“Digital payment system innovations: A marketing perspective on intention and actual use in the retail sector”

This study empirically investigated the marketing perspectives of behavioral intention and the actual use of digital payment solutions as electronic innovation for retail purchases in Thailand. This is important as leveraging digital innovation can be applied to minimize physical contact between retailers and customers, especially in the COVID-19 era. The UTAUT model was used and extended to include attitude, social distancing, and perceived risk variables. The study was conducted using primary data collected from 467 Thai respondents who used digital payment systems as a means of payment in retail purchases. The study data were collected employing a structured questionnaire. Techniques used in data analysis include Confirmatory Factor Analysis and Structural Equation Modeling. The results from the data analysis highlighted that behavioral intention to use digital payment innovation in Thailand was influenced by Perceived Risk (PR), Facilitating Condition (FC), Performance Expectancy (PE)


INTRODUCTION
In the recent past, innovations in information and financial technology have brought about a great revolution, leading to the development and advancement of e-commerce activities. This e-commerce, in turn, has created new fiscal needs that, in most cases, cannot be effectively met by conventional financial systems. Marketing has been instrumental to the spread of these digital technologies, especially on the internet and social media (Dahlström &  From this, various financial institutions, businesses, and other stakeholders are exploring the inherent potential and opportunities resulting from these technologies and are deploying an appropriate marketing strategy to introduce them to the market. Among the notably developed and well-adopted technologies are digital payment systems (Sivathanu, 2019). The digital payment infrastructure consists of an interconnected network of related entities, which are designed to accelerate the speed of data exchange between the concerned systems, and to initiate, sanction, and expedite cash transfer between different parties (Scholnick et al., 2008). Digital payments incorporate financial transactions initiated by an individual or several clients, which may cover business-to-business (B2B) transactions, individual-to-business transactions, and person-to-person payments. The intensity of the electronic payments is usually determined by the number of retail transactions to project the volume of digital payments taking place in the country levels. There is a wide range of payment solutions used in digital payments such as Point of Sale (PoS), Automated Teller Machines (ATMs), and other online and mobile banking applications. The consideration of these factors determines the level of adoption of digital payment systems.
Innovative digital solutions imply monetary transactions that are conducted using digital technologies. Through this means of payment, the sender and the recipient of money use digital technology to send and receive payment. Inferring from Nwaolisa and Kasie (2012), this type of financial transaction is also regarded as an electronic payment, which is mainly characterized by the absence of physical cash. Various factors have been considered by researchers to contribute to the growth of digital payments, they include: increased technology, the ever-increasing use of smartphones, institutions starting to offer more non-banking services such as online payments, as well as the adoption of regulations, which supports the growth of these technologies (Yu et al., 2002). Digital payments technology is considered to revolutionize the financial sector by bringing in appealing features such as user-friendliness, convenience, and fast delivery of payment, as compared to manual or traditional payment systems.
As the digital industry continues to develop in Thailand, the impact of Fintech advancement is to present digital payment systems that have revolutionized the business world such as e-money, e-wallet, among other payment systems (Chaveesuk et al., 2018;Kovács et al., 2007). In Thailand, digital payments systems have progressed appreciably in urban communities, and, according to Tounekti et al. (2019), they are used to non-cash payment facilities in their retail purchases. Digital payments are considered to facilitate the ease of the payment process, amid daily and demanding business activities. The most common digital payment systems for transacting payments in Thailand include electronic banking (e-banking), credit cards, debit cards, Prompt Pay, Alipay, PayPal, amongst other electronic payment systems available.
The successful implementation of digital payment systems is critical for the retail sector, especially when considering the transition of Thailand to Web 4.0. The adoption has been greatly aided by the marketing of services, especially digital marketing efforts (Feyen et al., 2021;Nathues, 2017). As a result, understanding customers' requirements and satisfying them in the retail context is important if the success of a digital payment system is considered a priority. Montazemi and Qahri-Saremi (2015) argue that to have increased adoption of digital payment systems, there should be effective management of the factors that influence customers'' adoption. Although huge investments have been made in the financial sector in terms of technological innovations; research shows that there is evidence of retail users being reluctant to adopt and use digital payment technologies (Kankanhalli & Gomez, 2020;Ligon, 2019;Macheel, 2017;Seethamraju & Diatha, 2019Staykova & Damsgaard, 2016). This highlights the need to investigate the factors influencing the behavioral intention to use and actual use of digital payment systems in the retail sector.

LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT
Technological developments in the financial industry have led to the emergence of e-com-merce activities such as digital payment systems (Sivathanu, 2019 (Howcroft, 2003). The high security of e-banking channels enhances trust, good trust perception, and these are likely to enhance the use of e-commerce platforms. Other studies investigated dynamics that prompt the embracing of e-banking based on the collected data from clients and customers of public and private banks ( The study concluded that PU strongly influences BI. According to TAM, PU and PEOU influence behavior by the way they are handled in the system of information. This developed behavior networks that analyze the intention of a person and results in the acceptance of the technology. Pikkarenen et al. (2004) used Technology Acceptance Model in Finland to determine that perceived usefulness is the key determinant of real behaviors that trigger bank customers towards utilizing comprehensive and modern e-banking and it offered increased independence and autonomy in financial transactions. Davis (1989) portrays PEOU as the extent to which an individual trusts that utilization of a given technology system will be free. Perceived ease of use refers to any task that is carried out effortlessly.
A study was carried out to discuss how the perception of users concerning online banking is impacted by the privacy policy and the PEOU. Goodrich and Boer (2003) concluded that PEOU cannot be of great significance when compared with security and privacy. Roboff and Charles (1998) contended that PEOU does not directly influence attitude intention though it was influenced more by a factor of indirect mediation from perceived use. PEOU states their insight and understanding of the process that leads to the result. Chin and Ahmad' (2015) findings showed that PEOU is crucial in determining consumer intent to use an e-payment system of. Hence, PEOU can be utilized as an intermediary of perceived enjoyment.

AIM
Using the cited literature as a guide and research gaps on the implementation of digital payment systems in Thailand, the study aimed to investigate the factors that influence BI in the use of digital payment solutions as well as the actual use of innovative digital payment solutions in the retail sector in Thailand from a marketing perspective.
Understanding these factors will influence stakeholder effort towards a seamless digital transaction experience both for consumers and retailers. Banks and other financial institutions can leverage these factors when developing digital payment solutions or improve the already existing technologies to meet the expectations of the 21st century. The conceptual framework for the study was developed based on the analyzed literature that guided the study. The conceptual model is shown in Figure 1.
The above conceptual model shows six independent variables, namely PE, EE, SI, FCs, SD, and PR. The study incorporated two dependent variables such as behavioral intention to use and actual use. The model integrated a mediating variable, attitude, which mediates the correlation between behavioral intention to use, and two independent variables such as performance expectancy and effort expectancy. The positive and significant role of performance expectancy and effort expectancy as direct determinants of behavioral intention in the adoption of technology has been validated in many studies ( H6: Perceived risk has a positive and significant influence on behavioral intention to use digital payment systems.
H7: Behavioral intention has a positive and significant influence on actual use.
H8: Attitude has a positive and significant influence on behavioral intention to use digital payment systems.

METHODS
This section discussed the methods adopted in the analysis and presentation of the results. The first part was a critical review of the literature on research conducted in the research areas of interest and reference to the adopted UTAUT model. This helped in identifying the study variables and developing research hypotheses.
The study was conducted using primary data. The data were collected using online surveys administered through Google forms. The data were collected from the people living in Bangkok province of Thailand as they were most familiar with using digital payment systems in their retail transactions. The study population consisted of people using or having experience with digital payment systems in their retail transactions such as QR codes, online banking, and mobile payments. A sample of 500 respondents was selected from the population from whom data was collected. A convenience sampling technique was adopted to identify people who have used digital payment systems in their retail purchases and transactions. Data was collected between August 5, 2020, and January 23, 2021.
A structured questionnaire was used to collect data from the respondents. The questionnaire consisted of two sections. The first section contained questions regarding the demographic characteristics of respondents, including age, gender, and occupation. The second section of the questionnaire contained questions regarding the study variables such as performance expectancy (PE), effort expectancy (EE), social influence (SI), facilitating conditions (FC), social distancing (SD), perceived risk (PR), attitude (AT), behavioral intention to use (BI) and actual use (AU). To ensure the validity of the questionnaire, it was submitted to five experts to validate its appropriateness. The questionnaire was designed using a 5-point Likert scale where 1 = strongly disagree and 5 = strongly agree. Once the data was collected, it was evaluated for missing values and outliers. A total of 467 responses were collected, after cleaning the data, a total of 400 responses were validated for analysis.
The data were analyzed using several techniques. The first technique was reliability and validity tests that evaluated the fitness of the model. Another technique adopted was Confirmatory Factor Analysis (CFA) that evaluated the suitability of the model. The relationship between the variables was evaluated using Structural Equation Modeling (SEM). The findings of the analysis were used to develop the discussions and conclusions of the study.

RESULTS
The preliminary result involved the descriptive statistics of the demographic characteristics of the respondents such as gender, age, education level, occupation, monthly income (in THB), and the digital payment system adopted by the respondents. The results are presented in Table 1 and discussed in the following section.
From the data, female respondents were the majority (63.7%) and males were 36.3%. The highest age group was 21-30 years comprising 56%, followed by those below 20 years accounting for 19.8%. Among the three education levels considered, the highest level was bachelor's degree (60.5%) and below bachelor's education accounting for 20.5%. The occupation of the respondents was evaluated, and the highest was company employed comprising 37.3%, followed by students comprising 35.3%. The government officer respondents accounted for 9.5%. The monthly income of the respondents showed that the highest were those earning below 15,000 THB accounting for 35.3%,followed by those earning between 20,001 and 30,000 THB accounting for 26%. The digital payment system consisted of internet banking, mobile banking, tablet application, and mobile application. The largest proportion were those using a combination of mobile app and internet banking (44%), followed by those using only mobile banking (36%). This section presents the model evaluation, this was to ensure that the constructs used in the study, the data collected, and the adopted model were suitable to give objective findings of the study. Several model evaluation methods were used such as the Confirmatory Factor Analysis (CFA), reliability analysis, and validity analysis (they are discussed in the following section).
The first analysis conducted was reliability analysis. In this analysis, the observed variables constructs were evaluated to determine if they were appropriate for the analysis. Reliability was evaluated using Cronbach's Alpha. The results show that the Cronbach's Alpha for the overall constructs was 0.936, which is excellent. When evaluating the independent observables in the Cronbach's Alpha if item deleted section, all constructs in the 0.934 to 0.937 range were found to be excellent. The results confirmed that the constructs were reliable for the study.
Reliability and validity of the model were also evaluated using the Composite Reliability (CR) and Average Variance Extracted (AVE) ( Table 2). According to Figure 2, CR for all variables was above 0.70 and AVE higher than 0.5 (Fornell & Larcker, 1981). Also, given that the Cronbach's Alpha was excellent, the reliability and validity of the variables are considered to meet the threshold in Fornell and Larcker (1981). The last evaluation of the model was carried out using Confirmatory Factor Analysis (CFA). This was intended to assess how well the measured constructs represent the number of constructs. The CFA model is shown in Figure 2.
The Fit indices reported by the CFA results were as follows: p-value = 0.000, CFI = 0.903, TLI = 0.891, IFI = 0.904, NFI = 0.805, RMSEA = 0.067, and X2/df = 2.768. From the above results, CFI, TLI, IFI, and NFI were all approximately greater or equal to the 0.9 goodness of fit threshold (Hu & Bentler, 1999). Additionally, the most reliable measure of the chi-square ratio to degrees of freedom was below 5.0 (X2/df = 2.768). Furthermore, the square root of the difference between the residuals of the sample covariance matrix and the hypothesized model (RMSEA) was below the cut-off threshold of 0.08. Therefore, these indices indicated that the proposed SEM model was fitting well with the study variables (Cselényi et al., 2002;Hu & Bentler, 1999).
After confirming the reliability and validity of the constructs, and the fitness of the proposed model in the previous sections, the next step was to carry out the Structural Equation Modeling (SEM). SEM was aimed at evaluating the research hy-  12 3 pothesis and achieving the objectives of the study. The SEM analysis results are presented in Figure 3 and Table 3.
From the model, the dependent variables include behavioral intention to use (BI) and actual use (AU) of the digital payment systems. Considering the direct effects, Table 3 shows that four variables significantly influenced behavioral intention to use BI. Perceived risk (PR) has a positive and significant influence on BI (β = 0.318, p < 0.01). Facilitating condition (FC) has a positive and significant influence on BI (β = 0.103, p < 0.05), performance expectancy (PE) has a negative and significant influence on BI (β = -0.150, p < 0.01), and attitude (AT) has a positive and significant influence on BI (β = 0.500, p < 0.01). The study also indicated that behavioral intention (BI) has a significant influence on the actual use (AU) of digital payment systems. It is also important to note that performance expectancy (PE) and effort expectancy (EE) has a significant influence on attitude (AT) (β = 0.121, p < 0.05) and (β = 0.121, p < 0.01), respectively. The mediating effect of attitude (AT) between PE and BI and EE and BI showed significant results. AT was found to positively and significantly mediate the relationship between EE and BI (β = 0.047, p < 0.01) and PE and BI (β = 0.168, p < 0.05).

DISCUSSION
This model was aimed at evaluating the BI and actual use of digital payment systems for retail purchases in Thailand from a marketing perspective by applying the UTAUT model. The research focus was to evaluate the factors influencing behavioral intention to use innovative digital payment systems in Thailand. From the findings, 'attitude' has the highest positive influence on behavioral intention to use digital payment systems (β = 0.500, p < 0.01). This implies that if the attitude of the respondents towards digital payment systems increased by one unit, then behavioral intention to use would increase by 0.5 units. These results confirmed H8 that attitude has a positive and significant influence on behavioral intention. The second factor in rank was the perceived risk. The results indicated that if perceived risk increased by one unit, the behavioral intention to use digital payment system increased by 0.318 units, which confirmed H6 of the study that perceived risk has a positive and significant influence on behavioral intention to use digital payment systems.
The results on perceived risk support the results in Ho et al. (2020), though their analysis indicated the indirect but positive and significant effect on behavioral intention to use mobile banking. The findings are also in line with Wong and Mo (2019) who opined that consumers' perception of service as being reliable and honest correspondingly increases the intention to make use of the service because of high levels of confidence in the services being offered. Thus, consumers' trust can significantly influence the intention to use digital payment solutions as a hallmark of innovation in the retail services business. Trust in digital technologies supports improved appraisals and attitudes towards digital payment solutions that are seen as an innovation, especially those in which the technology benchmarks fulfill all set obligations and are trustworthy and relatively meet all safety requirements and protocols, which will, in turn, have a positive influence on the consumers' intention to use such technologies.  However, these results were in conflict with those of Yang et al. (2012), who found that social influence had a significant influence on behavioral intention to use mobile payment services. According to their results, societal influence, for example, from friends, relatives, and other colleagues in the peer group, primarily affects the mindset of consumers, especially with regard to products considered innovative and new to the market, through digital and technological services. Social influence is considered the second most effective factor of consumers' intentions; it is, therefore, vital to encourage consumers' intentions to apply digital payment solutions, since it can foster emotive and coherent viewpoints among consumers in countries such as Thailand.
As a result, H2, H3, and H5 were rejected. The study also evaluated the relationship between behavioral intention to use digital payments and actual use of digital payments. According to the findings, a unit increase in behavioral intention would result in a 0.882 increase in the actual use of digital payments in Thailand.

CONCLUSION
The purpose of this study was to investigate the marketing perspectives of the behavioral intention and actual use of digital payment systems for retail purchases in Thailand. The study is important due to continuous developments in information and financial technology, which has led to advances in e-commerce and inventions, including digital payment technologies. The findings show that Perceived Risk (PR), Facilitating Conditions (FC), Performance Expectancy (PE), and Attitude (AT) significantly influenced behavioral intention to use digital payments in Thailand. The results also highlighted the significant influence of Behavioral Intention (BI) on the Actual Use (AU) of innovative digital payment systems.
This study has both theoretical and managerial implications. First, the UTAUT model was adopted and extended by including other variables such as social distancing perceived risk and attitude as constructs. Including these variables in the model provides valid observations of the results. Second, other studies could adopt this model to compare and evaluate the results. When considering managerial implications, two aspects stand out: perceived risk and attitude. Both variables significantly influence BI to use and the actual use of innovative digital payment solutions in the retail sector. Stakeholders in the financial sector and financial institutions need to consider the aspects of people's attitudes and perceived risk as they influence the adoption and use of digital payment systems in the retail sector. The study is limited by the fact that it was originally conducted in Thailand, so the application of the findings to other areas should be considered with caution.