Understanding mobile e-wallet consumers ’ intentions and user behavior

Purpose – This study aims to investigate the factors that in ﬂ uence behavioral intention (BI) and usage of e-wallets by extending the uni ﬁ ed theory of acceptance and use of technology (UTAUT) with constructs, namely, mobile self-ef ﬁ cacy, perceivedenjoymentandsatisfaction. Design/methodology/approach – This quantitative study used partial least squares structural equationmodeling ona sample of 576 mobilee-walletusers surveyed online. Findings – The key ﬁ ndings indicate that the model can explain 58.8% of the variance in behavioral intention and 53.8% in usage. Moreover, mobile self-ef ﬁ cacy has a signi ﬁ cant in ﬂ uence on perceived enjoyment. Perceived enjoyment signi ﬁ cantly affects satisfaction, effort expectancy and performance expectancy. Furthermore, effort expectancy signi ﬁ cantly in ﬂ uences customer satisfaction in contrast to performance expectancy. In addition, although performance expectancy, social in ﬂ uence and satisfaction signi ﬁ cantly impact consumers ’ behavioral intention, effort expectancy and facilitating conditions condition have an insigni ﬁ cant in ﬂ uence on consumers ’ behavioral intention. E-wallet stakeholders can use the ﬁ ndings of this study to make strategic decisionsregarding thee-walletecosystem. Originality/value – Although previous studies have independently addressed the impact of mobile self-ef ﬁ cacy, perceived enjoyment and satisfaction on consumers ’ behavioral intention and usage behavior, the expanded framework with the possible relationships proposed in thisstudy has never been adequately studied in previous research in the context of e-wallets in developing countries based on an empirical analysis. This study representsoneofthe ﬁ rstattemptstoimprovetheUTAUTbyempiricallyanalyzingtheserelationships.


Introduction
Digital transformation, financial inclusion, Fintech and e-wallets have risen with tremendous developments in information and communications technology and reliance on smartphones to access the internet.Esawe and Elwkeel's (2020) findings demonstrate that digital transformation is a dynamic phenomenon that evolves to create new forms and activities.Moreover, Fintech is rapidly gaining traction in both developed and developing countries and has long played an essential role in the financial services industry by bridging the gap between technological and financial components and removing restrictions that traditional payment activities cannot overcome (Esawe, 2022a).SJME 26,3 have been conducted to determine whether perceived enjoyment significantly influences satisfaction in the context of e-wallets.In addition, previous studies have shown that perceived enjoyment significantly influences perceived ease of use and usefulness (To and Trinh, 2021;Winarno et al., 2021) and behavioral intention (Lew et al., 2020).Moreover, in a learning context, few studies (Alotaibi et al., 2019;Chao, 2019;Fagan, 2019) have supported that perceived enjoyment significantly influences performance expectancy and effort expectancy.Therefore, theoretical foundations have been established, but according to our literature review, this influence has not been previously studied in studies related to e-wallets.
Moreover, as an e-wallet needs to complete a specific set or series of tasks using mobile devices, consumers' mobile self-efficacy could be a significant factor to study (Mushi, 2020).In the existing literature (Chao, 2019), users with greater mobile self-efficacy are more likely to feel more positive emotions and are more willing to use technology.According to Dang et al. (2016), the mobile self-efficacy of the users significantly influences perceived enjoyment.Therefore, we can argue that the perceived enjoyment of higher mobile selfefficacy consumers increases if they can perceive the e-wallet as more useful and valuable.Taken together, we can propose that the more the e-wallet system can achieve the tasks effectively, the more likely it is that the users' perceived enjoyment will increase, and therefore, their satisfaction will increase, which will increase their behavioral intention.
However, in a world where technologies are shifting faster, users' perceptions influence whether they adopt them.As a result, the particular result of adoption is more uncertain (Flavian et al., 2020).On the other hand, the existing research provides limited empirical insight into the optimum amount of these types of connections.The current study tackles this gap by attempting to improve knowledge of complicated interactions.The following research questions were developed to achieve the study's objectives: RQ1.What factors influence consumers' behavioral intention and usage of e-wallets?RQ2.Do mobile self-efficacy, perceived enjoyment and satisfaction impact the UTAUT model regarding the e-wallet?
This study adds to the existing literature by trying to identify satisfaction and perceived enjoyment as antecedents of behavioral intention and adoption of e-wallets; extending theoretical comprehension of behavioral intention and adoption through consumers concerning e-wallets; offering empirical evidence of the impact of external factors on effort expectancy and performance expectancy, which leads to usage-related satisfaction and behavioral intention; and providing a reference for e-wallet stakeholders to decide future development directions and approaches related to the implementation of e-wallets.The remainder of this paper is organized as follows.Section 2 reviews the literature and suggests a research framework and the formulation of the hypotheses.Following this is Section 3, which outlines the research methodology.Section 4 presents the data analysis and findings.Section 5 concludes the paper with a discussion of the results.We then discuss the implications for academics and practitioners, the study's limitations and the scope for further research.
2. Literature review and hypotheses development 2.1 Literature review 2.1.1Electronic wallets e-wallets.An e-wallet is defined as "a mobile device-based platform that facilitates cashless payments of a sales transaction-either in proximity or remotely, between consumers and merchants or service providers."(Ramli and Hamzah, 2021) An SJME 26,3 e-wallet is a virtual wallet that allows customers to preload a set amount to their accounts registered with the e-wallet's service providers and spend it online and offline to pay for goods and services (Phuong et al., 2020).Furthermore, e-wallet users can pay for the same receipt independently because they have the option to split expenses (Syifa and Tohang, 2020).
Scholars have paid the most attention to e-wallet adoption for several reasons, such as the e-wallet being one of the most recognizable Fintech inventions (Chawla and Joshi, 2019), with significant growth among users.On the one hand, introducing new Fintech solutions disrupts the traditional cash-based payment model (Abdullah et al., 2020) and capitalizes on the attractive patterns arising: consumers' reliance on smartphones to access the internet.On the other hand, Fintech solutions will consolidate an entirely new business model (Syifa and Tohang, 2020).Further, they will aid in striking the informal sector and accomplishing financial inclusion objectives while fostering economic growth (Esawe, 2022a) and affecting social dimensions, thereby increasing social equity.Also, mobile technology is integrating more unbanked consumers into the financial sector (Esawe, 2022b).
2.1.2Unified theory of acceptance and use of technology (UTAUT).The UTAUT model was created after empirically examining the eight theories.After rigorous testing, the eight models were combined and developed into a new model, UTAUT.The model contributed to a 70% improvement in predicting efficiency in the behavioral intention to use the technology (Venkatesh et al., 2003).According to Ramli and Hamzah (2021), numerous researchers use several theories to develop their study framework related to e-wallets, and UTAUT is among the most often used by previous studies.
The original UTAUT model has four moderating factors.However, to keep this study as concise as possible and in line with prior studies (Gupta et al., 2019), only the primary hypotheses were explored, and the effects of the moderators were not the focus of this study.In the following section, the four exogenous constructs of the UTAUT model (their definitions, root constructs and the source model) will be described in more detail.
2.1.3Self-efficacy theory.According to the self-efficacy theory, people who believe in their capability will perform well, and the best predictors of an individual's behavior are their capability self-appraisals (Bandura, 1977(Bandura, , 1982)).Furthermore, people's self-assessments of their operational capabilities serve as a set of key transcription factors of their emotional reactions.According to Lew et al. (2020), as the self-efficacy theory is concerned with individual beliefs, it may supplement models primarily concerned with technological factors.Therefore, this study used the self-efficacy theory as an underpinning to integrate mobile selfefficacy.
2.1.4Flow theory.As cited in Chen et al. (2018, p. 282), "the main idea of flow is enjoyment."In a flow experience, through modifying the process of interaction, a sense of immersion or telepresence is created, and flow can be formed when there is a balance between skill and challenge levels, while in an imbalance situation, a person may perceive either tedium or anxiety (Csikszentmihalyi, 1975).In this rationale, prior studies have linked mobile self-efficacy and perceived enjoyment (Chao, 2019).Perceived enjoyment is crucial in mobile services (Chen et al., 2018;Lew et al., 2020).Therefore, this study used the flow theory as underpinning to integrate perceived enjoyment.In line with Lew et al. (2020) work, we adopted perceived enjoyment to measure the flow level of e-wallet customers.
2.2 Hypotheses development 2.2.1 Mobile self-efficacy (MSE).MSE is "a degree to which an individual believes that he or she can perform a specific task/job using the mobile."(Mushi, 2020, p. 108).MSE refers to an individual's appraisal of their ability to use his/her skills to carry out a particular task well.

Mobile e-wallet consumers' intentions
E-wallet applications require knowledge in addition to just using smartphones.E-wallet users, for example, will struggle with some of the functionalities of e-wallet if they cannot perform a specific task using the mobile.Furthermore, higher levels of MSE may increase enjoyment and decrease user difficulty.Moreover, MSE has been shown to affect perceived enjoyment and is considered an antecedent to perceived enjoyment.Previous studies have confirmed that MSE directly influences perceived enjoyment (Chao, 2019;Dang et al., 2016).The above discussion demonstrates the importance of determining mobile self-efficacy's role.Therefore, the following hypothesis is proposed: H1.Mobile self-efficacy has a positive influence on perceived enjoyment of e-wallets.
2.2.2 Perceived enjoyment (P-Enj).One of the fundamental reasons driving technology adoption and use is P-Enj, also known as hedonic motivation (Venkatesh et al., 2012).P-Enj is "the degree to which the activity of utilizing a computer is regarded as enjoyable in and of itself" (Davis et al., 1992(Davis et al., , p. 1113)).In the TAM model, P-Enj can be considered one of the most important external factors that determine people's PU (Alalwan et al., 2018;To and Trinh, 2021), PEOU (To and Trinh, 2021;Winarno et al., 2021) and intention to use (Alfany et al., 2019;Sigar, 2016).Furthermore, P-Enj influences satisfaction (Chao, 2019).E-wallet usage can be the leading cause of customer enjoyment because it requires less time to accomplish the transaction (Chen et al., 2018).However, few studies have determined whether P-Enj is a significant external element of the UTAUT model (Alotaibi et al., 2019;Chao, 2019;Fagan, 2019).Moreover, to our knowledge, no prior studies on e-wallets have studied the effects of P-Enj on performance expectancy and effort expectancy.Consequently, the following hypotheses were proposed: H2.Perceived enjoyment has a positive influence on (H2a) performance expectancy; (H2b) effort expectancy; (H2c) consumer satisfaction.

Performance expectancy (PE)
. PE is "The degree to which an individual believes the system helps improve job performance."(Venkatesh et al., 2003, p. 447).PE is how consumers believed that using an e-wallet would provide better convenience, be more effective and beneficial for a transaction and be completed more quickly.Concerning previous academic literature, PE significantly influences satisfaction (Elok et al., 2021;Lee et al., 2021;Syifa and Tohang, 2020).Performance expectancy, with effort expectancy, social influence and facilitating conditions, is one of the antecedents of behavioral intention in the UTAUT model.The positive influence of PE on the adoption intention of e-wallets has been supported by previous studies (Abdullah et al., 2020;Chawla and Joshi, 2019;Widodo et al., 2019;Yang et al., 2021).Consequently, the following hypotheses were proposed: H3.Performance expectancy has a positive influence on (H3a) the consumer's satisfaction; (H3b) the consumer's intention to use an e-wallet.

Effort expectancy (EE)
. EE is "The degree of ease associated with the use of the system."(Venkatesh et al., 2003, p. 450).In which EE is how customers believed that learning the e-wallet would be simple, that they would be skilled at using the e-wallet, and that their interaction and navigation with the e-wallet would be simple and obvious.On the one hand, EE significantly influences satisfaction (Elok et al., 2021;Lee et al., 2021;Syifa and Tohang, 2020).However, Syifa and Tohang's (2020) finding shows an insignificant effect of EE on satisfaction.

SJME 26,3
On the other hand, some studies have regarded EE as an essential construct influencing the user's behavior or intention to adopt an e-wallet (Abdullah et al., 2020;Chawla and Joshi, 2019;Widodo et al., 2019;Yang et al., 2021).Consequently, the following hypotheses were proposed: H4. Effort expectancy has a positive influence on (H4a) the consumer's satisfaction; (H4b) the consumer's intention to use an e-wallet.
2.2.5 Satisfaction.Satisfaction is "the overall opinion and experience that a user feels while using a technological service."(Liébana-Cabanillas et al., 2021, p.139).Previous studies have shown that satisfaction significantly affects the user's intention (Alfany et al., 2019;Lee et al., 2021;Phuong et al., 2020).Consequently, the following hypothesis was proposed: H5. Consumer satisfaction has a positive influence on his/her intention to use an e-wallet.
2.2.6 Social influence.SI is "The degree to which an individual perceives that important other believe he or she should use the new system."(Venkatesh et al., 2003, p. 451).SI is consumers' perceptions of critical people's recommendations and support that will impact their decision to use the e-wallet.Moreover, how the government has encouraged the use of e-wallets.Also, how the current media trend will impact their decision, prior studies concluded that SI influences behavioral intention to use (Yang et al., 2021).Therefore, the following is proposed: H6. Social influence has a positive influence on the consumer's intention to use an ewallet.

Facilitating conditions (FC).
FC is "The degree to which an individual believes that an organizational and technical infrastructure exists to support the use of the system."(Venkatesh et al., 2003, p. 453).FC represents the consumers' perceptions of the e-wallet's compatibility with other technologies they are using and their availability of the resources and knowledge required to utilize it.FC influences usage behavior rather than the user's behavioral intention (Venkatesh et al., 2003).In various studies, FC influences users' behavioral intention (Abdullah et al., 2020;Chawla and Joshi, 2019;Esawe, 2022a;Widodo et al., 2019).Consequently, this study hypothesized that FC could significantly influence consumers' intentions and use behavior of e-wallets.Therefore, the following hypotheses were proposed: H7.Facilitating conditions have a positive influence on (H7a) the consumer's intention to use an e-wallet; (H7b) the usage behavior of the e-wallet.

Behavioral intention (BI) and user behavior (UB)
. BI is "The degree to which a person has formulated conscious plans regarding whether to perform a specified future behavior."(Chai and Dibb, 2014, p. 3).There is a strong relationship between BI and UB (Esawe et al., 2022;Orús et al., 2021).Intentions reflect users' willingness to engage in specific behaviors (Flavi an et al., 2019).According to the UTAUT model, BI positively affects users' UB.Moreover, users' BI has also been a common factor in determining users' UB of an e-wallet (Yang et al., 2021).Consequently, the following hypothesis was proposed: H8. Consumer's behavioral intention has a positive influence on his/her usage behavior of e-wallets.
Figure 1 illustrates the expected correlations between these constructs.

Sample participants and procedures
In its quest for financial inclusion, the Egyptian Government is trying to build a supportive environment for the prosperity of e-wallets by implementing initiatives that encourage or allow the use of e-wallets more efficiently (such as Meeza); this includes creating financial awareness campaigns and reaching out to individuals via social media and educate them about e-wallets.In addition to encouraging consumers to subscribe to the e-wallet without registration or account opening fees.According to Egypt's National Telecommunication Regulatory Authority (NTRA) report, the distribution of e-wallets for mobile phones in Egypt across firms is as follows: Vodafone had 65%, Orange 20%, Etisalat 11% and We 4% (NTRA, 2021).
In Egypt, e-wallet platforms are becoming popular NTRA's report revealed an increase in numerous indicators of e-wallet usage in 2021 compared with 2020.For instance, electronic transactions via wallets increased by 175% and transfers from one e-wallet to another increased by 300% (NTRA, 2021).Even though this is the bright side, however, according to the Oxford business group, the informal sector accounts for almost half of economic activity and cash transactions account for 55% of internet sales, which indicates that the potential of an e-wallet is not always fully realized (OBG, 2021).In addition, with 70% of Egyptians being unbanked (OBG, 2021), Egypt has the highest number of unbanked individuals and has enormous potential for merchandizing with multichannel payment systems (Esawe, 2022a(Esawe, , 2022b)), which combines two facts: the adoption and use of e-wallets are still at an early stage, and the size of the untapped market is postulated to be of more benefit to e-wallet stakeholders.
The sample was drawn from an online survey conducted in Egypt in June 2021.The study was conducted for approximately two weeks.The online survey was constructed using Google Forms and shared on social media platforms such as Facebook and Twitter.Consumers with a mobile device who can link an e-wallet to mobile payment services and use it are the target group of this study.Convenience sampling was used to collect the data, as it is not possible to have a sampling frame for all the e-wallet users.
We gathered 598 responses.Prior to actual cleaning and data screening, 22 responses were eliminated.Thus, the total number of usable responses was 576.Of those who participated, 48.3% were male and 51.7% were female.The most common age group among the respondents was under 20 years (85.4%).Regarding educational attainment, 85.8% did not have a university degree, which is consistent with most of the sample being under 20.A total of 94% respondents owned a smartphone with internet access.A total of 91.0% of participants were aware of what an e-wallet was (Table 1).

Measures
All the statements used to assess the nine constructs of the research framework were adapted from previous studies to maintain content validity.Each item was scored on a fivepoint Likert scale, with one indicating "strongly disagree," and five indicating "strongly agree."The questionnaire was structured into three parts: the first part dealt with a nominal scale to identify respondents' demographic information; the second part collected information about consumers' previous knowledge and usage of e-wallets; the third part included UTAUT constructs, mobile self-efficacy, perceived enjoyment and satisfaction.
Two professional and four academic experts evaluated the questionnaire to ensure that items were appropriate to the context of the e-wallet.Furthermore, the scales were translated using the back-translation method Behr (2017) to ensure that the English and Arabic versions did not contradict one another.In addition, the survey was pilot tested.A total of 60 questionnaires were distributed to assess the reliability and validity of the instrument.Internal consistency reliabilities (based on Cronbach's alphas) ranged from 0.731 to 0.866 (Abbasi et al., 2022).According to Nawi et al. (2020), in the pilot study, reliability results equal to or greater than 0.60 are acceptable.Furthermore, Hair et al. (2019b) and Hair et al. (2019a) assume similar thresholds, as they highlight that the recommended values for Cronbach's alpha are acceptable if they range between 0.70 and 0.90.
Confirmatory factor analysis (CFA) was conducted to identify the factors of the model.All measures of CFA such as the root mean square error of approximation, the normed fit index, the non normed fit index, the comparative fit index and the incremental fit index were appropriate as per the standard norms (Table 2).

Methodology
The partial least squares structural equation modeling (PLS-SEM) technique was employed, and a two-stage approach was adopted (Hair et al., 2019b) to analyze the obtained data using Smart-PLS 2.0, Excel sheets and SPSS V23.

Mobile e-wallet consumers' intentions
Recently PLS-SEM has flourished among social scientists (Hair et al., 2019b), and many scholars have used it in multidiscipline, e.g., hospitality and tourism (Flavi an et al., 2019), educational studies (Esawe et al., 2022).Moreover, it is now being used for quantitative study in Fintech (Abbasi et al., 2022;Esawe, 2022a) due to its ability to test complex models and reflective measurement (Henseler, 2018) and neither normally distributional assumptions nor requires large samples, as PLS-SEM incorporates both explanation and prediction (Hair et al., 2017).PLS-SEM is a two-stage approach: in the first stage, the outer model (measurement model) measures the reliability and validity, whereas, in the second stage, the inner model (structural model) tests the strength of the relationship between the constructs (Hair et al., 2019b(Hair et al., , 2019a)).
This study used Podsakoff et al. (2003) guidelines to reduce the risk of common method variance.The participants were informed that their responses would be treated confidentially and anonymously; there were no right or wrong answers, only positive or negative perceptions.In addition, the questionnaire was designed keeping in mind that IV and DV are furthest apart and in reverse order.Furthermore, the variance inflation factor (VIF) was used to address the potential common method bias.Table 5 shows that VIF values ranged between 1 to 2.4.according to Kock (2015), VIF values !1-5 indicate possible collinearity issues, but these are rarely important enough to warrant attention.

Measurement model
Hair et al. (2019b) stated a four-step assessment of measurement models in the first stage examining the indicator loadings.Second, internal consistency reliability was assessed by computing Cronbach's alpha and composite reliability (CR) (Fornell and Larcker, 1981).An acceptable level for factor loading, Cronbach's alpha and CR should be 0.70 or higher.Third, we computed the average variance extracted (AVE) for all items to verify the convergent validity of each construct.The acceptable levels of AVE should exceed 0.50.According to Table 3, the results of outer loadings, Cronbach's a and CR values have reached acceptable levels, internal consistency is established, and the scale has adequate construct validity.Additionally, the AVE for all latent variables exceeded good convergent validity values.
Fourth, discriminant validity was assessed by computing the correlations' heterotraitmonotrait (HTMT) ratio.Table 4 shows that the HTMT results indicate that all constructs' values were less than 0.90.Therefore, they are defined by their respective constructs without any overlap.Except, SI is highly correlated with UB.Also, perceived enjoyment and mobile self-efficacy have a high correlation among them.Although these constructs are conceptually distinct, it may be difficult to distinguish them empirically.As a result, and in line with Henseler et al. (2015) suggestion, a more liberal criterion for a model such as UTAUT is warranted.According to Franke (2019), HTMT confidence interval values should not include 1.Therefore, all the values are acceptable as they differ from 1.

The structural model
In the second stage, the structural model assessment procedures include evaluating collinearity issues, the statistical significance and relevance of the path coefficient, the coefficient of determination R 2 , effect size f 2 and the blindfolding-based cross-validated redundancy measure Q 2 .First, multicollinearity issues were assessed using a VIF.VIF values !3-5 indicate possible collinearity issues, but they are rarely significant enough to warrant attention (Hair et al., 2019a(Hair et al., , 2019b)).The VIF results in Table 5 indicate that collinearity was not an issue.
Second, examining R 2 , Hair et al. (2019aHair et al. ( , 2019b) ) highlight that R 2 values of 0.75, 0.50 and 0.25 are considered substantial, moderate and weak, respectively.R 2 values of 0.90 and higher are typically indicative of overfitting.Figure 2 illustrates that The R 2 value for UB (0.373) is considered moderate, while the R 2 value for UB (0.730) is considered substantial to measure the variance.Note: The numbers in parentheses represent (HTMT) confidence interval values Table 3.

SJME 26,3
Third, the Q 2 values were obtained using the blindfolding technique in Smart-PLS.As a rule of thumb, all values should be more significant than zero.Moreover, Hair et al. (2019aHair et al. ( , 2019b) ) suggest that "the Q 2 values higher than 0, 0.25, and 0.50 depict small, medium, and large predictive relevance of the PLS-path model".Finally, the bootstrap resampling approach with 5,000 resamples was employed to establish direct pathways' relevance and estimate standard errors (Ringle et al., 2005).Table 5 shows that all Q 2 values have significant predictive relevance, as they are greater than zero, confirming the model's out-of-sample predictive relevance.Moreover, all Q 2 values higher than 0.50 depict the considerable predictive accuracy of the PLS path model.
Fourth, we examined how removing a specific predictor construct influences the R 2 value of an endogenous construct to compute f 2 .Hair et al. (2019aHair et al. ( , 2019b) cite Cohen's rule of thumb, f 2 values greater than 0.02, 0.15 and 0.35, representing small, medium and substantial effects, respectively.Based on the f 2 effect sizes in Table 5, mobile self-efficacy was the most critical factor and substantially affected perceived enjoyment, followed by the paths between perceived enjoyment and performance expectancy.Moreover, BI for UB and FC for UB had a medium effect size.Furthermore, neither effort expectancy in UB, performance expectancy in satisfaction, nor FC in BI had an effect size.
Figure 2 reveals the hypotheses testing results, which shows that all hypotheses were supported except H3a, H5b and H8a.These findings can be observed from the t-test value (t-values !1.96; p < 0.001).From Figure 2, it can also be seen that the paths with the highest Standardized Regression Weight are the hypothesized paths between mobile self-efficacy

Not supported Supported
Mobile e-wallet consumers' intentions and perceived enjoyment (0.671), followed by the paths between perceived enjoyment and performance expectancy (0.669).
4.3 Importance-performance map analysis (IPMA) Following Hair et al. (2019aHair et al. ( , 2019b) ) suggestion, this study employed importanceperformance map analytics (IPMA) by using BI and UB as the target variables to provide a more in-depth dimension to the analysis and identify the potential improvement areas that need more attention.In other words, applying the IPMA approach allows for a better understanding of the PLS-SEM results.Instead of just assessing the importance measure, as shown in the path coefficients, IPMA considers the average value of the constructs and their items (i.e.performance measure) (Ringle, 2016).

Importance-performance map analysis for user behavior
The results of the IPMA analysis are presented in Table 6 and depicted in the graph in Figure 3.The results indicated that behavioral intention was relevant in predicting UB, with 28.53 importance and a 56.41 performance level.Then, facilitating condition, with 16.22 importance and a 55.72 performance level, is shown in Figure 3.

Importance-performance map analysis for behavioral intention
The results of the IPMA analysis are presented in Table 6 and depicted in the graph in Figure 3.The importance and performance of all the independent variables (i.e.PE, EE, FC, SI, MSE and P-Enj) were measured.The results indicate that although effort expectancy had the most outstanding index values (performance), it was not vital in predicting behavioral intention in the model, with a total effect (importance) of 9.61.Social influence was relevant in predicting behavioral intention, with a total effect (importance) of 56.85, as denoted in the IPMA map in Figure 3.

Discussion and conclusions
Overall, this study adds to the corpus of knowledge by extending UTAUT to include new relationships and improving the understanding of the factors influencing behavioral intention and e-wallet adoption among Egyptian consumers.Mobile self-efficacy is critical in adopting e-wallets (Sigar, 2016).That is to say, without mobile self-efficacy, it would be impossible to implement innovative technologies such as ewallets.Therefore, last year's research primarily focused on studying the influence of mobile self-efficacy on attitude, actual e-wallet use or BI to adopt an e-wallet.However, it is worth noting that a few scholarly attempts have already been made to evaluate the direct influence of mobile self-efficacy on perceived enjoyment.As a result, the current study studied this direct influence.The empirical evidence provided above suggests that mobile self-efficacy significantly influenced perceived enjoyment (t-values of 11.44; p-value < 0.001), which fits with findings reported by Chao (2019) and Dang et al. (2016).Consequently, e-wallet service providers should always aim to produce e-wallet applications capable of performing tasks in the shortest period to maintain consumers' positive perceived enjoyment.
According to previous studies (Alalwan et al., 2018;To and Trinh, 2021;Winarno et al., 2021), perceived enjoyment is a critical construct that significantly influences the e-wall PU and PEOU.The findings revealed that perceived enjoyment had a significant influence on satisfaction (t-values of 2.208; p-value < 0.05) this results supported by Chao (2019) and Dang et al. (2016), effort expectancy (t-values of 2.885; p-value < 0.05) and PE (t-values of 13.76; p-value < 0.001) this results supported by Alotaibi et al. (2019), Chao (2019) and Fagan (2019).As a result, perceived enjoyment can be considered a critical external construct in the UTAUT model, implying that customer enjoyment of e-wallets is predicted to grow as they become a more prevalent form of transaction.Customers not only find e-wallets easy to use but also allow customers to complete transactions more rapidly.
According to our previous results, satisfaction significantly influences consumers' behavioral intention toward using e-wallets (t-values of 2.698; p-value < 0.05).Several previous studies have concluded that if an e-wallet meets a consumer's needs, they are more likely to use it again (Alfany et al., 2019;Lee et al., 2021;Phuong et al., 2020).The findings of this study also confirm customers' agreement that they are content with using e-wallets and will recommend them to their colleagues.
Moreover, the findings show that effort expectancy significantly affects satisfaction with an e-wallet (t-values of 3.978; p-value < 0.001).This result is supported by the study results of Elok et al. (2021) and Lee et al. (2021) and, in contrast, with the results of Syifa and Tohang (2020).Contrary to the general literature, the empirical results in Table 6 show that effort

Mobile e-wallet consumers' intentions
expectancy had an insignificant influence on consumers' behavioral intention toward using e-wallets (t-values of 0.889; p-value > 0.05).This result is inconsistent with Abdullah et al. (2020) and Widodo et al. (2019).The insignificant influence of effort expectancy on consumers' behavioral intention in this study was presumed to be because most consumers become more accustomed to and knowledgeable about e-wallet usage because they had already been exposed to e-wallets.Furthermore, most respondents were under 20, a more flexible and adaptive age for FinTech.Consequently, consumers do not consider whether using e-wallets is complex or straightforward.This result is supported by previous studies such as Raihan and Rachmawati (2019) and Syifa and Tohang (2020).Nonetheless, we recommend that e-wallet providers strive to create user-friendly and straightforward applications to maintain positive behavioral intention among consumers.This study's results suggest that performance expectancy had an insignificant impact on customer satisfaction with the e-wallet (t-values of 0.341; p-value < 0.05).This result contrasts with Elok et al. (2021), Lee et al. (2021) and Syifa and Tohang (2020).This not significant could be attributed to the fact that performance expectancy refers to the cognitive beliefs of inexperienced consumers, whereas satisfaction is based on consumers' first-hand experience with e-wallets.According to this study's findings, e-wallets are rapidly becoming an extremely effective method for customer transactions.In addition, customer satisfaction with e-wallets and their behavioral intention toward utilizing them will increase if they are educated on using the e-wallet effectively.As a result, in terms of future e-wallet usage growth, service providers are advised to establish customer communities through online forums to discuss and share their experiences.This measure could boost e-wallet diversification while increasing customer satisfaction and BIs' willingness to use e-wallets.
Furthermore, performance expectancy was found to significantly influence consumers' behavioral intention toward using e-wallets (t-values of 2.784; p-value < 0.05) because the use of e-wallets saves time and makes transactions more effective.Therefore, the current study is consistent with the findings of Abdullah et al. (2020), Syifa and Tohang (2020) and Widodo et al. (2019).As a result, e-wallet providers should create applications that enhance consumer performance in completing their transactions.
Moreover, the results indicate that social influence significantly influences consumers' behavioral intention toward using e-wallets (t-values of 4.196; p-value < 0.001).In other words, one of the factors driving behavioral intention to use e-wallets is the social influence of important people, such as family, relatives and friends.This result is supported by previous studies such as Abdullah et al. (2020) and Raihan and Rachmawati (2019).
In addition, facilitating conditions was found to have an insignificant influence on consumers' behavioral intention toward using e-wallets (t-values of 0.862; p-value < 0.05) and significantly influenced consumers' UB of e-wallets (t-values of 4.754; p-value < 0.001).These results align with the UTAUT model, whereas facilitating conditions are considered a direct determinant of "UB" construction rather than the "behavioral intention."(Venkatesh et al., 2003).This also corresponds to other studies, such as Esawe (2022aEsawe ( , 2022b) ) and Raihan and Rachmawati (2019) and, in contrast with the findings of Abdullah et al. (2020), Chawla and Joshi (2019) and Widodo et al. (2019).This contradiction implies that future research should focus on the impact of facilitating conditions on BI.
According to the results of the IPMA analysis for UB, service providers should focus on BI and facilitating conditions, as they performed poorly compared to effort expectancy and satisfaction but are the most crucial factors to UB.As a result, improvements are required to transform behavioral intention and facilitate conditions into a better state of UB for customers.In addition, IPMA analysis Results for BI suggested that social influence needs more consideration and improvement from service providers.Although it is the most SJME 26,3 fundamental factor in determining the customers' behavioral intentions, social influence performs low compared with other less crucial factors such as effort expectancy, satisfaction and facilitating conditions.

Theoretical implications
Theoretically, this study has used multiple conceptualizing lenses by integrating three theories: self-efficacy theory, flow theory and UTAUT, to help explain how these constructs in the framework influence consumers' behavioral intention and usage of e-wallets.Although the integration of previously mentioned theories has been highlighted in the literature (Chen et al., 2018;Lew et al., 2020), this is one of the first studies in the context of ewallets to integrate these theories to investigate the relations between the structures in this order, according to the researcher knowledge.Furthermore, this study found that performance expectancy, social influence, mobile self-efficacy, perceived enjoyment and satisfaction significantly influence consumers' behavioral intentions.Moreover, facilitating conditions and behavioral intention significantly positively influence the behavioral usage of e-wallet services.In addition, the findings of this study added to the literature regarding the relationships between effort expectancy, performance expectancy, mobile self-efficacy, perceived enjoyment and satisfaction to explain the potential of consumers' behavioral intention and usage of e-wallets.

Managerial implications
In terms of practical implications, this is the first study to report the factors influencing e-wallet adoption in Egypt.The findings of this study should improve decision-makers understanding of these factors' roles and could help them design successful tactics to encourage customers' behavioral intention and UB of e-wallets.Moreover, the study findings broaden e-wallet providers' understanding by incorporating these factors into upgrading their services to perform ubiquitously as consumers desire.Moreover, as developing countries shift towards financial inclusion, digital transformation will impact the future, requiring businesses and consumers to be digitally aware and knowledgeable because of the inevitability of increasing noncash transactions to purchase products and services.Esawe et al. (2018) assert that developing countries, particularly Egypt, would not progress unless massive innovation occurs, frequently resulting in a paradigm shift.This innovation may begin at the individual level; however, by coordinating efforts and focusing on continual development, a transformation that is not just local but also global is achievable.

Limitations and suggestions for future research
This study has some limitations.First, we did not include the four moderating factors in the original UTAUT in our model.future research could investigate the moderating effect of those four factors on consumer behavior.Second, it cannot confirm long-term causal relationships among factors because it is a cross-sectional study.Customers' perceptions of constructs might shift over time as they gain new knowledge and experiences.Therefore, future research could use a longitudinal strategy to obtain more precise results from a specific cohort.Third, because the study was conducted in the Egypt e-wallet market, the study's generalizability is limited.However, future research could replicate our framework with a larger and more geographically diverse sample.Fourth, the most common age group among the respondents was under 20 years (85.4%).Because this age group adopts technology more than other age groups, future studies should look at other age groups and the average age.Fifth, this study focuses on customers' perceptions, and future research Mobile e-wallet consumers' intentions could investigate retailers' perceptions.Sixth, future studies may be conducted comparing consumers and retailers.Seventh, this study used a quantitative method using questionnaires to conduct the research.It is helpful for future studies to be conducted from multiple perspectives, such as using qualitative or mixed methods.Finally, moderating and mediating variables can be added to the model to evaluate the mechanisms related to the present situation.
Figure 1.Conceptualized extended UTAUT model Venkatesh et al. (2003) and Venkatesh et al. (2012) a = 0.760; CR = 0.845; AVE = 0.578 Assuming I had access to the e-wallet, I intend to use it 0.762 Given that I had access to the e-wallet, I predict that I would use it 0.774 I plan to use the e-wallet in the future 0.788 I recommend e-wallets to my colleagues 0.715 Use Behavior (UB) Adapted from Venkatesh et al. (2012) a = 0.706; CR = 0.836; AVE = 0.631 UB 1-I use an e-wallet frequently 0.831 UB 2-I use many functions of e-wallets 0.711 UB 3-I depend on e-wallets 0.834Note: Frequency ranged from "never" to "many times per day" Figure 2. Results of path analysis Figure 3. IPMA for BI and UB