Investigating Acceptance of Digital Asset and Crypto Investment Applications Based on the Use of Technology Model (UTAUT2)

Abstract

In recent years, cryptocurrency has increased in popularity in Indonesia. In Indonesia, based on data from the Ministry of Trade (Kemendag), until the end of May 2021, the number of investors in cryptocurrency assets or crypto money was 6.5 million people. This number has increased by more than 50 percent when compared to 2020 when there were 4 million people. The Pintu application is the first crypto mobile application in Indonesia that is committed to solving crypto investment problems, especially for beginners and ordinary people. Even though it provides benefits, investing in cryptocurrency can provide high profits. In an instant, it can also make a profit. The motion, which is like a roller coaster, requires strong mental readiness to invest in cryptocurrencies. This should also be a critical consideration for investors, especially young investors. Therefore, it is necessary to understand what factors contribute to building stronger attitudes and behavioral intentions toward the PINTU application. This research analyzes the data using the use of technology 2 method with the partial least square (PLS) analysis technique method, which will later be processed in the form of data results in the form of responses of the user when using the application. Facilitating conditions and social influence are the most influential indicators. The results of the study show that behavioral intention to adopt has a relationship with behavioral intention to recommend, and behavioral intention to adopt positively and significantly influences the intention to recommend.

1. Introduction

Currently, mobile applications and websites are gaining importance [1] and mobile phones have become the most important instruments for communication and relationship formation in the modern world [2]. Smartphones, tablets, e-book readers, handheld gaming devices, and portable music players are virtually ubiquitous in today’s society [3]. The evolution of information and communication technologies, particularly mobile phones, has altered how people interact. These tools enhance people’s access to, acquisition of, and communication of information while facilitating the formation of new communication networks [4]. These recent technological advancements enable faster access to a growing volume and diversity of data [5]. Mobile phones enable portable access to knowledge across borders, disciplines, and organizations. Globally, 2.4 billion people used digital banking in 2020, and this number is expected to increase to 3.6 billion within the next four years [6]. This accelerating growth in global digital finance has been fueled by the growth of mobile phone penetration and the remarkable development of mobile internet, such as 4G and 5G connections [7].

During the past decade, there has been an exponential increase in the use of digital payments and technological advancements associated with banking operations, which have made users’ lives simpler. This growth has been fueled by omnichannel merchandising and cyberspace transactions, including the increasing prevalence of mobile payments (m-payments), which eMarketer predicts will have up to three trillion users by 2024 and a value of $1.31 billion by 2031 [8]. Various factors influence the adoption of technologies such as mobile payments [9]. In recent years, the use of mobile payments has increased significantly, particularly in emerging markets [10]. This growth can be ascribed to performance and function benefits for users [11]. M-payments are available through both single-party applications owned by banks and third-party applications held by licensed digital wallet providers [12]. M-payment applications require users to integrate their information with the app to provide authorized services. This integration has also raised concerns regarding the trust and privacy features of the m-payment system, which enable behavioral penetration among prospective users.

In recent years, cryptocurrency has increased in popularity in Indonesia. The same thing happened in the international market. In Indonesia, based on data from the Ministry of Trade (Kemendag), until the end of May 2021, the number of investors in cryptocurrency assets or crypto money was 6.5 million people. This number has increased by more than 50 percent when compared to 2020, which was four million people [13]. Cryptocurrency is a digital asset designed to work as a medium of exchange that uses strong cryptography to secure financial transactions, control the creation of additional units, and verify asset transfers. In Indonesia, cryptocurrency regulations are issued by the Commodity Futures Trading Supervisory Agency (Bappebti) of the Ministry of Trade. Pintu (application) is a mobile-based application for buying, selling, storing, and sending cryptocurrencies in Indonesia to help millennials and retail investors easily invest in various crypto assets such as bitcoin, ETH, and other assets. The Pintu application is the first crypto mobile application in Indonesia that is committed to solving crypto investment problems, especially for beginners and ordinary people. PINTU is a means to buy, sell, store, and send cryptocurrency. Everything can be done through just one smartphone application [14]. According to [15], currently, 13 registered companies are selling crypto assets, of which PT Pintu Kemana Saja (PINTU) is ranked fifth in the list of crypto asset companies. In the current pandemic situation, lots of people are starting to take advantage of blockchain technology. Judging from [16], people’s awareness of crypto assets is getting higher. As evidenced by a survey conducted from early 2020 to April 2021, 69% of respondents read news about crypto assets and blockchain daily, followed by rapid media developments that make the information very accessible. Despite its benefits, investing in cryptocurrency can bring high profits instantly. The motion, which is similar to a roller coaster, requires strong mental readiness to invest in cryptocurrencies. This should also be a critical consideration for investors, especially young investors. Therefore, it is necessary to understand what factors contribute to building stronger attitudes and behavioral intentions toward the PINTU application.

Venkatesh [17] created the unified theory of acceptance and use of technology (UTAUT) in the organizational context, with an emphasis on the utilitarian value (extrinsic motivation) of organizational users. The proliferation of consumer technologies necessitated the extension of the UTAUT model to the consumer context, with an emphasis on the hedonistic value (intrinsic motivation) of technology users. This resulted in the addition of three new constructs, including hedonic motivation, price value, and habit, to the original UTAUT; the new, expanded version is commonly known as UTAUT2. However, in UTAUT2, the voluntariness of use was eliminated as a moderator because consumers are not mandated by their organizations and consumer behavior is often voluntary [18]. The predictive ability of UTAUT2 is significantly greater than that of UTAUT, accounting for 74% of the variance in consumers’ behavioral intentions and 52% of the variance in consumers’ technology utilization of the focal technology [19]. UTAUT2 theory articulates explicitly the inside boundary conditions of a class of things, extending individual technology acceptance and use to consumers from the organizational user context of UTAUT. To empirically validate the UTAUT2 model, 1512 “Mobile Internet” consumers, a specific form of technology users, were examined. In addition to class, the UTAUT2 precisely defined twelve internal attributes, including nine constructs: performance expectancy (PE), effort expectancy (EE), social influence (SI), facilitating conditions (FC), hedonic motivation (HM), price value (PV), habit (HA), behavioral intention (BI), and use behavior (UB) with measurement items; and three moderators, including age, gender, and experience [20]. From the discussion above, gaps can be identified. The following research questions are posed: Which of the UTAUT2 factors contributes to building stronger attitudes and behavioral intentions toward the PINTU application? By investigating these questions, this research contributes to literature and theory in multiple ways. (1) By investigating the precursors of the behavioral intention to use the PINTU application. (2) In employing the UTAUT2 framework, it provides a more holistic framework for digital finances with the inclusion of attitude. (3) The study makes a meaningful contribution to the literature associated with Indonesians’ intention to use m-payment in cryptocurrency.

2. Literature Review

The unified theory of acceptance and use of technology (UTAUT) was developed by Venkatesh [17] after a review of eight prominent theories on technology adoption. It is one of the most prominent and extensively used technological acceptance theories. The initial version of the model focused on the organizational perspective; later, Venkatesh et al. [18] developed the most recent version of the model with the individual customer’s perspective in mind. The UTAUT2 model comprises seven independent variables to measure the behavioral intention of customers to adopt new technologies: performance expectation, effort expectation, social influence, facilitating conditions, price, habit, and hedonic motivation. Mansyur and Ali [21] used the UTAUT2 model to determine the adoption of Shariah-compliant FinTech among Indonesian millennials. In addition, Mohd Thas Thaker et al. [22] used the UTAUT2 model to determine the factors influencing Malaysian consumers’ adoption of internet banking. Moreover, many studies have employed the UTAUT2 model to assess the adoption intention or acceptability of technology [23,24]. Ref. [25] suggested using the UTAUT2 model as the foundational paradigm for technology adoption studies. Thus, UTAUT2 served as the foundational paradigm for this investigation. Even though the hedonic motivation was included in the UTAUT2 model as a predictor of behavioral intention [26], the model did not adequately demonstrate the factors that contribute to enjoyment. In addition, the concept of “habit” cannot be used to evaluate recent innovations in the technological marketplace [27]. In addition, some studies do not employ the moderating variables of UTAUT2 in their analysis of FinTech adoption [28,29].

The unified theory of acceptance and use of technology, or UTAUT, is used to identify the motivation for using technology, as developed by [17]. According to [30], the UTAUT theory was developed through a comprehensive synthesis and integration of the theory of reasoned action (TRA), the technology acceptance model (TAM), the motivational model (MM), the theory of planned behavior (TPB), a combination of TAM and TPB (C-TAM TPB), the model of PC utilization (MPCU), innovation diffusion theory (IDT), and social cognitive theory (SCT). The most critical part of the UTAUT model is the relationship between usage intention and two independent constructs: performance expectancy and effort expectancy. The UTAUT model has two more constructs than TAM: facilitating conditions (environmental factors that make actions easy) and social influence (the extent to which a person perceives that significant others believe that he or she should adopt a method or system) [31].

Various independent variables, such as perceived security, perceived trust, perceived risk, etc., have been used in technology adoption studies to measure behavioral intention. However, perceived credibility can be regarded as one of the most important variables for identifying the effect on consumers’ intention to adopt new technologies. Perceived credibility refers to the belief that a business associate is dependable and possesses the necessary skills to complete transactions [32]. According to a previous study conducted by [33], customers deny the adoption of new technologies due to a perceived lack of credibility. According to [34], integrating perceived credibility into the UTAUT enhances the ability to predict the behavioral intentions of consumers.

According to [35], UTAUT summarizes eight previous theories and models that include TAM and proposes that the adoption of new technology is determined by performance expectancy (PE), which is similar to relative advantage; effort expectancy (EE), which is similar to PEOU; social influence (SI); and facilitation conditions (FC), which are expected to influence behavioral intention (BI) and usage behavior (UB). This model describes the acceptance of a technology based on a better use side with an improvement percentage from 56% to 74% for acceptance in the form of user behavior intention and an improvement in acceptance in the form of user behavior with a percentage from 40% to 52% [18]. The unified theory of acceptance and use of technology 2 (UTAUT2) is appropriate for use because this theory or model is a model of acceptance of the latest technology, which is a unification, synthesis, or summary of eight pre-existing theories or models of technology acceptance. Unlike UTAUT1, whose context is organizational, UTAUT2 can explain technology acceptance, whose context is consumer use. Figure 1 shows the UTAUT2 Model.

2.1. Performance Expectancy

It is defined as the extent to which the use of technology will provide benefits for users in carrying out certain activities [18]. Performance expectancy is the confidence level of an individual in using and believing in technology because it can help their work. Research by [17] shows that performance expectancy has a positive and significant effect on the use of a system. Research by [36] also proves a significant effect of the construct of performance expectancy on mobile payments. Therefore, to support the consistency of the previous research hypothesis, the formulation of Hypothesis 1 is stated as follows:

H1. 

Performance expectancy has a positive effect on the behavioral intentions of PINTU users.

2.2. Effort Expectancy

Venkatesh [17] said that the ease of using information technology creates a feeling in a person that the system has benefits and therefore makes working with it comfortable. The effort expectancy in this study is related to system utilization that can facilitate one’s work and influence the user’s behavioral intention. Therefore, to support the consistency of the previous research hypothesis, the formulation of Hypothesis 2 is stated as follows:

H2. 

Effort expectancy has a positive effect on the behavioral intentions of PINTU users.

2.3. Social Influence

Social influence is the extent to which users perceive that significant people (e.g., family and friends) believe they should use certain technologies [18]. According to Venkatesh [17], social influence concludes that the construct of social influence is a strong predictor that influences individual decisions based on the interests of users of technology systems. Research by [35] shows that social influence plays a role in influencing the interests of mobile-payment users. Therefore, to support the consistency of the previous research hypothesis, the formulation of Hypothesis 3 is stated as follows:

H3. 

Social Influence has a positive effect on the behavioral intentions of PINTU users.

2.4. Facilitating Conditions

Facilitating conditions refer to the user’s perception of the resources and support available for using the technology [18]. According to Venkatesh [17], supporting conditions or facilitating conditions are the extents to which an individual believes that organizational and technical infrastructure can support system or application users. In contrast, this is not in the research on mobile-payment user interest conducted by [35]. Based on the differences in some of these studies, this study tries to re-examine the construct by formulating a hypothesis:

H4: 

Facilitating conditions have a positive effect on the behavioral intentions of PINTU users.

2.5. Hedonic Motivation

According to Venkatesh [17], hedonic motivation is pleasure motivation obtained from users of a system or technology. The concept of hedonic motivation consists of several intrinsic elements, such as pleasure, excitement, and entertainment [17]. Hedonic motivation has been considered a necessary predictor of interest among technology users [18]. Research by [35] provides strong evidence supporting the role of hedonic motivation in shaping individual decisions to adopt the technology. Therefore, to support the consistency of the previous research hypothesis, the formulation of Hypothesis 5 is stated as follows:

H5. 

Hedonic motivation has a positive effect on the behavioral intentions of PINTU users.

2.6. Price Value

According to [18], price value can be used as a predictor of behavioral intention variables in using technology. The price value is defined as the level of consumer awareness of the trade-off between the perceived benefits of using technology and the costs [18]. The use of technology will be of high value when the price is higher while the monetary costs are low [35]. Therefore, to support the consistency of the previous research hypothesis, the formulation of Hypothesis 6 is stated as follows:

H6. 

Price value has a positive effect on the behavioral intentions of PINTU users.

2.7. Habit

According to [18], experience is operated at three levels based on the passage of time: post-training is when the system is initially available to use, 1 month later, and 3 months later. Habits have been defined as the extent to which people tend to perform behaviors automatically due to learning [37]. Ref. [38] also noted that feedback from previous experience will influence various beliefs and future habitual performance. In this context, habit is a perceptual construct reflecting the results of experience. Therefore, it is stated with the following hypothesis:

H7. 

Habit value has a positive effect on the behavioral intentions to use PINTU.

2.8. Behavioral Intention of the User

Users with a higher intention to adopt new technology are more likely to become adopters and recommend the technology to others [39]. Social networks bring several challenges and opportunities to companies as they represent a means of communication that allows users to express their opinions and experiences of mobile-commerce services, products, and technology [35]. Therefore, to support the consistency of the previous research hypotheses, the formulation of Hypothesis 8 is stated as follows:

H8. 

The behavioral intention of PINTU users has a positive effect on the behavioral intention to recommend PINTU to others.

3. Research Model

A unified theory of technology acceptance and utilization of UTAUT was devised in 2003 by Venkatesh [17] to predict the adoption of information technology by business users. UTAUT incorporated the following eight previous relevant theories: IDT [40]; TRA [41]; TPB [42]; SCT [43]; TAM [44]; MPCU [45]; MM [46]; and C-TAM model combined with TPB [47]. For a systematic approach, a model design previewing research models from the UTAUT acceptance model defines four main constructs: performance expectation, effort expectation, social influence, and facilitating conditions, which are the factors that determine technology adoption. The user’s behavior depends on his/her intention and technology usage, and it influences all four factors mentioned: PE, EE, SI, and FC. The UTAUT model takes into account variables from categories across personal identity variables (gender, age, experience, and voluntary use) to moderate the influence of the four constructs in addition to behavioral intention and technology use (Venkatesh [17]). Since UTAUT arises in a context generic to organizations, Venkatesh et al. [18] devised UTAUT2 to include three new constructs: hedonic motivation, price/value, and habit, factors oriented toward the acceptance of technology within an evolved framework to key in consumers behavior (Figure 2).

Figure 3 shows the conceptual model that is used in this research. From the conceptual model above, eight hypotheses were obtained, as shown in

Table 1. Hypotheses.

Hypotheses	Descriptions of Relationship
H1	Performance expectancy has a positive effect on the behavioral intentions of PINTU users.
H2	Effort expectancy has a positive effect on the behavioral intentions of PINTU users.
H3	Social influence has a positive effect on the behavioral intentions of PINTU users.
H4	Facilitating conditions has a positive effect on the behavioral intentions of PINTU users.
H5	Hedonic motivation has a positive effect on the behavioral intentions of PINTU users.
H6	Price value has a positive effect on the behavioral intentions of PINTU users.
H7	Habit values have a positive influence on behavioral intentions to use PINTU.
H8	The behavioral intention of PINTU users has a positive effect on their behavioral intention to recommend PINTU to others.

.

Operational Variable

Operational variables define measurable concepts by determining the dimensions and characteristics of the idea [48]. Measurement of research variables can be done by identifying operational variables by considering the processes in a variable [49]. The author determines operational variables by identifying them through a study of the journal literature. The operating variables used in this study are described in

Table 2. Operational Variable.

No.	Indicators of Statement	Variables
1	The mobile internet is beneficial in my daily life.	Performance expectancy
2	I think using the mobile internet helps me complete tasks faster.
3	I think using the mobile internet will increase my productivity.
4	I think using the mobile internet increases the chances of getting something significant.
5	My interaction with the mobile internet will be clear and understandable.	Effort Expectancy
6	It is easy for me to become skilled at using the mobile internet.
7	I find mobile internet easy to use.
8	I think learning to operate mobile internet will be easy for me.
9	People who are important to me think that I should use the mobile internet.	Social Influence
10	People who influence my habits think that I should use the mobile internet.
11	People whose opinions I value recommend that I use the mobile internet.
12	I have the necessary resources to use the mobile internet.	Facilitating Condition
13	I have the necessary knowledge to use mobile internet.
14	The mobile internet is compatible with other systems I use.
15	I got help when I had trouble using the mobile internet.
16	Using the mobile internet is fun.	Hedonic motivation
17	Using the mobile internet is convenient.
18	Using the mobile internet is very entertaining.
19	Mobile internet has a reasonable price.	Price Value
20	Mobile internet is affordable.
21	At current prices, mobile internet provides good value.
22	Using mobile internet has become my habit.	Habit
23	I am addicted to using the mobile internet.
24	I have to use mobile internet.
25	Using mobile internet has become commonplace for me.
26	In the future, I will use mobile internet intensely.	Behavioral Intention to Adopt
27	I am trying to use mobile internet constantly in my daily life.
28	I plan to use the mobile internet frequently.
29	I will recommend to my friends that they use mobile internet services if they are available.	Behavioral Intention to Recommend
30	If I have a good experience using the mobile internet service, I will recommend it to my friends.

.

4. Result and Discussion

4.1. Pilot Test

Before distributing the questionnaires, it is necessary to carry out a pilot test to determine the validity and reliability of the questionnaires and the level of understanding of the respondents. Respondents in this study are users of the PINTU application who are anonymous to protect the confidentiality of their data. According to [50], experimental and comparative research requires a sample of 15 to 30 respondents in each group. The number of respondents used in the pilot test of this research was at least 30, distributed through Google Forms media with the snowball sampling technique. At this initial stage, 30 data points were obtained. The data processed at this stage corresponds to the number of respondents determined in the previous chapter, namely 30 samples. The highest percentage of respondents’ usage in this early-stage survey was 3–6 months, with a gain of 33%.

Table 3. Validity of Pilot Test.

Variables	Question Items (Indicators)	Pearson Correlation	Notes
Performance Expectancy (PE)	PE1	0.771	Valid
	PE2	0.752	Valid
	PE3	0.843	Valid
	PE4	0.730	Valid
Effort Expectancy (EE)	EE1	0.792	Valid
	EE2	0.752	Valid
	EE3	0.858	Valid
	EE4	0.752	Valid
Social Influence (SI)	SI1	0.956	Valid
	SI2	0.982	Valid
	SI3	0.910	Valid
Facilitating Conditions (FC)	FC1	0.704	Valid
	FC2	0.695	Valid
	FC3	0.809	Valid
	FC4	0.897	Valid
Hedonic Motivation (HM)	HM1	0.881	Valid
	HM2	0.896	Valid
	HM3	0.842	Valid
Price Value (PV)	PV1	0.833	Valid
	PV2	0.658	Valid
	PV3	0.826	Valid
Habit (H)	H1	0.889	Valid
	H2	0.930	Valid
	H3	0.817	Valid
	H4	0.968	Valid
Behavioral Intention to Adopt (BIA)	BIA1	0.909	Valid
	BIA2	0.936	Valid
	BIA3	0.712	Valid
Behavioral Intention to Recommend (BIR)	BIR1	0.754	Valid
	BIR2	0.914	Valid

shows the validity of the pilot test.

Reliability tests can be performed to measure the stability and consistency of the respondents’ responses to questions related to constructs between variables. All variables will be known for their accuracy in measuring what the researcher measures, which will determine whether the questionnaire will be used as a research tool at a later stage. The following is a summary of the reliability test results for all variables in this study.

Table 4. Reliability of Pilot Test.

Variables	Cronbach’s Alpha	Notes
Performance Expectancy (PE)	0.735	Reliable
Effort Expectancy (EE)	0.807	Reliable
Social Influence (SI)	0.946	Reliable
Facilitating Conditions (FC)	0.780	Reliable
Hedonic Motivation (HM)	0.850	Reliable
Price Value (PV)	0.772	Reliable
Habit (H)	0.923	Reliable
Behavioral Intention to Adopt (BIA)	0.818	Reliable
Behavioral Intention to Recommend (BIR)	0.793	Reliable

shows the reliability of the pilot test.

The reliability of the questionnaire was determined by analyzing Cronbach’s alpha value for each variable. As is visible from the reliability test results in Table 4, all variables are reliable because their Cronbach alpha value exceeds 0.700. The social impact variable (SI) has the highest value of 0.946. Meanwhile, the performance expectancy (PE) variable has the lowest Cronbach’s alpha value of 0.735. Therefore, it can be concluded that the generated questionnaire can be used in the next stage of the research because of its consistency and accuracy in measuring what the researcher wants to measure.

4.2. Profile of Respondents and Descriptive Statistics of the Field Test

In the final stage of this study, 100 samples of data were collected for further analysis. The process of distributing and collecting questionnaire data lasted for approximately one week. It is known that 65% of respondents have used the application for 1–3 months; the remaining 8% have used it for <1 month; 12% have used it for 3–6 months; and 15% have used it for >6 months. The following is a summary of the profiles of the field-test respondents.

Table 5. Profiles of the Field-Test Respondents.

Categories	Frequencies	Percentages
Age
15–20 years old	10	10%
21–25 years old	52	52%
26–30 years old	27	27%
31–35 years old	11	11%
Gender
Male	88	88%
Female	12	12%
Length of Use
<1 Month	8	8%
1–3 Months	65	65%
3–6 Months	12	12%
>6 Months	15	15%

shows the profiles of the field-test respondents.

Like the recapitulation of respondents at the initial stage, the field-test respondents were also those who were in their productive years and had a disproportionately large proportion of female and male respondents.

Table 6. Descriptive Statistics of Field Test.

Variables	Indicators	N	Minimum	Maximum	Std. Deviation	Means
Performance Expectancy (PE)	PE1	100	2.00	5.00	0.66	4.56	4.51
	PE2	100	3.00	5.00	0.54	4.48
	PE3	100	3.00	5.00	0.64	4.45
	PE4	100	2.00	5.00	0.61	4.53
Effort Expectancy (EE)	EE1	100	3.00	5.00	0.63	4.53	4.56
	EE2	100	3.00	5.00	0.63	4.54
	EE3	100	1.00	5.00	0.68	4.60
	EE4	100	3.00	5.00	0.67	4.58
Social Influence (SI)	SI1	100	3.00	5.00	0.58	4.69	4.63
	SI2	100	3.00	5.00	0.55	4.59
	SI3	100	2.00	5.00	0.62	4.60
Facilitating Conditions (FC)	FC1	100	3.00	5.00	0.59	4.60	4.58
	FC2	100	3.00	5.00	0.64	4.45
	FC3	100	3.00	5.00	0.56	4.65
	FC4	100	4.00	5.00	0.49	4.63
Hedonic Motivation (HM)	HM1	100	3.00	5.00	0.58	4.61	4.63
	HM2	100	3.00	5.00	0.54	4.64
	HM3	100	3.00	5.00	0.58	4.63
Price Value (PV)	PV1	100	3.00	5.00	0.56	4.53	4.48
	PV2	100	2.00	5.00	0.59	4.46
	PV3	100	3.00	5.00	0.57	4.44
Habit (H)	H1	100	2.00	5.00	0.67	4.45	4.49
	H2	100	2.00	5.00	0.72	4.40
	H3	100	2.00	5.00	0.67	4.52
	H4	100	3.00	5.00	0.67	4.58
Behavioral Intention to Adopt (BIA)	BIA1	100	3.00	5.00	0.66	4.48	4.52
	BIA2	100	3.00	5.00	0.62	4.41
	BIA3	100	2.00	5.00	0.61	4.66
Behavioral Intention to Recommend (BIR)	BIR1	100	2.00	5.00	0.59	4.65	4.63
	BIR2	100	1.00	5.00	0.68	4.60

shows the descriptive statistics of the field test.

The recapitulation of the descriptive statistics above shows the tendency of respondents to assess each variable through the average indicator. All question indicators were answered with the highest score on a Likert scale of five (representing the answers of strongly agree), while the lowest answers obtained were on a Likert scale of one (strongly disagree). Of the variability, the sample data obtained has a quite large standard deviation range, between 0.49 and 0.68. Based on the recapitulation, it is visible that the 100 responses given by respondents were mostly in the very significant category because most had a mean value higher than 4.21 [51]. Regardless of this, the social influence variable (SI1) has the highest mean value compared to the other variables, with a value of 4.69.

4.3. Partial Least Square–Structural Equation Modeling (PLS-SEM) Analysis

This study uses structural model analysis to determine whether all factors (indicators/manifest variables and variables/latent variables) are interrelated and influence the performance of the retail industry. The PLS-SEM analysis method was chosen because this method is advisable to determine the relationship between variables and identify the main driving factors in the construct [52]. PLS-SEM analysis will illustrate a path diagram and construct values between factors. The relationship hypothesis for this study is depicted by an arrow (blue circle) connecting one latent variable with another. Each latent variable has a measure called the inventory variable (yellow box). Numerical values in the figure represent factor loading values (located on arrows from latent variables to explicit variables), R² (located on latent variables), and path coefficient values (located on the arrows between latent variables). As explained in the research method, the PLS-SEM method performs two model evaluations, namely the evaluation of the external model and the evaluation of the internal model. The external model is the relationship between latent and explicit variables, which is called the evaluation of the measurement model, while the internal model is the path between latent variables, better known as the evaluation of the structural model [53]. Figure 4 shows the initial path model.

4.4. Evaluation of the Measurement Model

The evaluation process of the measurement model is conducted by evaluating the validity and reliability of the model. Model validity was evaluated based on convergent and discriminant validity. The convergent validity analysis was carried out by looking at the parameters of loading factor (outer loading) and average variance extracted (AVE) and discriminant validity using the parameter of cross-loading value and the Fornell–Larcker criterion. The results of the convergent validity recapitulation are shown in

Table 7. Recapitulation of Initial Convergent Validity.

Variables	Indicators	Outer Loading	AVE	Notes
Performance Expectancy (PE)	PE1	0.769	0.571	Valid
	PE2	0.73		Valid
	PE3	0.768		Valid
	PE4	0.754		Valid
Effort Expectancy (EE)	EE1	0.896	0.684	Valid
	EE2	0.896		Valid
	EE3	0.706		Valid
	EE4	0.795		Valid
Social Influence (SI)	SI1	0.696	0.589	Invalid
	SI2	0.788		Valid
	SI3	0.813		Valid
Facilitating Conditions (FC)	FC1	0.744	0.529	Valid
	FC2	0.727		Valid
	FC3	0.715		Valid
	FC4	0.722		Valid
Hedonic Motivation (HM)	HM1	0.512	0.55	Invalid
	HM2	0.802		Valid
	HM3	0.863		Valid
Price Value (PV)	PV1	0.702	0.584	Valid
	PV2	0.814		Valid
	PV3	0.772		Valid
Habit (H)	H1	0.76	0.635	Valid
	H2	0.821		Valid
	H3	0.815		Valid
	H4	0.79		Valid
Behavioral Intention to Adopt (BIA)	BIA1	0.875	0.637	Valid
	BIA2	0.779		Valid
	BIA3	0.736		Valid
Behavioral Intention to Recommend (BIR)	BIR1	0.802	0.722	Valid
	BIR2	0.895		Valid

. Meanwhile, the model reliability used the composite reliability and Cronbach’s alpha parameters. In this case, an indicator with an outer loading of less than 0.7 is considered invalid and needs to be removed from the model [54]. However, the automatic deletion of indicators with weak outer loading values must consider the effect of removing these items on AVE scores and composite reliability [55]. In Table 7, it is visible that there are two invalid indicators, namely the social influence (SI) variable in S1 (0.696) and hedonic motivation (HM) in HM1 (0.585). Based on these considerations, invalid indicators need to be removed from the model. The following is the result of convergent validity after being corrected. Table 7 shows the recapitulation of initial convergent validity.

The removal of invalid indicators causes an increase in the outer loading value on several indicators, and it can be ascertained that all indicators have an outer loading of more than 0.7 and are valid. The EE1 indicator has the highest outer loading value among other indicators, namely 0.897, which indicates that this indicator is the most capable of explaining what the researcher wants to know. The lowest outer loading is on the PV1 indicator (0.701) but it can still explain what the researcher wants to study. The improvements made also led to an increase in the AVE value on the social influence (SI) variable, which was initially worth from 0.589 to 0.727, and the hedonic motivation (HM) variable, which was initially worth from 0.55 to 0.734. Therefore, all variables have met the acceptance criteria for the AVE parameter and can be said to have good convergent validity. Furthermore, discriminant validity analysis was carried out to ensure the validity of the model construct.

Table 8. The Discriminant Validity Based on Cross-Loading.

	Behavioral Intention to Adopt	Behavioral Intention to Recommend	Effort Expectancy	Facilitating Conditions	Habit	Hedonic Motivation	Performance Expectancy	Price Value	Social Influence
BIA1	0.875	0.584	0.786	0.309	0.729	0.3	0.483	0.692	0.243
BIA2	0.781	0.507	0.607	0.344	0.623	0.25	0.255	0.499	0.263
BIA3	0.733	0.731	0.495	0.187	0.581	0.308	0.239	0.367	0.02
BIR1	0.566	0.802	0.36	0.166	0.482	0.067	0.363	0.434	0.261
BIR2	0.708	0.895	0.706	0.228	0.721	0.38	0.473	0.432	0.118
EE1	0.718	0.373	0.897	0.156	0.592	0.243	0.27	0.545	0.081
EE2	0.723	0.384	0.896	0.147	0.585	0.255	0.243	0.542	0.075
EE3	0.708	0.895	0.706	0.228	0.721	0.38	0.473	0.432	0.118
EE4	0.633	0.505	0.795	0.271	0.79	0.438	0.279	0.33	0.089
FC1	0.227	0.191	0.109	0.743	0.149	0.356	0.178	0.26	0.138
FC2	0.296	0.098	0.347	0.728	0.338	0.638	0.348	0.333	0.357
FC3	0.256	0.279	0.086	0.715	0.323	0.268	0.573	0.237	0.321
FC4	0.235	0.069	0.174	0.723	0.204	0.455	0.43	0.106	0.426
H1	0.57	0.467	0.53	0.338	0.761	0.442	0.389	0.447	0.359
H2	0.684	0.726	0.578	0.292	0.821	0.375	0.524	0.494	0.242
H3	0.685	0.572	0.686	0.253	0.815	0.281	0.36	0.471	0.064
H4	0.633	0.505	0.695	0.271	0.79	0.438	0.279	0.33	0.089
HM2	0.272	0.221	0.329	0.485	0.405	0.823	0.414	0.248	0.435
HM3	0.337	0.268	0.347	0.505	0.412	0.889	0.289	0.257	0.322
PE1	0.343	0.353	0.206	0.438	0.378	0.471	0.769	0.436	0.859
PE2	0.131	0.273	0.043	0.545	0.288	0.407	0.731	0.221	0.62
PE3	0.309	0.463	0.393	0.268	0.485	0.2	0.768	0.212	0.332
PE4	0.371	0.369	0.375	0.452	0.312	0.202	0.754	0.4	0.337
PV1	0.331	0.304	0.196	0.255	0.265	0.264	0.311	0.701	0.318
PV2	0.596	0.379	0.499	0.306	0.485	0.156	0.32	0.815	0.373
PV3	0.53	0.461	0.519	0.203	0.458	0.286	0.398	0.772	0.271
SI2	0.215	0.185	0.192	0.341	0.26	0.449	0.663	0.351	0.896
SI3	0.161	0.172	−0.035	0.387	0.117	0.267	0.493	0.368	0.807

shows the results of discriminant validity based on cross-loading.

The condition for accepting cross-loading parameters is when the correlation value between indicators on the same variable is higher than that indicator with other variables. Based on the recapitulation of cross-loading values in Table 8 above, it is visible that all indicators have met the requirements for accepting cross-loading parameters. For example, the correlation value of the BIA1 indicator with the BIA variable is 0.875, which is greater than the BIA1 correlation–BIR variable (0.584), the BIA1 correlation–EE variable (0.786), the BIA1 correlation–FC variable (0.309), the BIA1 correlation–H variable (0.729), the BIA1 correlation–HM variable (0.3), the BIA1 correlation–PE variable (0.483), the BIA1 correlation–PV variable (0.692), and the BIA1 correlation–SI variable (0.243). It has been ensured that the same happens with the correlation of other indicators. Therefore, it can be stated that the correlation between indicators and variables in the model is valid. Furthermore, the Fornell–Larcker parameter analysis was carried out to see the correlation between variables.

Table 9. Discriminant Validity Based on Fornell–Larcker.

	Behavioral Intention to Adopt	Behavioral Intention to Recommend	Effort Expectancy	Facilitating Condition	Habit	Hedonic Motivation	Performance Expectancy	Price Value	Social Influence
Behavioral Intention to Adopt	0.798
Behavioral Intention to Recommend	0.757	0.849
Effort Expectancy	0.745	0.653	0.827
Facilitating Condition	0.351	0.235	0.24	0.727
Habit	0.711	0.724	0.71	0.359	0.797
Hedonic Motivation	0.359	0.288	0.394	0.577	0.475	0.857
Performance Expectancy	0.42	0.498	0.384	0.539	0.494	0.401	0.756
Price Value	0.662	0.507	0.566	0.332	0.55	0.294	0.447	0.764
Social Influence	0.224	0.209	0.11	0.42	0.231	0.433	0.688	0.418	0.853

shows the result of discriminant validity based on the Fornell–Larcker.

The removal of invalid indicators causes an increase in the outer loading value on several indicators, and it can be ascertained that all indicators have an outer loading of more than 0.7 and are valid. The EE1 indicator has the highest outer loading value among other indicators, namely 0.897, which indicates that this indicator is the most capable of explaining what the researcher wants to know. The lowest outer loading is on the PV1 indicator (0.701) but it can still explain what the researcher wants to study. The improvements made also led to an increase in the AVE value on the social influence (SI) variable, which was initially worth from 0.589 to 0.727, and the hedonic motivation (HM) variable, which was initially worth from 0.55 to 0.734. Therefore, all variables have met the acceptance criteria for the AVE parameter and can be said to have good convergent validity. Furthermore, discriminant validity analysis was carried out to ensure the validity of the model construct.

Table 10. Recapitulation of the Final Convergent Validity.

Variables	Indicators	Outer Loading	AVE	Notes
Performance Expectancy (PE)	PE1	0.769	0.571	Valid
	PE2	0.731		Valid
	PE3	0.768		Valid
	PE4	0.754		Valid
Effort Expectancy (EE)	EE1	0.897	0.684	Valid
	EE2	0.896		Valid
	EE3	0.706		Valid
	EE4	0.795		Valid
Social Influence (SI)	SI2	0.896	0.727	Valid
	SI3	0.807		Valid
Facilitating Condition (FC)	FC1	0.743	0.529	Valid
	FC2	0.728		Valid
	FC3	0.715		Valid
	FC4	0.723		Valid
Hedonic Motivation (HM)	HM2	0.823	0.734	Valid
	HM3	0.889		Valid
Price Value (PV)	PV1	0.701	0.584	Valid
	PV2	0.815		Valid
	PV3	0.772		Valid
Habit (H)	H1	0.761	0.635	Valid
	H2	0.821		Valid
	H3	0.815		Valid
	H4	0.79		Valid
Behavioral Intention to Adopt (BIA)	BIA1	0.875	0.638	Valid
	BIA2	0.781		Valid
	BIA3	0.733		Valid
Behavioral Intention to Recommend (BIR)	BIR1	0.802	0.722	Valid
	BIR2	0.895		Valid

shows the recapitulation of the final convergent validity.

Acceptance of the Fornell–Larcker parameter occurs when the correlation value between variables with the same variable is greater than the correlation value between these variables and other variables. In other words, the diagonal value of each variable must be greater than the other parallel values. As can be seen in Table 10, the values in gray are greater than the other parallel values. For example, in the second row, the BIR variable has the same correlation value between variables, which is 0.849, which is greater than the BIA–BIR variable correlation value (0.757), the EE–BIR variable correlation (0.653), the FC–BIR variable correlation (0.235), the H–BIR variable correlation (0.724), the HM–BIR variable correlation (0.288), the PE–BIR variable correlation (0.498), the PV–BIR variable correlation (0.507), and the SI–BIR variable correlation (0.209). It has been confirmed to occur in the correlation of other variables, and it can be concluded that the correlation between variables in the model is valid. Therefore, the discriminant validity of the model fully meets the requirements and is declared valid.

Table 11. Reliability of Formal Questioners.

	Cronbach’s Alpha	Composite Reliability	Notes
Behavioral Intention to Adopt	0.713	0.84	Reliable
Behavioral Intention to Recommend	0.721	0.838	Reliable
Effort Expectancy	0.842	0.896	Reliable
Facilitating Condition	0.707	0.818	Reliable
Habit	0.809	0.874	Reliable
Hedonic Motivation	0.741	0.846	Reliable
Performance Expectancy	0.762	0.842	Reliable
Price Value	0.755	0.807	Reliable
Social Influence	0.731	0.841	Reliable

shows the results of the reliability of formal questionnaires.

Composite reliability is used as a supporting parameter in terms of the interrelationship of the outer loading. It is visible in Table 11 that the composite reliability value obtained ranges from 0.807 (price value variable) to the highest value of 0.896 (effort expectancy variable). This finding states that the reliability is very good and strong because the value of composite reliability is in the range from 0.7 to 0.9 [56]. Therefore, it can be concluded that all variables are very reliable and have a very strong relationship.

4.4.1. Structural Model Evaluation

A structural model evaluation, or inner model, is conducted after all manifest variables are declared valid and reliable. This evaluation aims to evaluate the influence of constructs between latent variables in the research model. An overview of the final construct model that is valid and reliable is visible in Figure 5. The structural model evaluation is conducted by analyzing the parameters of the coefficient of determination or R-square, the path coefficient, and predictive relevance (Q2). Figure 5 shows the final model.

The path coefficient parameter values range from −1 to +1, with values close to +1 indicating a strong relationship and vice versa. The further explanation says that a value of less than 0.15 is declared weak, a value from 0.15 to 0.45 is stated as moderate, and if it is higher than 0.45, it is stated as strong (Rodliyah, 2016). Based on the recapitulation of the path coefficient values in

Table 12. Recapitulation of Bootstrapping.

Hypotheses	Description of the Relationship	T Statistics	T Table	Evidence
H1	Performance expectancy has a positive effect on the behavioral intentions of PINTU users.	1.977	1.96	Evident
H2	Effort expectancy has a positive effect on the behavioral intentions of PINTU users.	4.038	1.96	Evident
H3	Social influence has a positive effect on the behavioral intentions of PINTU users.	2.84	1.96	Evident
H4	Facilitating conditions have a positive effect on the behavioral intentions of PINTU users.	2.341	1.96	Evident
H5	Hedonic motivation has a positive effect on the behavioral intentions of PINTU users.	2.117	1.96	Evident
H6	Price value has a positive effect on the behavioral intentions of PINTU users.	3.368	1.96	Evident
H7	Habit values have a positive effect on behavioral intentions to use PINTU.	2.053	1.96	Evident
H8	The behavioral intention of PINTU users has a positive effect on the behavioral intentions to recommend PINTU to others.	2.381	1.96	Evident

, it is visible that the hedonic motivation (HM) and performance expectancy (PE) variables are negative, which means they have a weak relationship value. In addition, it is also visible that the technology adoption variables of price value (PV) (0.209), habit (H) (0.298), and facilitating condition (FC) (0.410) have a moderate relationship with the behavioral intention to adopt (BIA) variable. While the relationship between the effort expectancy (EE) variable and the behavioral intention to recommend (BIR) variable and the behavioral intention to adopt (BIA) variable with the behavioral intention to recommend (BIR) variable has a strong relationship because it is greater than 0.45.

4.4.2. Hypothesis Testing

Hypothesis testing aims to define the significance level of exogenous variables (variables of behavioral intention to adopt and behavioral intention to recommend) to endogenous variables (variables PE, EE, SI, FC, HM, PV, and H). The test was carried out using the bootstrapping technique and has acceptance conditions of the T-statistic value ≥ T table or p value ≤ significance level (α). The significance level used in this study was 5% using a two-tailed test, so the T table value used was 1.96 [57]. The following are the results of hypothesis testing based on the path between exogenous and endogenous variables. Table 12 shows the recapitulation of bootstrapping.

In this study, the factors influencing behavioral intention (behavioral intention to adopt) are visible. As we can see in the Figure 5, seven factors directly influence each other, with the following details:

• Performance Expectancy

Performance expectancy is the extent to which the use of technology will benefit consumers in carrying out certain activities [18]. It means that using the PINTU application will help gain benefits by improving the task performance of buying and selling shares, thereby influencing the behavioral intention to adopt the application. Then, for the value of the effect, it is visible in Table 4.12 on the path coefficient value, where performance expectancy has a negative influence on the behavioral intention of PINTU users by −0.113; this means that, according to [58], it shows that an indication of being rejected due to the use of the PINTU application must be supported by an application installed on the smartphone and connected to the internet, where sometimes unstable network conditions affect the use of the application.

• Effort Expectancy

Effort expectancy is the level of convenience associated with the use of technology by consumers [18]. Based on C. L. Miltgen, A. Popovič, and T. J. D. s. s. Oliveira (2013) [39], it will contribute to the correct prediction of the intention to adopt new technology. When users find the PINTU application easy to use and effortless, they will have higher expectations of obtaining the desired performance [17]. Then, for the value of the effect, it is visible in Table 4.12 on the path coefficient value, where effort expectancy has a positive influence on the behavioral intention of PINTU users by 0.52, which means that the application user gets convenience in using the application and gets the appropriate performance results against the user’s wishes.

• Social Influence

Social influence is the extent to which consumers perceive that significant others (e.g., family and friends) believe they should use certain technologies [18]. It reflects the effect of environmental factors such as the opinions of friends, relatives, and users on behavior [17] when they can positively encourage users to adopt the technology. Then, for the value of the effect itself, it is visible in Table 4.12 on the path coefficient value, where social influence has a positive influence on the behavioral intention of PINTU users by 0.085, which means that environmental factors are very influential on users when they want to use the application. The assumptions of surrounding people about the application are considered very important when wanting to adopt it.

• Facilitating Conditions

Facilitating Conditions (FC) refers to consumer perceptions of the resources and support available to perform a behavior [18]. If the operational infrastructure exists and supports the PINTU application, the behavioral intention to adopt the application will increase. Then, for the value of the effect, it is visible in Table 4.12 on the path coefficient value, where facilitating conditions have a positive influence on the behavioral intention of PINTU users by 0.148, which means that PINTU application users have adequate facilitating facilities; for example, internet, mobile devices, and others that are adequate to support the use of the application.

• Hedonic Motivation

Hedonic motivation (HM) is defined as pleasure or fun derived from the use of technology. In this context, consumer hedonic motivation is a critical determinant of technology adoption and use [18]. As an activator of a new form of buying and selling crypto shares, the PINTU application is fun for users, which can encourage them to adopt the application. Then, for the value of the effect, it is visible in Table 4.12 on the path coefficient value, where hedonic motivation has a negative influence on the behavioral intention of PINTU users by −0.138, which means that, according to Watmah et al. (2020), it shows an indication of being rejected. This is because the use of the PINTU application has not provided users with pleasure and satisfaction; this may be due to an unattractive interface display or problems in using the application, such as bugs, lagging, and so on.

• Price Value

Ref. [18] define price value as the cognitive consumer trade-off between the perceived benefits of technology and the monetary costs of using it. The perceived benefits of using technology are greater when the price value is higher and the perceived monetary cost is lower. Therefore, price value positively influences the intention to adopt mobile payments. Then, for the value of the effect, it is visible in Table 4.12 on the path coefficient value, where the price value has a positive influence on the behavioral intention of PINTU users by 0.148, which means that PINTU application users have felt a significant impact on the use of the application, which has a direct impact on ignoring costs issued on the application.

• Habit

Habit refers to the extent to which individuals tend to perform behaviors automatically using technology [18]. It becomes a conscious awareness that reflects the results of previous experiences. To study more about habits, technology users must have a long history of using technology [59]. Then, for the value of the effect, it is visible in Table 4.12 on the path coefficient value, where it can be concluded that the habit value has a positive influence on the behavioral intention to use PINTU by 0.333, which means that users of the PINTU application have a good experience when using the application so that it forms a habit.

Consumers with higher intentions to adopt new technologies are more likely to become adopters [59] and to recommend those technologies to others [39]. Social networks bring several challenges and opportunities to companies [60] because they represent a means of communication that allows users to express their opinions and experiences about mobile payment services, products, and technologies. From Table 5.1, it can be concluded that the biggest influencing factor is a behavioral intention to adopt with a positive influence value of 0.506, which means that when the PINTU application user has decided to intend to adopt or use the application’s technology, the user will organically recommend the application to others who have not used it yet. Meanwhile, the influence factor of the facilitating conditions, as seen from the table above, has a negative value of −0.063, which means that according to [58], it shows an indication of being rejected. It is due to the use of the PINTU application, where the greater the value of the supporting facilities, the smaller the value of the intention to recommend. For example, when a user has supporting facilities such as the internet, mobile devices, and so on, it will not affect his/her intention to recommend this application to others.

5. Managerial Implications

The managerial implications in this section are expected to be able to provide theoretical contributions that can improve the PINTU application performance. The compilation of managerial implications is based on indicators with the highest factor loading values on exogenous and moderator variables. The following are suggestions given by researchers:

• Effort expectancy (EE): when users feel the PINTU application is easy to use and effortless, they have higher expectations of obtaining the desired performance [17]. It means that application users get convenience in using the application and get the appropriate performance results for their wishes. Therefore, the developer company must be able to make the application easy for users to use so that it can attract interest in using it, which will have an impact on the number of users who recommend the application so that this application can meet the performance needs well according to what the user wants. Excellent companies are those that succeed in satisfying and delighting their customers. Customer satisfaction contributes to a number of crucial aspects, such as creating customer loyalty, increasing company reputation, reducing price elasticity, reducing future transaction costs, and increasing employee efficiency and productivity [61];
• Environmental factors are highly influential on users’ decisions to use the application. Assumptions about the application from the people around are considered very important when intending to adopt the application. Word-of-mouth communication spreads through business, social, and community networks, which are considered very influential, suggesting that word-of-mouth communication is personal communication between customers or members of a group. Information obtained by customers through trusted people such as experts, friends, and family tends to be received more quickly [62]. Companies must be able to communicate their products well so that they can be easily accepted by users, which will later be supported by social influence factors (SI);
• In this context, consumer hedonic motivation (HM) is a critical determinant of technology adoption and use [18] for the PINTU application. As an activator of a new form of buying and selling crypto shares, PINTU is fun for users, which can encourage them to adopt the application. Here, the role of the company is very critical for continuing to develop the appearance and performance of the application so that it can attract users to continue using it. The company is expected to be able to be consistent and continue to carry out periodic developments in the future.

6. Conclusions

Based on the results of obtaining the path coefficient and testing the hypothesis with the bootstrapping technique using the SmartPLS 0.3 software, it can be concluded that the seven variable hypotheses have a relationship to the user’s behavioral intention (behavioral intention to adopt). The first hypothesis concluded that performance expectations (PE) have a positive and significant effect on user behavioral intentions. In the second hypothesis, it was concluded that business expectations (EE) have a positive and significant effect on user behavioral intentions. The third hypothesis concluded that social influence (SI) has a positive and significant effect on user behavioral intentions. The fourth hypothesis concluded that the supporting conditions (FC) have a positive and significant effect on the user’s behavioral intention. The fifth hypothesis concluded that the influence of hedonic motivation (HM) has a positive and significant effect on the user’s behavioral intention. The sixth hypothesis concluded that price value (PV) has a positive and significant effect on user behavioral intentions. Furthermore, the last hypothesis concluded that the habit value (H) has a positive and significant effect on the user’s behavioral intention.

Based on the results of obtaining the path coefficient and testing the hypothesis with the bootstrapping technique using SmartPLS 0.3 software, it can be stated that the user’s behavioral intention (behavioral intention to adopt) has a relationship to the behavioral intention to recommend (behavioral intention to recommend). In the eighth hypothesis of this study, it can be concluded that the user’s behavioral intention (behavioral intention to adopt) positively and significantly influences the intention to recommend (behavioral intention to recommend) PINTU to others. In other words, the user’s behavioral intention (behavioral intention to adopt), which is supported by PE, EE, SI, FC, HM, PV, and H, can influence the user’s behavioral intention to recommend PINTU to others.

In data processing, the loading factor values for all latent variable indicators are obtained that describe behavioral intentions to recommend (behavioral intention to recommend), especially related to behavioral intentions of users (behavioral intention to adopt). Based on the analysis of indicators that influence behavioral intention to recommend, it can be seen that each variable indicator has the most influence based on the loading factor value. The performance expectancy (PE) variable has the most influential indicator, namely the PE1 indicator “mobile commerce (PINTU) is useful in my daily life.” The effort expectancy (EE) variable has the most influential indicator, namely EE1 “My interaction with mobile commerce (PINTU) will be clear and understandable.” The social influence (SI) variable has the most influential indicator, namely the SI2 indicator “People who influence my habits think that I should use mobile commerce (PINTU)”. The facilitating conditions (FC) variable has the most influential indicator, namely the FC1 indicator “I have the necessary resources to use mobile commerce (PINTU)”. The hedonic motivation (HM) variable has the most influential indicator, namely the HM3 indicator “Using mobile commerce (PINTU) is very entertaining”. The price value (PV) variable has the most influential indicator, namely the PV2 indicator “Mobile commerce (PINTU) is affordable for finance”. Finally, the habit variable (H) has the most influential indicator, namely H2 “I am addicted to using mobile commerce (PINTU)”.

This research has limitations and needs to be refined through further research. This research focuses only on testing the usability of currency trading applications (Cryptocurrency) with the indicator The unified theory of acceptance and use of technology 2 (UTAUT2). In addition, the company is expected to continue to improve and maintain the image of the company itself and the image of the products it produces because these two components have proven to be benchmarks for consumers when deciding to buy a product.