V and VI respectively.
The effects of indirect influence between various dimensions are shown in Table VI. Information quality has a positive indirect effect on continued intention to use (β = 0.23, p < 0.001). Service quality has a positive indirect effect on continued intention to use (β = 0.16, p < 0.001). System quality has a positive direct effect on continued intention to use (β = 0.16, p < 0.001). Whereas privacy and security concerns have a negative indirect effect on continued intention to use (β= −0.06, p > 0.05).
Discussion
This research mainly explored why people would continuously use mobile apps. Based on an ISS model and from the perspectives of privacy, security and utilitarian value, this research proposed an integrated model to figure out the reasons for people's continued use of the Starbucks App. The research results hereof indicated that the quality factor, privacy and security, utilitarian value, and degree of satisfaction played an important role in continued use intention of users. Findings in this study will be discussed below.
Influence of utilitarian value and satisfaction on the intention of continued use
The utilitarian value will positively affect the satisfaction of users, and research results correspond to those of previous studies (Karjaluoto et al., 2018). If people are satisfied when experiencing the utilitarian value of information technologies, the utilitarian value will positively affect the satisfaction of users. Next, the utilitarian value will positively affect the continued intention to use the Starbucks App, and research results correspond to those of (Ozturk et al., 2016). When users believe that the Starbucks App will bring use benefits, for example, which include replacing previous entity cards or cash payment with digital payment and providing diversified additional functions such as online amount inquiry, product information, promotional information, etc. The satisfaction of customers and their perception of practicability will indeed be improved, and additional opportunities will be brought to enterprises.
This research indicated that the satisfaction of users for mobile apps would positively affect the continued use intention of users. If people can effectively meet their purposes through functions provided by the app, their satisfaction will be increased and thus their continued use intention for the app strengthened (Osatuyi and Qin, 2018; Ozturk et al., 2016).
Influence of system quality, information quality and service quality on utilitarian value and satisfaction
This research indicated that the system quality, information quality, and service quality influenced utilitarian value and satisfaction of users. Research results showed that in the scenario with mobile apps, the system quality, information quality, and service quality played a very important role. First, when the system quality provided by mobile apps meets users' demands. For example, the system platform access speed is high, and users do not have to spend too much time learning system designs with good operations and navigations, the utilitarian value and satisfaction of users will be affected positively (Aldholay et al., 2018; Gao et al., 2015; Huang et al., 2015;). Next, in terms of information quality, research results showed that when people obtain correct and sufficient information contents through the app in a real-time manner, their perception of value provided by the app would be improved and thus their continued use intention strengthened. Finally, in terms of service quality, personalized services provided by the app will positively affect the utilitarian value and satisfaction of users. This research concluded that the personalized services provided by the app, such as accumulated bonus points, coupons or consumption records, would improve users' perception of the service quality in the app and strengthen their satisfaction and feeling for value on the platform, thus affecting their use behaviors. To this end, this research believed that the system quality, information quality and service quality are three main drivers of achieving the utilitarian value and satisfaction of users.
Effects of privacy and security concerns on utilitarian value and satisfaction of users
Privacy and security concerns will negatively affect utilitarian value, and research results in this research correspond to those of previous academics (Karjaluoto et al., 2018; Ozturk et al., 2016). When using mobile apps, users pay special attention to privacy risks, like leakage of personal data and records. However, this research did not find that privacy and security would negatively affect the satisfaction of users, which was the opposite against the assumptions. This study infers that customers must log in before making payment via the Starbucks App. The process requires passing on personal data and financial transaction information. Therefore, customers’ perceived utilitarian value will be negatively affected when doubts arise about the privacy and security of the app. Oghuma et al. (2016) explored the continued use of instant messaging software indicated that perceived security has no direct effect on satisfaction. This is mainly because of the good impression the brand has built in the customer’s mind that leads to trust in the sufficient security protection measures and coping strategies of the enterprise. This study made the further deduction that when consumers have good impressions on the Starbucks brand, they tend to not worry about disclosure of personal data or financial loss, hence the privacy and security concerns have no significant impact on satisfaction. Karjaluoto et al. (2018) explored mobile financing software, and their results showed that perceived risks would directly affect the utilitarian value and thus the satisfaction. All of these studies were consistent with the results of this research. Accordingly, it could be concluded that customers would only assess their privacy and security in the use of mobile apps after generating practical use intentions through benefit evaluation.
Control variables
Control variables in this research include gender, age, and educational status. The research explored whether these variables would affect the continued intention of users to use mobile apps. Research results showed that the influence of gender, age, and educational status did not reach a significant level, and thus would not affect the continued use intention of users.
Implications
Academic implications
This research provides several important academic implications. First, based on an ISS model, this research explored user's perception of quality characteristics in mobile apps, and developed an integrated model framework by incorporating utilitarian value, privacy and security, thereby exploring the reasons for continued intention to use mobile apps. Findings show that in the Starbucks App context, ISS factors (system quality, information quality and service quality), privacy and security concerns, utilitarian value and satisfaction all play critical roles in affecting users’ willingness to continue use the Starbucks App. Second, the three constructs of quality characteristics can predict consumers’ utilitarian value and satisfaction towards mobile apps, with system quality having the most significant impact. It follows that timely download ability and ease-of-use provided by the mobile apps have a major impact on user satisfaction, which in turn affects users’ willingness to continue use the apps. In addition, the previous mobile app-related studies (Cidral et al., 2018) found that service quality had no significant effect on user satisfaction. In contrast, this study finds that service quality does have a significantly positive impact. It can be inferred that service quality is still a very important determinant in affecting customers' feeling during the service period. Third, many scholars (Kleijnen et al., 2007; Yang et al., 2016) pointed out that wireless networks users worried about disclosure of personal data in the process of transmitting personal data, which result in dubious attitude towards the risks of using mobile networks (such as personal privacy and security risks). Therefore, this study incorporated privacy and security concerns into the ISS model and proposed an integrated theoretical framework to better understand users’ willingness to continue use the mobile app software. Finally, the model in this research shows good explanatory capabilities and can indeed predict users' behavior to use mobile apps, providing a new thinking direction for future research.
Practical implications
This research will provide practitioners with some implications. First, the utilitarian value and satisfaction of users are the main factors affecting the continued intention to use mobile apps. Therefore, in an action-oriented information system, when mobile apps assist users in completing some task-related targets, users will feel their practical benefits. This not only improves the satisfaction of users but also strengthens their continued use intention. Second, the utilitarian value is one of the important factors having an influence on the satisfaction of users. Research results showed that when the more the utilitarian benefits provided by mobile apps, the higher the satisfaction of users. Thus, mobile application software providers should continuously develop diversified applications for users' demands. In addition, this research found that system quality was another important factor affecting the satisfaction of users. Third, quality characteristics will effectively increase the utilitarian value of mobile apps, with the influence of system quality and information quality greater than that of service quality. This result provides mobile application software providers with important information and displays mobile apps that users consider to be action-oriented, and people pay more attention to benefits of system quality (such as system platform access speed, ease of use, navigation, etc.) and information quality (such as correctness, immediacy, and correlation of information contents). Building a good environment of system quality and information quality is thus more effective than simply emphasizing service quality. Finally, privacy and security will negatively affect users' knowledge of utilitarian value. We can thus see that customers tend to pay particular attention to the risks of privacy and security, such as leakage of personal data and transaction records, repeated deduction, or loss of mobile phones. In other words, the more the users sense security risk, such as personal data leakage when using the mobile as a payment tool, the lower evaluation they would ascribe to the payment service provided by the mobile app software. That is, the consumer's perceived privacy and security concerns have a negative impact on the utilitarian value. Therefore, by strengthening the privacy security functions (e.g. proper use and protection of customer' data on the platform, and well-designed security function and professional payment flow), will assist service providers in arousing customer's perceived utilitarian value, and thereby enhancing their willingness to continue use the app. Therefore, enterprises should consider emphasizing that information will not be disclosed to outside parties under strict security protocols so that users know it is safe to use mobile apps.
Conclusion and limitations
This study reports valuable findings and implications; however, it also includes several limitations. First, whether the findings can be generalized to all types of mobile payment service is unclear. Further research is necessary to verify the generalizability of these findings. Second, regarding the research methodology, a quantitative statistical research model was used. Online questionnaires were used to collect data, and information of the online survey was announced across a variety of mobile payment forums. The purpose was to increase sample diversity and avoid skewed results caused by insufficient sample attributes. Self-selection bias, however, may well be present in instruments of this type. A combination of qualitative and quantitative viewpoint in future research may offer more comprehensive evidence to support such research. Third, this study proposed two sets of antecedents that can influence mobile payment usage behaviors: positive factors (system quality, information quality and service quality) and negative factors (privacy and security). Nevertheless, these factors were inadequate in providing full explanatory power for all usage attitudes and behaviors. Therefore, it is suggested that future studies could add other constructs to the model (e.g. types of mobile payment service and trust), to better explain and predict mobile-oriented user behaviors. Last but not least, future studies could incorporate interference variables, such as habit or involvement, to foster the understanding of the topic in question.
Research model
Results of research model
Demographic characteristics of respondents
| Demographic | Category | Frequency | (%) |
|---|---|---|---|
| Gender | Male | 61 | 32.1 |
| Female | 129 | 67.9 | |
| Age (years) | <20 | 15 | 7.9 |
| 21-30 | 95 | 50 | |
| 31-40 | 73 | 38.4 | |
| >41 | 7 | 3.7 | |
| Education level | High school or under | 10 | 5.3 |
| College/university | 123 | 64.7 | |
| Graduate degree | 57 | 30 | |
| Occupation | Student | 33 | 17.4 |
| Office makers | 112 | 58.8 | |
| Home makers | 5 | 2.6 | |
| Self-employment | 40 | 21.1 | |
| <20,000 | 38 | 20 | |
| Monthly income | 20,001∼40,000 | 67 | 35.3 |
| 40,001∼60,000 | 59 | 31 | |
| 60,001∼80,000 | 17 | 9 | |
| >80,001 | 9 | 4.7 |
FLs, AVE, CR and Cronbach’s alpha values
| Construct | Items | Factor loading | t-value | CR | AVE | α |
|---|---|---|---|---|---|---|
| System quality (SQ) | SQ1 | 0.74 | 16.30 | |||
| SQ2 | 0.85 | 43.05 | 0.89 | 0.67 | 0.83 | |
| SQ3 | 0.85 | 38.46 | ||||
| SQ4 | 0.82 | 33.57 | ||||
| Information quality (IQ) | IQ1 | 0.81 | 27.87 | |||
| IQ2 | 0.83 | 29.14 | 0.88 | 0.65 | 0.82 | |
| IQ3 | 0.81 | 25.19 | ||||
| IQ4 | 0.78 | 23.07 | ||||
| Service quality (SEQ) | SEQ1 | 0.83 | 33.23 | |||
| SEQ2 | 0.80 | 22.12 | 0.84 | 0.64 | 0.71 | |
| SEQ3 | 0.76 | 20.41 | ||||
| Privacy and security concerns (PSC) | PSC1 | 0.84 | 18.23 | |||
| PSC2 | 0.91 | 34.07 | ||||
| PSC3 | 0.93 | 69.67 | 0.94 | 0.75 | 0.92 | |
| PSC4 | 0.94 | 70.78 | ||||
| PSC5 | 0.71 | 9.49 | ||||
| Utilitarian value (UV) | UV1 | 0.71 | 19.73 | |||
| UV2 | 0.77 | 18.90 | 0.83 | 0.56 | 0.74 | |
| UV3 | 0.71 | 13.47 | ||||
| UV4 | 0.80 | 26.93 | ||||
| Satisfaction (SF) | SF1 | 0.93 | 90.70 | |||
| SF2 | 0.93 | 77.49 | 0.94 | 0.83 | 0.90 | |
| SF3 | 0.87 | 40.27 | ||||
| Continued intention (CI) | CI1 | 0.81 | 15.92 | |||
| CI2 | 0.92 | 90.53 | 0.89 | 0.73 | 0.81 | |
| CI3 | 0.82 | 22.41 |
Discriminant validity
| Construct | CI | IQ | PSC | SEQ | SF | SQ | UV |
|---|---|---|---|---|---|---|---|
| CI | 0.853 | ||||||
| IQ | 0.481 | 0.806 | |||||
| PSC | −0.211 | −0.236 | 0.869 | ||||
| SEQ | 0.394 | 0.672 | −0.097 | 0.797 | |||
| SF | 0.596 | 0.660 | −0.145 | 0.634 | 0.911 | ||
| SQ | 0.467 | 0.581 | −0.161 | 0.548 | 0.661 | 0.816 | |
| UV | 0.615 | 0.672 | −0.296 | 0.601 | 0.705 | 0.594 | 0.747 |
Notes:Diagonal elements (italic) are the square root of average variance extracted (AVE) between the constructs and their measures. Off-diagonal elements are correlations between constructs. For discriminant validity, diagonal elements (AVE) should be greater than off-diagonal elements. All correlations are significant at p < 0.01
Summary of collinearity analysis
| Construct | CI | IQ | PSC | SEQ | SF | SQ | UV |
|---|---|---|---|---|---|---|---|
| CI | |||||||
| IQ | 2.418 | 2.153 | |||||
| PSC | 1.123 | 1.070 | |||||
| SEQ | 2.082 | 1.975 | |||||
| SF | 2.027 | ||||||
| SQ | 1.766 | 1.624 | |||||
| UV | 2.037 | 2.253 |
Direct effect between the dimensions
| Relationship | β | Mean | SD | t-value | ρ | Result of hypothesis |
|---|---|---|---|---|---|---|
| IQ → SF | 0.17* | 0.16 | 0.08 | 2.10 | 0.036 | Support |
| IQ → UV | 0.34*** | 0.34 | 0.08 | 4.14 | 0.000 | Support |
| PSC → SF | 0.06 | 0.06 | 0.05 | −1.20 | 0.231 | Not support |
| PSC → UV | −0.15** | −0.16 | 0.05 | −3.18 | 0.001 | Support |
| SEQ → SF | 0.17* | 0.17 | 0.08 | 2.09 | 0.036 | Support |
| SEQ → UV | 0.22** | 0.22 | 0.07 | 2.94 | 0.003 | Support |
| SF → CI | 0.32*** | 0.32 | 0.08 | 3.84 | 0.000 | Support |
| SQ → SF | 0.28*** | 0.28 | 0.06 | 4.61 | 0.000 | Support |
| SQ → UV | 0.25*** | 0.25 | 0.07 | 3.80 | 0.000 | Support |
| UV → CI | 0.39*** | 0.39 | 0.08 | 5.11 | 0.000 | Support |
| UV→ SF | 0.34*** | 0.34 | 0.08 | 4.50 | 0.000 | Support |
| Gender → CI | 0.00 | 0.00 | 0.06 | 0.06 | 0.950 | |
| Age → CI | −0.03 | −0.03 | 0.06 | 0.55 | 0.581 | |
| Education → CI | 0.03 | 0.03 | 0.06 | 0.55 | 0.585 | |
| Revenue → CI | 0.02 | 0.02 | 0.06 | 0.34 | 0.732 |
Notes:***p < 0.001; ** p < 0.01; *p < 0.05
Indirect effect between the dimensions
| Relationship | β | Mean | SD | t-value | ρ |
|---|---|---|---|---|---|
| IQ → CI | 0.23*** | 0.23 | 0.05 | 4.52 | 0.000 |
| PSC → CI | −0.06* | −0.06 | 0.03 | −1.96 | 0.050 |
| SEQ → CI | 0.16*** | 0.16 | 0.05 | 3.56 | 0.000 |
| SQ → CI | 0.21*** | 0.22 | 0.04 | 5.21 | 0.000 |
Notes:***p < 0.001; ** p < 0.01; *p < 0.05
Measurement items for the constructs
| Constructs | Measurement items | References |
|---|---|---|
| System quality | SQ1: Starbucks App quickly loads all the text and graphics | Gao et al. (2015) |
| SQ2: Starbucks App is easy to use | ||
| SQ3: Starbucks App is easy to navigate | ||
| SQ4: Starbucks App is visually attractive | ||
| Information quality | IQ1: Starbucks App provides me with relevant information | Gao et al. (2015) |
| IQ2: Starbucks App provides me with sufficient information | ||
| IQ3: Starbucks App provides me with accurate information | ||
| IQ4: Starbucks App provides me with up-to-date information | ||
| Service quality | SEQ1: Starbucks App provides on-time services | Gao et al. (2015) |
| SEQ2: Starbucks App provides prompt responses to my questions | ||
| SEQ3: Starbucks App provides professional services | ||
| Privacy and security concerns | PSC1: I am concerned over the security of personal information exchange on Starbucks App | Gao et al. (2015) |
| PSC2: I am concerned that my personal information may be shared with business without my consent as a result of purchasing through Starbucks App | ||
| PSC3: I am concerned that the information I disclosed to this mobile vendor may be misused | ||
| PSC4: I am concerned about providing personal information to this mobile vendor, because it could be used in a way I can not foresee | ||
| PSC5: I am worried about the security of financial transactions carried out on mobile site | ||
| Satisfaction | SF1: I feel satisfied with the services provided by Starbucks App | Gao et al. (2015) |
| SF2: I feel contented with the services provided by this Starbucks App | ||
| SF3: I feel pleased with the services provided by this Starbucks App | ||
| Utilitarian value | UV1: Using Starbucks App enables me to pay immediately | Hong et al., 2017 |
| UV2: Using Starbucks App enables me to answer calls and messages immediately | ||
| UV3: Using Starbucks App could synchronize my physical member's card and consumption bonus | ||
| UV4: Using Starbucks App enables me to receive the newest information | ||
| Continued intention to use starbucks APP | CI1: I intend to continue purchasing through Starbucks App rather than discontinue its use | Gao et al. (2015) |
| CI2: My intentions are to continue purchasing through Starbucks App than use any alternative means | ||
| CI3: If I could, I would like to discontinue my purchase through Starbucks App |
© Emerald Publishing Limited 2019