Digital Payment

Financial technology and e-commerce have revolutionized the global economy by enhancing customer experiences, simplifying transactions, and incorporating online payment systems into the public sector. Particularly in Vietnam, FinTech and e-commerce have seen exponential growth, driven by a tech-savvy population and increasing internet exposure, causing digital payments to be a vital component of the country’s economy. Digital payments have also become a crucial component of Vietnam’s expanding economic sector, driven by the rapid digitalization of financial services and evolving consumer habits or behaviors2. Platforms, particularly Shopee, hold the dominant market share with 63% of the Gross Merchandise Volume (GMV) and have significantly contributed to the transition toward online payment platforms. Shopee excels in seamlessly integrating payments within its system, facilitating a more convenient shopping and payment experience for users. As consumer confidence in online transactions increases, these platforms can capitalize on promotions and user-friendly interfaces, further solidifying their dominance in the Vietnamese e-commerce landscape 3. The rise in online shopping customers, particularly amongst the younger population, has led to a greater acceptance of cashless transactions. Shopee’s dominance in sectors such as home, beauty, fashion, and lifestyle highlights the impact of online payments on consumer spending, demonstrating these systems have integrated into Vietnam’s evolving retail market 3.

Biometric Authentication

Biometric authentication in e-commerce transactions employs unique biological characteristics such as the eye’s iris 4, hand geometry, fingerprints, face, and voice recognition to verify a consumer’s identity 5. This process involves capturing user's above’ traits using a device’s webcam and extracting critical features through Principal Component Analysis (PCA), thereby identifying patterns within the image4. The extracted features would then be encrypted using the RSA algorithm and transmitted along with the user’s transaction details, to the bank system for verification/authentication. The encrypted biometrics data is compared with pre-stored information in the bank system’s database afterward to authenticate the user’s identity, enhance security, and mitigate fraud during transactions4. For users to proceed with the transaction, their biometrics information must undergo the verification process conducted by the bank system associated with the e-commerce platform, and this depends on whether the authentication step succeeds or fails4. The Figure 1 presented based on integrating previous studies 4, 6.

Biometric authentication has rapidly become a critical factor in Vietnam’s e-commerce landscape. This authentication method enhances security and builds customer trust, which are vital to the continuance usage intention (CUI) of online payment systems. Approximately 78% of Vietnamese consumers are recorded to prefer using biometric methods such as fingerprint and facial recognition compared to traditional PINs and passwords, citing these as more secure for identity verification during online transactions​ 7. The increase in e-commerce fraud and identity theft has prompted biometrics adoption in Vietnam, with over 38 million bank accounts and nearly 4 million e-wallets being linked to biometric authentication8. The widespread implementation of biometrics has significantly reduced fraud, as evidenced by a reported decrease in fraudulent bank accounts due to the Vietnamese government’s regulation for mandatory biometrics usage in high-value transactions7.

Through e-commerce platforms such as Shopee dominating the market, consumer trust (CT) has become a critical factor for the CUI regarding online payment methods. Biometric authentication plays a crucial role in reinforcing CT by ensuring that the users’ identities are secure during transactions. This is particularly significant in a rapidly expanding digital payment market dominating the e-commerce sector, with 50 companies providing such services in Vietnam 9. Consumers have higher probabilities toward the CUI of e-commerce platforms given that they possess confidence in financial data’s security, with biometrics serving as a reassurance through offering a distinctive and secure authentication method 10. Ultimately, biometric authentication in Vietnam enhances security, maintains customer trust (CT), and contributes to the continued growth of Vietnam’s e-commerce and FinTech sectors 8.

Theory of Trust

As proposed by Larue Tone Hosmer, the Theory of Trust is significantly crucial to understanding personal, organizational, and economic behaviors. Golembiewski and McConkie have asserted that no single variable has as profound an impact on interpersonal and group behavior as trust 11. Trust can be defined as an optimistic expectation concerning others’ actions and behaviors. This factor is particularly relevant under contexts characterized by dependence and vulnerability. Trust arises from implicit moral obligations that require individuals to safeguard other users’ interests, serving as a critical advocate for cooperation in economic and social interactions 11. Its significance in economic exchange is stated through Hirsch’s reemphasized that “trust was a ‘public good, necessary for the economic transactions’ success” 11. Opportunistic actions within a single market may generate short-term benefits. Nonetheless, these incur long-term costs in the form of diminished trust that can hinder “future acquisitions of cost-reducing and/or quality-enhancing assets”. Trust is thus, the probability that one economic consumer makes decisions and undertakes actions that are beneficial or, at a minimum, not detrimental to another11.

Moreover, Hill concluded that “reputation has an economic value”, highlighting its significant role in impacting others’ willingness to enter an exchange or transaction. This concept fundamentally arose from consistent trustworthy behavior, wherein trust in this circumstance is defined as the economically rational decision to commit to contractual obligations or promises11. Failure to adhere to such actions would ultimately lead to a reputation loss, “thereby diminishing future contracting opportunities” 11. Cummings further asserted that a higher level of trust diminishes the costs associated with monitoring performance and eliminates the necessity for control systems based on short-term financial outcomes11. Nevertheless, such systems could, as referenced by Hoskisson, have undesirable adverse impacts on reducing innovation and collaboration 11. It is significant to acknowledge that “trust did not replace the market or the hierarchy”, rather this factor complements and enhances authority, price, and economic transactions 11. Therefore, a critical aspect of trust’s definition is the expectation that the consequences of breaking trust would far exceed the benefits of maintaining it; otherwise, the decision to trust would merely reflect simple economic rationality 11.

Consumer Trust

Consumer trust (CT) is conceptualized as an exchange of belief between online payment providers and customers to satisfy consumers’ expectations12. In a cashless environment where digital payments have become increasingly prevalent, users’ intentions to adopt these platforms are highly driven by trust13, 14, 15. This fundamental aspect is ranked as the third most significant barrier to e-commerce success that drives consumer activities and engagement16, thereby “generating commitment that leads to strong, long-term” behavior17. A heightened technology fear due to a lack of CT may disrupt the relationship between CUI and actual usage regarding e-commerce platforms17. It is recorded that those having prior Internet usage experience oftentimes accumulate greater exposure to e-commerce, in turn fostering positive and favorable attitudes toward these platforms16. CT, being both a social and personal factor, is based on users’ anticipation regarding the deriving benefits in online payment usage, which can positively and directly impact continued usage intention (CUI) 17. CT is therefore essential for fostering and maintaining “a sustainable competitive advantage, increased revenue, and consumer satisfaction alongside loyalty”, thereby being a significant predictor regarding CUI in the e-commerce context 17.

Perceived Risk

Perceived Risk (PR), a multidimensional construct, is conceptualized as the probability negative outcomes might arise due to an economic event, thereby “impacting various entities such as individuals, businesses, organizations, or governments” 18. This factor has been a central focus in several empirical studies to deepen the understanding of consumer behaviors, particularly in the marketing field. Within the current digital payment context, PR plays a significant role in research concerning the acceptance of new technologies or innovations acceptance alongside shaping consumer behavior and trust. Consumers frequently associate digital payments with potential security vulnerabilities, including risks related to fraud, privacy breaches, and transaction errors19. These risk perceptions can significantly hinder users from engaging in online payment platforms20, and as a result, trust emerges as a vital mitigating factor given the context. PR can therefore be considered a function of uncertainty regarding a given behavior’s potential outcomes and their associated negative consequences21. It represents consumer uncertainty related to the loss or gain in a specific transaction.

Moreover, “the temporal separation between consumers and e-retailers”, challenges in anticipating contingencies and ambiguities in cybersecurity laws have contributed to an inherent uncertainty surrounding online transactions22. Therefore, secure and user-friendly digital payment platforms can therefore significantly reduce PR through implementing strong encryption, transaction guarantees, and effective security measures, fostering consumer confidence and trust in such context 23, 24. Empirical research has indicated that higher levels of CT correspond to lower PR, encouraging consumers to pursue and engage in online transactions frequently25. Within the digital payment context, PR plays a crucial role in influencing consumers’ decision-making, as users weigh the potential dangers against digital transactions' convenience. In circumstances where PR is high, customers would shift away from adopting and continuously using online payment platforms, despite their inherent benefits. Consequently, by effectively managing and minimizing PR through robust security measures, CT can be maintained and ensures the successful implementation of digital payment systems in this era.

Prior studies have asserted that PR is examined through multiple subdimensions 18, 26. This paper therefore examines the impact of six risk facets, including performance, financial, time, social, psychological, and security risks as mediators between CT and biometric authentication CUI. It is crucial to recognize that not every PR component mentioned previously influences the relationship between CT and biometric authentication CUI, as their impacts vary depending on the goods or services involved in online transactions.

Research hypothesis

Perceived Risk and Continuance Usage Intention

PR within the context of technology-based services refers to “the potential for loss in the pursuit of a desired outcome from using the service” 32. Previous research has indicated that as consumers assess a technology-based platform, beliefs regarding the service and its potential usage are formulated, which may encompass risk-related perceptions. These significantly influence users’ assessment of usage risks associated with online payment applications, thereby impacting their intentions toward CUI 32. Adopting digital payment platforms inherently involves risks, including unforeseen negative outcomes or financial losses for. Such that, individuals exhibit greater risk aversion than risk-seeking behavior, PR is an essential variable in anticipating and determining prospective behavior of online payment users 32. Having that said, numerous studies have indicated that PR directly and negatively impacts CUI, highlighting the fact that consumers’ willingness to adopt and use digital payment applications diminishes as their risk perception increases.

Similar to the relationship with CT, an argument exists that PR positively impacts CUI. 33 certain studies have reported that the relationship between PR and CUI is either insignificant or even positively correlated31. Further results suggest that users neglect potential risks due to their perception regarding the possibility that such events would not occur, alongside the “confidence in service providers and governmental support” 31. Consumers therefore are more concerned regarding trust, and oftentimes disregard risks, given that there is reliability towards online payment providers31. Nevertheless, in alignment with common findings regarding the relationship between PR and CUI, the hypothesis is proposed as below:

H3: Perceived Risk (PR) negatively influences Continuance Usage Intention (CUI).

Nonetheless, there is an ongoing argument regarding PR’s role alongside its relationship with CT and CUI. Compared to studies prior to the year 2020, recent research has evoked that CT positively influences PR and, in turn, has a favorable impact on CUI. Therefore further investigation is essential to elucidate the significance and importance of PR among Gen Z residing in Ho Chi Minh City. Thus, another hypothesis is suggested:

H4: The relationship between Consumer Trust (CT) and Continuance Usage Intention (CUI) is mediated by perceived risk (PR).

Consistent with the previously presented arguments and explanations, the conceptual framework is provided in the Figure 2.

Respondent

The target respondents for this study are Gen Z individuals born between 1997 and 2006, residing or studying in Ho Chi Minh City, and currently using digital payment methods alongside authenticating transactions via biometrics. These targeted participants must have a monthly income higher than 5 million VNĐ. With biometric authentication being mandatory in Vietnam’s current context for transactions exceeding 10 million VNĐ, individuals with a monthly income above this threshold have a higher probability of possessing greater opportunities for savings or adequate account balances to facilitate payments that necessitate biometric verification. To diversify the survey’s population, respondents’ majors are classified into distinct categories, including sciences, technology, social sciences and humanities, economics and business administration, law, medicine, and pharmacy, among others.

Instrument Development

The questionnaire content has been translated into Vietnamese to accommodate the targeted participants, with them being native speakers of this language. Before the survey’s formal distribution, a pilot test was conducted with 50 individuals to ensure respondents comprehended the questionnaire substance smoothly and effectively. Afterward, Google Forms is employed as a web-based platform for questionnaire distribution and data collection platform. The Likert scale, ranging from “strongly disagree” to “strongly agree” is employed to measure the question items based on the theoretical framework 34. The survey will be prolonged within a month starting in late July 2024.

Measurement Scales

This study adopted the scale derived from prior research to evaluate the constructs and their respective components. As mentioned previously, PR is a multidimensional concept evaluated through six components, including PER, FR, TR, SR, PYR, and SER. Since this study primarily focuses on digital payment, the adopted measuring scale must align with the research objective. Our study incorporates updated measurement scales specific to digital-oriented systems alongside the original measurement scale,. Referring to PR’s measurement scale, Featherman and Pavlou in the initial successfully developed subconstructs to assess PR, that is measuring PER, FR, TR, and PYR. Rooted in this origin, several academicians have expanded and refined the measurement scale for PR over the past two decades, resulting in six major components, referred to as PER, FR, TR, SR, PYR, and SER. Consequently, including revised measurement scales is essential for this research. CT and CUI are considered as unidimensional variables, each with specific items for measurement. Table 1 presents factors and sources adopted for this paper’s measurement scale.

Measurement Model

Structural Model

Alongside the measurement model, it is crucial to evaluate the structural model, with inner VIF is used for identifying the collinearity’s existence. Subsequently, the statistical significance and relevance of the path coefficients are assessed through bootstrapping for hypothesis testing purposes. The power for independent variables in explaining dependent variables is quantified R-squared values. Furthermore, the effect size is examined through the f-square outcomes with the aim of clarifying the importance of the independent variable over the dependent variables.

The Inner VIF when less than 3.0 indicates the absence of collinearity 39, 40. Table 13 demonstrates that the Inner VIF values are lower than 3.0, confirming that collinearity does not exist between independent variables (CT and PR) and dependent variables (PR and CUI).

Based on the path coefficients in Table 14, both CT and PR significantly influence CUI, as evidenced by p-values below 0.05 36. Notably, CT generates a higher influence on CUI than PR, with both variables resulting in a positive direction towards CUI. The PR’s indirect specific effects as a mediator require further evaluation.

Table 15 provides evidence to conclude that PR mediates the relationship between CT and CUI, with a p-value of 0.01 and an original sample value of 0.05, indicating that PR delivers an indirect effect from CT to CUI. Regarding explanatory power, CT and PR can explain 50% of CUI, whilst their explanatory capacity in the relationship between CT and PR is weaker (R = 0.14) — results are included in Table 16. According to Cohen’s criteria, CT has a more significant impact on CUI, as its f-square approach is 0.73. ​​Conversely, the impact sizes of CT on PR and PR on CUI are negligible, as their f-square values are below 0.2 (Table 17).

Implicitly, CT maintains a significant position in determining CUI, exerting the strongest impact. Consequently, any changes related to CUI in biometric authentication for digital payments should be considered from CT’s aspect. Remarkably, despite the weak influence compared with CT, PR with its components — PER, TR, and SER positively contribute to enhancing CUI. As demonstrated in Figure 3and Table 18, H1 and H4 are supported whereas the results for H2 and H3 are reversely supported. To be more specific, CT positively impacts CUI (β = 0.65, t = 15.78, p-value < 0.05), where this relationship is mediated by PR (β. =0.05, t = 2.44, p-value < 0.05). In contrast, both CT (β = 0.37, t = 7.01, p-value < 0.05) and PR (β = 0.13, t = 2.73, p-value < 0.05) indicate positive impacts on PR and CUI respectively, contrary to the negative direction proposed the hypotheses. Surprisingly, CT increases PR, indicating that higher CT leads to higher demand for PR among users. This positive impact is inconsistent with the majority findings in previous research 29, 30. Nevertheless, our study stands out as one of the limited investigations providing additional evidence supporting the correlation between higher trust and increased risk perception 31. It can be implied that there is an emerging tendency to view it as a notable signal that distinguishes Gen Z from other generations. Accordingly, risk disclosure is preferable among Gen Z’s users, as they believe that an increase in trust toward technology is accompanied by a desire for awareness or information regarding risks, rather than solely focusing on perceived advantages.