Abstract
The research aims to empirically investigate the effect of perceived easiness of use (PEU), perceive usefulness (PU), data security (DAS) on adoption intention of Fintech services through Fintech promotion (FP), and customer trust (CT) in commercial banks of Pakistan using Technology Acceptance Model (TAM) in Pakistan. The data of 297 banking services users were collected through self-administered survey. This quantitative study employs cross sectional research design, and data was analyzed through Partial Least Square (PLS)-Structural Equation Modeling (SEM) technique. The regression results indicate that DAS, PEU, PU, FP, CT have positive and significant influence on adoption intention of Fintech Services. The PEU, DAS, PU also have positive and significant effect on CT. In addition, customers’ perception about DAS and PEU also have a positive and significant influence on customers’ perception about importance of FP. On the contrary, FP has insignificant effect on CT and PU also has insignificant effect on FP. Based on researchers findings, this study would be an addition to the existing literature pertaining to adoption of Fintech services by providing novel understanding about the determinants of users’ behavior for adoption of Fintech services. Therefore, this study could help to regulators that how data security through proper IT infrastructure could help to increase the adoption of Fintech. Moreover, findings of the study also indicates that it is precisely important for Fintech innovators to understand their customers’ attitude regarding transparency of data and its security. Such operational activities enable customers to be aware of adequate use of their data and its secure storage.
Introduction
Financial technology, commonly known as Fintech, has been transforming the financial industry worldwide in recent years. Fintech services are considered as technology-driven innovations that provide a range of financial products as well as services to customers (Park, 2009). These services are intended to improve the efficiency, speed, and traditional financial services convenience (Hu et al., 2019). With the initiation of new and modern technologies and the rise of the digital economy, Fintech has become a crucial part of the financial industry in Pakistan. The E-Banking concept is not different to Pakistan’s financial markets. The first ATM in the country was installed earlier in 1987. Fintech channels dealt with the transactions of 253.7 million worth and among these transactions 53% were processed with ATM which shows the highest share in overall transaction volume. Furthermore, ATMs collectively dealt with transactions of 134.9 million having the value of Rs. 1.7 trillion. In whole, ATM withdrawals are having 96% of transaction volume which is a maximum share in total transactions (Nataraj, 2017). During pandemic, to satisfy the need of customer’s mobile banking networks, ATMs were available as the alternatives to conventional banking. In such challenging situation the encouraging policies implemented by State Bank of Pakistan, the FinTech’s promotion could help to enhance the number of users in near forthcoming (Allen et al., 2021).
This is the reason, reported by “Statista,” showed that in Pakistan, there has been a note worth upsurge in the number of mobile internet and social media users, with smart phones being the largest medium in the financial sector since 2014 (Statista, 2016). Investments in Fintech projects reached US$55.3 billion, which were more than doubled in 2018, (Accenture, 2019). Figure 1 displays the smart phone penetration proportion in Pakistan from 2014 to 2020 as a percentage of connections. Along with these increasing trends, Pakistan is not an exception and has countersigned a significant growth of Fintech services in recent years (Yousaf & Fazal, 2022). These services offer a range of financial products and services to customers such as digital wallets, mobile banking, and online payments. However, the adoption of Fintech services by customers in Pakistan is still in its early stages (Karim et al., 2022; Mehrban et al., 2020). This lack of adoption can be attributed to several factors such as “perceived usefulness (PU), perceived ease of use (PEU), data security (DAS), customer trust (CT), and Fintech promotion (FP)” (Denaputri & Usman, 2019; Hussain et al., 2021; Kazi & Mannan, 2013).
Smartphone penetration rate.
Among above factors, PU means that the point on which people think using a certain technology would increase their performance or productivity (Denaputri & Usman, 2019). According to the Technology Acceptance Model (TAM), PU is one of the key factors that influence users’ intention to adopt a technology (Park, 2009). Different studies have found a significant positive relationship between PU and the adoption of Fintech services (Al-Okaily et al., 2021; Laksamana et al., 2022). Moreover, PU refers to the level where people are more likely to have faith of adopting the service. On the other hand, PU is important factor that influence customers to perceive a particular Fintech service can help them achieve their goals or solve their problems. If customers believe that a Fintech service is useful and can help them achieve their financial goals will decide to adopt Fintech services (Venkatesh & Zhang, 2010). If consumers perceive a Fintech service as easy to use, and useful they are now more likely to use Fintech services and to FP (Hu et al., 2019). Therefore, it is essential to explore the impact of PU on the CT and FP in order to see adoption of Fintech services in the perspective of commercial banks in Pakistan.
Furthermore, PEU is one more critical factor that affects users’ intent to adopt a technology. It refers to the extent to which people think employing a specific technology will be effortless and easy to learn (Davis, 1989). Studies have revealed that PEU and users’ intention to adoption of Fintech Services are positively linked (Laksamana et al., 2022; Tun-Pin et al., 2019). Moreover, PEU refers to the ease with which customers can use a Fintech services. If customers find a Fintech services easy to use, they are more likely to trust and adopt it. Because, when customers perceive a Fintech service is easy to use, their trust in the service is likely to increase, leading to increased adoption. This is because customers are more likely to trust a service that they perceive to be helpful and easy to use. On the other hand, PEU is also an important factor that can influence consumers’ decision to adopt Fintech services Venkatesh et al. (2003). If consumers perceive a Fintech service are easy to use, then they are more expected to promote Fintech services and will adopt Fintech services (Lemma, 2020). Therefore, it is necessary to investigate the impact of PEU on the customer FP to adoption of Fintech services in the setting of commercial banks in Pakistan.
Additionally, DAS is also a crucial element that consumers consider before adopting any Fintech service. DAS refers to the personal protection of financial information from unauthorized disclosure or access. In the context of Fintech services, DAS is a critical concern for users as they are required to provide sensitive materials like personal identification information, bank account details and credit card information (Cao, Wang, Ding, Lv, Dong et al., 2021; Cao, Wang, Ding, Lv, Tian et al., 2021; Li et al, 2022; Lv, Wu, Li, & Song, 2022; Mannan & Van Oorschot, 2008; Oly Ndubisi & Sinti, 2006; Zhang et al., 2023). Studies have shown that CT and adoption of Fintech services are significantly influenced by their perceptions of DAS (Abdul-Halim et al., 2022; Dwivedi et al., 2019). The study based on TAM, Shahzad et al. (2022) investigated behavioral intentions of the consumers while using Fintech service, by integrating brand and service trust together. The study concluded that DAS, CT, their interface designs, value added, and FP were main factors to be attributed toward a progressive adoption of Fintech services in Germany. The previous studies identified that when the trust is improved then the Fintech services also gets improved because CT is confidence that an individual has on the technology to meet their expectations (Chuang et al., 2016). In the matter of Fintech services, CT could be seen as the degree to which customers believe that these services are safe and reliable. Therefore, the adoption of Fintech services by customers is dependent on their trust in these services (Hallikainen & Laukkanen, 2018) which could increase the Fintech services (Roh et al., 2022).Additionally, FP that highlights robust DAS measures can positively impact their adoption rate, leading to increased usage and growth in the Fintech industry (Nangin et al., 2020). Few recent studies have identified important finidngs in terms of risk prediction and customer intentions (Li, Wang & Yang, 2023; Meng, Xiao & Wang, 2022), reliable and secure communication (Cao, Wang, Ding, Lv, Dong et al., 2021; Cao, Wang, Ding, Lv, Tian et al., 2021; Yu et al., 2021), credit rating (Li & Sun, 2021), multiscale feature in online services (Lu et al., 2023; Qin et al., 2022; Zheng et al., 2022; Ni et al., 2022), credible data on block chain (Yan et al., 2021), data privacy and security (Chen et al., 2021; Lv, Cheng & Song, 2022; Ma & Hu, 2022), technology innovations (Gong & Rezaeipanah, 2023; Wang et al., 2022; Liu, 2021; Zhenggang et al., 2022; Lv et al., 2020; Cheng et al., 2021). Hence, it is necessary to investigate the impact of DAS on the CT and FP to adoption of Fintech services in the context of commercial banks in Pakistan.
Prior researches have several gaps, like previous studies have inconsistent empirical findings (Farida, 2022; Saleem, 2021; Singh et al., 2021; Suprapto, 2022). Furthermore, previous studies primarily focused on developed economies (Ioannou & Wójcik, 2022; Moreira-Santos et al., 2022; Singh et al., 2020), neglecting developing economies, especially Pakistan (Oladapo et al., 2022; Zakariyah et al., 2022). Although some studies have been conducted in Pakistan (Ali et al., 2021; et al., 2022), they focused on other indicators rather than the combined effect of DAS, PEU, PU, CT, and FP on the adoption of Fintech services. Hence, our research aims to fill the literature gap and analyze customers’ perceptions about DAS, PU, PEU, and their importance for CT and FP influencing their intentions to use Fintech services in Pakistan. Our study builds upon earlier research conducted by Chuang et al. (2016), Steward and Jürjens (2018), and Hu et al. (2019). Previous studies also shown lack of adoption of Fintech services in Pakistan can be attributed to several factors such as PU, PEU, and DAS (Siyal et al. 2019). Therefore, the research objective is to inspect the impact of PEU, PU, and DAS on adoption purpose to Fintech services through FP and CT in commercial banks of Pakistan.
The research finding contributed literature in the extant works on the Fintech services adoption in Pakistan. The study will also provide insights into the factors which affect the implementation of Fintech service facilities and will also help commercial banks in Pakistan to understand the customer’s needs and preferences. The study will also help commercial banks in Pakistan to design and implement effective strategies to promote the implementation of Fintech services by their customers. Moreover, the findings of this study will also be useful for commercial banks in Pakistan to design and apply effective approaches to promote the implementation of Fintech services by their customers.
The research paper was divided in five sections. In this study the first section contains introduction of the study. The next section based on literature review, and the third section comprises of research methodology, fourth section based on data analysis and interpretation, and fifth section describes the discussion and conclusion.
Literature Review and Conceptual Framework
Ajzen and Fishbein (1980) recommended the Theory of Reasoned Action (TRA), it concentrates on figuring out what influences a person’s conduct while embracing new technology. Specially, TRA recognizes attitudes and particular norms as important pointers of person’s intentions to adopt such specific technologies (Fishbein & Ajzen, 1975). This model recommends that the comportment of an individual is decided by their attitude and subjective norms, which represent the recognition of benefits associated with performing the intended behavior. However, TRA may not be appropriate for our study as it is limited in its ability to predict behavior when individuals have voluntary control over their actions (Ajzen, 1991). Furthermore, TRA does not consider the effect of specific attitudes on an individual’s beliefs, which is an important consideration in our study.
Davis (1989) proposed the Technology Acceptance Model (TAM) to explore the impact of external variables on an individual’s internal behavior and beliefs (refer to Figure 2). The concept states that a person’s attitude to use information systems affects their intentions and behavior, a finding which has been confirmed by Yang (2005) by means of the most valid, robust, and strongest model in the relevant literature on technology adoption. TAM identifies two primary elements that influence an individual’s behavioral attitude toward technology adoption—PU as well as PEU. These elements are thought to significantly affect decision of individual person to adopt new technology tools (Venkatesh & Bala, 2008). PU represents a person’s belief that implementing a certain technology will improve their ability to execute their job, while PEU represents an individual’s belief that adopting a specific technology would make their work efforts easier (Davis, 1989).
Technology Acceptance Model (TAM) (Davis, 1989).
Overall, the “Technology Acceptance Model (TAM)” emphasizes the importance of” PEU and PU in deciding an “individual’s attitude” and aim to adopt new technology, such as Fintech services (Chuang et al., 2016; Riquelme et al., 2010). These factors have been shown to have positive effects on attitude and willingness to adopt new technology, especially when access to the technology is made easy and when users trust the service providers. Similarly, comparison shown between “Theory of Planned Behavior” (TPB) model, “Decomposed Theory of Planned Behavior” (DTPB) model and TAM model describe in the study of Taylor and Todd (1995), by using a computer resource center, concluded that PEU has positive effects on PU (Basak et al., 2016; Koksal, 2016). Hanafizadeh et al. (2014) and Hu et al. (2019) have also drawn attention to the unintended consequences of using Fintech services. Therefore, the TAM model can be useful in predicting and understanding consumers’ behavior in the direction of the use of Fintech services.
As TAM offers a thorough knowledge of the elements that affect customers’ propensity to accept current technological systems, it has been extensively utilized in nowadays studies, specifically in terms of their PU and PEU (Zhang et al., 2018). The model has been adapted and refined over time to suit different research contexts and issues of analysis, making it a flexible and useful tool for researchers studying technology adoption.
On the contrary, Luarn and Lin (2005) attempted to shed light on the constraints in TAM. They argued that TAM focus only on PU and PEU but ignores various constraints that cause hindrances in proper utilization of information technology system. Even Liu et al. (2009) interrogated the connotation of TAM in use of mobiles in banking services, and accordingly highlighted several effects observed in the use of computer-based as well as wireless-based systems. Luarn and Lin (2005) also emphasized the scope in advancement of TAM, for the incorporation of an important element, viz., “trust,” and two-asset components (“perceived self-viability” and “perceived financial costs”). Their research concluded that such trust showed indirect effect toward the intention of the customers to go for mobile banking adoption, which is initially based on PEU. However, information and awareness regarding data is insufficient on security actions for Fintech transactions and activities, leading to slower pace of its adoption among users. Besides, Wang et al. (2003), Tang et al. (2004) have emphasized to adopt system of mobile banking through utilization of TAM as a blueprint only, simply by including data security measures and various protection concerns of the customers. Furthermore, Luarn and Lin (2005) also highlighted that the risks of data security measures and concerns regarding data transmission were considered as vital factors and significant elements, which placed adequate impact on those users who are willing to adopt electronic messaging and conveyance channels.
Clark (2002) and Lanford (2006) emphasized that usability and user design interface should be considered as extra elements that required incorporation to describe users’ data security concerns. Therefore, as proposed by Steward and Jürjens (2018) we explore the elements affecting behavioral intention of customers to adopt FinTech services in Pakistan for the incorporation of various parts within data security measures and their trust, through the extension of TAM. Our study is obviously a consistent continuation of the earlier existent research works undertaken and prepared by Chuang et al. (2016), Steward and Jürjens (2018), and Hu et al. (2019). If we assume that these factors are those elements which have great influence on the behavioral willingness of customers following adoption of this service. As outlined in the Figure, expansion of TAM is internal as well as, external elements that focus a clear influence through the Fintech’s adoption services and denoted by a determinant, value added (VA), and presented accordingly in our model. Both the important TAM elements (PU and PEU), constructed therein, represent those internal elements that tend to determine Value Added aspect. Thus, to evaluate customers’ perception about important factors (trust, customers’ perception about data security, PU, customers’ perception about the importance of FP and PEU), which affect their intentions to adopt Fintech services in Pakistan; the following hypotheses have been identified (see Figure 3).
Conceptual model.
Research Methodology
In the extant literature, two methodologies are used in analysis, the one is qualitative and the other is quantitative. The quantitative data analysis method is based upon numerical data and procedures. In contrast, qualitative analysis is based upon narrative procedures and descriptive data (Berrios & Lucca, 2006). At the same time, quantitative study is an approach for assessing the link among variables in objective theories. In turn, these variables, normally on instruments that could be calculated in such a way as to analysis numbered data with statistical procedures. The final paper consists of an introduction, literature and theory, methods, findings, and discussions. Therefore, based on this discussion, in this current study we applied the quantitative research methodology. Moreover, “cross sectional research design” had been applied in the current study for which data had been collected for the first time. The cross-sectional research design is considered to be useful and better for survey analysis. The questionnaire has been adopted and modified (Chuang et al., 2016; Hu et al., 2019; Stewart & Jürjens, 2018).
Questionnaire Description
To evaluate the proposed model, a comprehensive questionnaire was developed for quantitative analysis. The questionnaire was administered in English, the official language of Pakistan, and comprised of two parts. The first part gathered demographic information about the respondents, including age, gender, education, income, occupation, and marital status. The second part measured customers’ perceptions of perceived usefulness (PU), data security (DAS), perceived ease of use (PEU), customer trust (CT), Fintech promotion (FP), and adoption intention of Fintech Services using a “five-point Likert scale.” CT, measured with 4 items, is crucial for a bank’s engagement and customer loyalty (Hu et al., 2019; Stewart & Jurjens, 2018). DAS, measured with 14 items, protects digital data throughout its lifecycle against unauthorized access, corruption, or theft (Li et al., 2021; Stewart & Jürjens, 2018). PU, measured with 7 items, reflects the user’s subjective perception of a technology’s potential to enhance work performance (Hu et al., 2019). PEU is measured with 3 items, assesses the ease of using a system (Davis, 1989); and it was measured by adopting 3 items from the study of (Hu et al., 2019). FP, measured with 3 items, enhances the delivery of financial products and opens up new markets (Stewart & Jurjens, 2018). Fintech adoption intention, measured with 3 items, uses communication accessibility, safe and easy financial transactions, internet use, and automated data processing to facilitate commerce in the financial industry (Dwivedi et al., 2021; Hu et al., 2019). The questionnaire is attached in Appendix 1.
Population and Sampling Technique
To evaluate the perception of the customers with respect to various important factors like, data security, perceived usefulness, Fintech promotion, trust and perceived ease of use while in intent to adopt Fintech. We collected the data by using the “convenience sampling method” through a survey by using a questionnaire. The questionnaire consisted of 33 relevant questions to obtain the requisite information. Out of the 375 respondents from the original sample, only 313 duly completed questionnaires were received, wherein 16 questionnaires were found to be incomplete and hence, wasted. Thus, only 297 questionnaires were accepted for analyses.
Data Analysis and Results
The study model was assessed using SPSS as well as Smart PLS following the recent studies (Siyal, 2023; Siyal et al., 2020, 2022, 2023; Siyal & Peng, 2018; Siyal, Saeed et al., 2021; Siyal, Xin et al., 2021). Adequate analysis of a theoretical model is possible with variables as one of the fundamental factors for understanding systemic existence and causality direction (Esposito Vinzi et al., 2010). Our study defines best statistical approaches to be used for a comprehensive understanding and to enhance the structure of the model (Henseler et al., 2015). We have used reflective transformation right from the construct toward the indicators in our analysis, (1) the variable are the gauges and/or variables observed. The variables of this form are defined by: 1) the position or effect of the reflection or representation of the unrelated structures; 2) the reflective variables are categorized because their construction indicators are strongly correlated; they can be interchanged and, if it is removed, content of construct was not altered (Wetzels et al., 2009).
Descriptive Statistics
SPSS was used to analyze 297 responses from 375 questionnaires distributed, yielding a 79.2% response rate. Most respondents were male (186) and aged among (31–40) Majority of respondents were married (225) with a minimum graduation qualification (144). Most respondents were employed (174) and earned between PKR 51,000/- to PKR 75,000/-. Table 1 presents the results.
Descriptive Analysis (N = 297).
Source. Own illustration.
Common Method Biased and Multicollinearity
Common method bias (CMB) is a type of methodological bias that can occur when a single method or instrument is used to measure multiple variables in a study. This can lead to counterfeit correlations or associations between the variables, as they may be influenced by shared methodological factors rather than their actual relationships (Podsakoff et al., 2003). CMB can be a particularly insidious form of bias, as it can go unnoticed or unaddressed by researchers. It can occur in any type of research that involves self-reported or survey data, as respondents may be impacted by a lot of factors, such as social desirability, acquiescence bias, or response set bias. To mitigate the effects of CMB, researchers can employ a number of strategies. One approach is to use multiple methods or instruments to measure the variables of interest, in order to reduce the influence of shared methodological factors. Another approach is to control for potential confounding variables that may be associated with both the predictor and outcome variables, such as demographics or personality traits. Several statistical methods have also been developed to detect and correct for CMB, such as the “Harman single-factor test,”“partial correlation analysis,” and “structural equation modeling”(Kock, 2015). Therefore, single-factor test developed by Harman was used to check for CMB issues. As a consequence, the findings showed that the total variance explained by all components was 35.26%, which was under the crucial criterion of 50%. As a result, the CMB is not suspicious in the available data. Consequently, based on these findings, the CMB is reliable. In other context, value of Variance inflation factor (VIF) is also less than 3.33 which also indicates that there is no issue of CMB (Kock, 2015). The CMB findings are comprised in Table 2.
Common Method Biased.
For the multicollinearity, VIF is a diagnostic measure used to evaluate the degree of multicollinearity in a PLS-SEM model. It is commonly used in Smart PLS software to identify the manifestation of multicollinearity among the predictor variables Hair et al. (2017). VIF gauges how much the association among the predictor variables inflates the variance of an estimated regression coefficient. VIF values greater than 5 or 10 indicate the presence of significant multicollinearity, which can lead to biased parameter estimates and decreased predictive power of the model Hair et al. (2017). All VIF values were smaller than 5 which indicate that Multicollinearity issue is not present. The VIF values are prophesied in the following Table 3 as under.
Variance Inflation Factor.
Inferential Analysis
The inferential statistics has been comprises using Smart PLS.3 which tested the hypothesis of this study. The smart PLS results were tested using Partial Least Square (PLS)-Structural Equation Modeling (SEM) technique was employed where analysis was done in two models, that is, measurement and structural Model. By adopting this technique, it would allow the researcher to draw a conclusion by undertaking theoretical justification rooting through empirical results with the PLS-SEM method. With this, it would permit the researcher to adopt this method and gain results once the theoretical construct is deemed valid, as inked by Hair et al. (2006).
Measurement Model
While using measurement model, certain factors have been kept into consideration to further support and prove the hypothesis: individual reliability (load), reliability of construct internal consistency and scale (as prescribed under the principle of “Cronbach’s alpha and composite reliability”) (Brown, 2002). In the first instance, Cronbach’s Alpha test was performed where the outcome was considered satisfactory; outcome was above .700 (Hair et al., 2006). In the second instance, the composite reliability analysis was conducted that proved passable than the Cronbach’s alpha test for the PLS technique, as all indicators receive a similar weightage (Chin, 1998; Henseler et al., 2016), it is a singular consistent reliability measure stated by Dijkstra and Henseler (2015). For the AVE suggested value is set to .5, as per the findings brought forth by Hair et al. (2006). Table 4 legibly reflects values which shows greater values than those recommended above. To summarize, the constructs had a convergent validity.
Construct Reliability and Validity Measures.
Note. CT = customer trust; DS = data security; FA = Fintech adoption intention; FP = Fintech promotion; PEU = perceived ease of use; PU = perceived usefulness.
Discriminant Validity
The term “discriminant validity” had been evolved through some theoretical concepts pertaining to validity that focus on the construct and determines whether any highly correlated constructs exist among them or not, in the model. Also, it compares AVE Square root with one specific construct toward the association of other constructions (Huet al., 2019). A general suggestion exists to show that AVE square root values should be greater than other diagonal (Fornell & Larcker (1981), as revealed in Table 5. Hence, this indicate that the construct has discriminant validity.
Discriminant Validity of Constructs.
Note. CT = customer trust; DS = data security; FA = Fintech adoption intention; FP = Fintech promotion; PEU = perceived ease of use; PU = perceived usefulness. Bold values are discriminant values of the respective constructs.
Structural Model
Commonly goodness-of-fit (GOF) measures used in PLS-SEM which represent the overall measure of how well the model fits the data. The GOF could be assessed from the following criteria’s, R Sqaured, Q squared, and Standardized Root Mean Square Residual (SRMR) (Hair et al., 2017). Among these criteria’s R-Squared shown the variance in endogenous variable due to exogenous variables in the model. According to Hair et al. (2017) for reflective measurement models R-square value of 0.25 or higher is reflected as a good fit. On the other hand, Miles (2005) further explained that R square value 0.26 signifies substantial influence, 0.13 shows moderate influence while 0.02 shows weak influence. All R square values were greater than 0.26 which indicates that models shows good fit. On the other hand, Q-Squared which measure of predictive accuracy for endogenous constructs. Hair et al. (2017) suggest that a Q-square value of 0.25 or higher is considered a good fit for PLS-SEM models because the minimum recommended values for Q Square is that this values should be greater than 0. The Q square values falls in different ranges, for instance 0.35 indicates larger predictive relevance, 0.15 represents medium predictive relevance, while 0.02 represents smaller predictive relevance for the dependent variable (Hair et al., 2017). The Table 6 predicted values indicates that all Q Squared values were greater than 0.15 which indicates that model is a good fit. Additionally, Standardized Root Mean Square Residual (SRMR) should be used in conjunction with other goodness-of-fit measures such as R-square, and Q Squared to assess the overall fit of the model. SRMR is a measure of the average difference between the observed and predicted covariance’s, with values close to 0 indicating good fit (Hair et al., 2017). The values was near to 0 which also representing the good of fit model. These above measures reported in Table.6 summarize PLS-SEM model GOF.
Goodness of Model.
After testing the goodness of model, the next step is to test the hypothesis of the study. For this purpose, the PLS-SEM is a type of statistically testing method to investigate relations among variables through their co-variance matrix. It is also considered as a very significant element for analyzing multivariate data. Following the validity and reliability tests, empirical research for Fintech’s adoption model, based on sample analyses was conducted. The PLS-SEM were used to test our hypothesis and accordingly the t-value and Standardized Path Coefficient (β) were calculated through Smart PLS, by using PLS-SEM model. In broad terms, t > 1.96, then the “coefficient test” proved as significant at confidence level of p < .05. When t is more than 2.58, then coefficient test considered to be significant at confidence level of p < .01. Finally, when t > 3.1, then coefficient test showed significant at confidence level of p < .001 (Hu et al., 2019). The estimated results of our hypotheses had been tested accordingly and presented in Figure 4.
Structural results of the proposed model.
Customer Trust (CT) (β = 0.665, t = 10.081), Data Security (DAS) (β = 0.286, t = 2.679), Fintech Promotion (FP) (β = 0.097, t = 1.60), Perceived ease of use (PEU) (β = 1.03, t = 10.851), and Perceived usefulness (PU) (β = 0.128, t = 3.683) all have a positive significant impact on Fintech Adoption intention. Thus, hypotheses H1, H4, H6 H8, and H11 are supported. Similarly, DAS (β = 0.315, t = 6.508), PEU (β = 0.432, t = 7.444), and PU (β = 0.832, t = 31.811) has a positive significant influence on CT. Therefore, H3, H7, and H10 are supported. On the contrary, FP (β = 0.066, t = 1.475) has an insignificant impact on CT. Moreover, DAS (β = 1.572, t = 28.772) and PEU (β = 0.788, t = 10.297) have positive as well as significant influence on FP, while PU (β = 0.012, t = 0.413) has an insignificant influence on FP. Thus, H5 and H9 are supported. However, H2 and H12 are not supported since p > .05 and t < 1.96. These discussed results are predicted in Figure 4.
Discussion
Financial institutions worldwide have invested heavily in upgrading their IT infrastructure to improve their efficiency, but there is still uncertainty around the potential ROI (return on investment) of these expenditures. Pressure from customers and competitors to improve IT infrastructure has led to many banks adopting FinTech to increase profits, better serve customers, and advertise more effectively. However, ensuring data security (DAS) on FinTech services is crucial to maintain customer trust (CT) and increase fintech promotion (FP), which can in turn increase fintech adoption. “Perceived usefulness (PU) and perceived use of ease (PEU)” also play important roles in increasing CT and FP (Stewart & Jurjens, 2018; Wilson et al., 2021). Therefore, this study aims to empirically examine the key factors influencing of financial technology adoption in Pakistan using TAM. Moreover, data was collected from bank users in Pakistan and employed PLS-SEM technique.
Being consistent with the research results of Hu et al. (2019), Stewart and Jürjens (2018) and Sikdar et al. (2015), our paper also shows that CT has an significant influence on purposes to adopt Fintech services. DAS also has significant and positive effect on CT, FP, and intent to Fintech services adoption. It is evidently found that customers’ perception about existence DAS significantly influences on CT, FP, and intention to adopt Fintech services. Customers’ perception about the importance of FP does not influence CT, but contrary to Hu et al. (2019) and Stewart and Jürjens (2018) we found that FP does influence on intentions to go for Fintech services.
The outcomes further designated that (PEU) also has positive as well as significant influence on CT, FP, and use of financial technology services. Hence, the results of this experiment are congruent with those of Sikdar et al. (2015) and Wilson et al. (2021) who clearly revealed that PU had significant influence on CT and purpose to adopt FinTech services, though DAS affected CT. The execution of trust simply guides users to go for Fintech services since PU have significant and positive effect on trust, showing essential role of data security in reducing trust level. Financial institutions necessarily required to take strong steps and measures for reducing users’ cyber security concerns for excessive trust in products and services, so as to increase their willingness to go for such services. The results also show that PEU seems to have a good and substantial impact on CT and FP and adoption of Fintech services. Findings remain constant with the earlier investigation of (Wilson et al., 2021) indicating that PEU leads to intentions to go for use of financial technology services. On the other hand, scholars assumed that PEU in early stages of Fintech services adoption or technology, has less influence on users’ attitude, due to its unfamiliarity or lesser opportunities in using it (Davis,1989). There are clear reflections that in current scenario, Pakistan is not in primary stages of Fintech services, since many bank users showed experience in its usage.
Based on our results, we can precisely propose that DAS, PEU, and CT strongly affects Fintech adoption intention, whereas PEU and PU also strongly influence on CT. Likewise, PEU and DAS create strong influence on FP. Hence, these factors need more attention for effective planning in future to divert CT and confidence toward Fintech. With regards to PU and PEU, Egger (2002) argued that users seem more inclined toward attractive products, showing best standards of usability for creating trust. They look toward simple preventive methods from fraudulent measures for DAS which increases their trust, which in turn, tends to enhance their diversion toward Fintech adoption intention. Likewise, DAS holds stronger impact on FP. According to Stewart and Jürjens (2018), if more knowledge and adequate assurance on DAS is provided to customers, then their trust toward Fintech will obviously increase. Therefore, innovators of Fintech should appropriately understand attitudes of customers regarding data security and increase its transparency for customer awareness how the data is being stored and used safely.
Conclusion
In this research, we have empirically analyzed “influence of perceived usefulness (PU),”“perceived ease of use (PEU),”and “data security (DAS)” on “adoption intention of Fintech services” through customer trust (CT) and Fintech promotion (FP) in Pakistan commercial banks. For data analysis, PLS-SEM was used in two measurement and structural models. Results indicated that PEU, PU, CT, DAS, and FP, saw a significant impact on the adoption of Fintech services. Likewise, DAS, PU, and PEU had also significant effect on CT. On the other hand, PEU, and DAS also have positive and significant effect on FP. In contrary, The FP did not has positive and significant effect on CT and PU also did not has significant effect on FP. Thus, the significant and positive outcomes of the study indicate that the customers’ perception of PU, PEU, FP, DAS, and CT for intention to Fintech adoption is crucial to enhance the use of their Fintech services. As a result, the research gives significant knowledge on variables impacting “adoption of Fintech services” in Pakistan and highlights the significance of PEU, PU, FP, DAS, and CT in shaping customer aim to adopt such services. These insights could be used by commercial banks and Fintech services providers to improve their services and increase their customer base.
Implications and Future Directions
The present research contributed in the extant literature because previous researches have considered very few aspects that could affect the “adoption of Fintech technology,” such as data security (DAS), Fintech promotion (FP), and customer trust (CT) (Stewart & Jurjens, 2018). While have little attention on “perceived usefulness (PU) as well as perceived ease of use (PEU).” In particular, in the setting of Pakistan. However, the present research is contributing a significant part into the existing literature as it covers factors that were not considered in previous studies, especially in the perspective of developing countries like Pakistan. The study’s findings can provide theoretical as well as practical ramifications that could aid regulators and scholars in comprehending the significance of DAS, PEU, PU, and CT in increasing the adoption of FinTech. The study’s practical implications can assist regulatory bodies in improving their policies and permit banks to attain economies of scale for worldwide concentration. The study’s results could also help financial institutions improve their use of information technology, which is critical in the current competitive landscape. The results of this research have significant ramifications for Pakistani commercial banks and Fintech service providers as well. Therefore, results suggest that these institutions should prioritize the development of robust data security measures to address the customers’ concerns and build trust in their services. They should also focus on building customer trust by providing transparent and reliable services to increase the intention to adopt Fintech services.
Hence, Fintech innovators have active potentials to optimize the process of financial inclusions with predictive high expectations through financial solutions. However, this research has some limitations; this research was limited to Pakistan and has only focused on the individual-level factors affecting Fintech adoption, neglecting the organizational-level factors. Future research can consider these factors, such as organizational culture, readiness, and support, to understand how they influence the adoption of Fintech by financial institutions. Furthermore, this study has only focused on the aspects influencing the intent to adopt Fintech, neglecting the actual adoption behavior. Future research can consider the actual adoption behavior and its determinants to have a clearer comprehension of adoption procedure. Moreover, current research has only focused on the banking sector, neglecting other sectors, such as insurance and investment, where Fintech adoption is also prevalent. Future research can consider the factors affecting Fintech adoption in other sectors to have a holistic understanding of Fintech adoption in different industries. Furthermore, Future research might explore the effect of demographic parameters including age, gender, income, and education level in affecting customers’ intentions to embrace Fintech services as this study ignores their influence on intention to use Fintech services. Additionally, to examine the variations in consumer perception and acceptance of Fintech services among various age groups, genders, income levels, and educational backgrounds, a comparative research might be carried out. Finally, longitudinal studies can be conducted to track the evolution of customers’ insight and adoption of Fintech services over time, providing insights into the future of financial services in Pakistan.
Recommendations
The following suggestions should be made to commercial banks and Fintech service providers operating in the area depending on the findings of the study: As the research discovered, consumer trust and propensity to consume Fintech services were significantly positively impacted by perceived utility, therefore commercial banks and Fintech service providers should prioritize the development of services that offer tangible benefits to customers. This could include increasing the speed and convenience of transactions, offering personalized services, and improving access to financial products. The study also found that data security significantly influenced customer trust, which in response positively impacted aim to adopt Fintech services. Therefore, commercial banks and Fintech service providers should prioritize the development of robust data security measures, such as encryption, authentication, and access control measures, to protect customer data from unlawful access. Moreover, to build trust among customers, commercial banks and Fintech service providers should provide transparent and clear information about their services, including how customer data is used and protected. They should also communicate their security measures and data protection policies clearly to customers to build their trust. It is also recommended that there should be a collaboration between commercial banks and Fintech service providers which can enhance the adoption of Fintech services by building trust among customers. Commercial banks can leverage Fintech service providers’ expertise to offer innovative and personalized services, while Fintech service providers can leverage commercial banks’ brand reputation and customer base to increase their reach. Overall, the recommendations outlined above can help commercial banks and Fintech services providers in Pakistan improve their services and increase customer adoption of Fintech services.
Footnotes
Appendix
Survey Instrument.
| Construct | Questions | Source |
|---|---|---|
| Perceived usefulness | 1. Fintechuse could meet my needs of service. 2. Fintech services help in saving time. 3. Fintech services help in improving efficiency. 4. Generally, Fintech services are beneficial to me. 5. Fintech services are easy to use. 6. Use of Fintech services could meet my service needs. 7. Fintech services help to save my time. |
Hu et al. (2019) |
| Perceived ease of use | 1. I think the operation interface of Fintech is understandable and friendly. 2. Its equipment are easy to use for Fintech services. 3. It is quite easy to access the equipment to practice Fintech services. |
Hu et al. (2019) |
| Customer trust | 1. I trust services of Fintechretain my private information secure and safe. 2. In total I am sure thatFintech services are trustable. 3. This bank provides good products and services. 4. I trust in my financial security whenever using Fintech services. |
Hu et al. (2019) |
| Fintech services adoption | 1. If I already used Fintech services, then I am agreeing to remain using them. 2. I would like to use Fintech services soon. 3. I will praise Fintech services to my friends. |
Hu et al. (2019) |
| Data security | 1. I believe that banks provide enough security on their websites or mobile applications. 2. My bank provides compensation for losses due to mobile banking fraud. 3. I believe that I am secured using mobile online banking. 4. I believe that bank should take steps in order to increase mobile banking security. 5. I have major concerns with data security when using mobile banking. 6. My bank is responsible for my security while I am using mobile banking. 7. Security aspect is most important concern for me when using mobile banking. 8. I think that my bank’s mobile banking is safe. 9. There are secure infrastructure platforms in mobile baking systems. 10. Application downloaded from my mobile phone’s app is good enough to secure my data. 11. I carry out my banking transactions on my mobile device safely. 12. I am much concerned about the security of mobile banking. 13. Confidentiality of information is protected in mobile banking systems. 14. I believe, while using my mobile banking someone may capture my calls or data. |
(Stewart & Jürjens, 2018) |
| Fintech promotion | 1. Promotion is important to increase my interest in mobile banking. 2. Fintech services offer better value for money compared to traditional financial services. 3. I am likely to recommend fintech services to others. |
Li et al. (2021) and Stewart and Jürjens (2018) |
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Social Science Fund of China of the Youth Project “A Comparative Study on the Laws of Global Cyberspace Security Governance and Its Enlightenment to China” (Grant No. 19CXW039). The National Social Science Fund of China Project “Research on the Multi subject Collaborative Governance System of Internet Information in China ”(18BXW121). The Ningbo Yongjiang Talent Introduction Project Innovative Talents Project: “Governing the Internet by Law, Benefiting the Internet by Virtue and Cooperating with the Five Governance: Theoretical and Practical Research on the Construction of a Comprehensive Network Governance System ”(2021B-009-G).
