Mobile communication devices are so ubiquitous that there are more people with mobile devices than with toothbrushes in the world (MMA Asia, 2011). In response to this trend, marketers are significantly increasing the use of mobile devices to meet the demands of mobile consumers. Data from eMarketer (2015) show us that mobile business spending now accounts for one-quarter of overall digital spending.

In fact, innovative mobile technologies deliver new tools (applications) that make it possible to separate time of purchase and the time of actual consumption by allowing consumers to make a purchase from a mobile device and collect the product at home or in the store, in addition to the traditional model of purchase and consumption in the store (Pantano and Priporas, 2016).

Although the use of mobile in marketing has grown dramatically, research on this topic has evolved gradually. Studies have focused on issues such as mobile marketing (Shankar and Balasubramanian, 2009), delivery of mobile services (Kleijnen et al., 2007), use of the mobile interface and usability (Venkatesh et al., 2012), the mobile navigation experience (Adipat et al., 2011), applications for sales (Shankar et al., 2010), interfaces for mobile devices (Brasel and Gips, 2014), demand for mobile applications (Garg and Telang, 2012), mobile advertising and promotions (Andrews et al., 2015; Fong et al., 2015), and purchases via mobile devices (Wang and Sun, 2016).

For Shankar et al. (2016), the term mobile encompasses several aspects, including device, medium, technology, and channel. Device refers to the device, such as a smartphone, tablet, and notebook. The medium is about the media, such as applications, email, and print. Technology is the hardware and software behind communication, such as fixed broadband and wireless broadband. The channel concerns the transaction mode, such as mobile, desktop, phone, and physical store. All these factors affect mobile marketing.

Mobile marketing can be seen as an evolution of e-commerce, in which traditional internet based marketing shifted to the mobile channel in response to the emerging trend of consumers losing interest in traditional marketing channels (Hinz et al., 2011). For Maduku et al. (2016), mobile marketing is a set of marketing practices that uses mobile technologies and wireless networks to create interactive and personalized communication between the organization and its target audience, resulting in the creation of value for both parties in this ubiquitous environment.

When mobile marketing meets online retail via mobile devices, there is mobile commerce. Mobile commerce is a growing format involving mobile computing technologies, with growing popularity and significant potential (Nilashi et al., 2015). In this scenario, mobile devices (e.g., smartphones, tablets, etc.) are becoming essential on a daily basis, making communication more convenient (Dinh et al., 2013; Wong et al., 2015) and allowing for various types of mobile services such as SMS, MMS, GPS, mobile payment, and mobile banking (Hsiao and Chen, 2015) and the emergence of sales and related platforms with exclusive advantages, mainly via applications and e-commerce sites with mobile versions, among others.

The use of these mobile devices in m-commerce corroborates the formation of two central features in this platform, which are ubiquity and location. These aspects are not found in e-commerce; thus, companies are able to connect with consumers anytime, anywhere (Huang et al., 2015).

Because of the monetary transactions that occur in mobile commerce, the main influencing factors for consumer acceptance are related to risk aversion, which include trust in the transaction, privacy concerns, and network security (Grob, 2015; Ng, 2016). These aspects are included in the perceptions of the probability of acceptance of the m-commerce format by people and potential consumers. Therefore, the next section addresses the theories and models most commonly used in research to understand consumer acceptance of technologies.

Here are the theories and models that this research adopts as the antecedents constructs of perceived value.

Technology Acceptance Model (TAM)

The Technology Acceptance Model (TAM) was designed by Davis (1989) to specifically apply to computer-related behavior by incorporating findings accumulated over more than a decade into information systems. It can be used to model computational acceptance.

The work of Davis (1989) aimed to develop and validate new scales for two specific variables: perceived utility and perceived ease of use. These two variables were tested as fundamental determinants of user acceptance.

In his article that gave rise to the TAM, Davis asks what makes an individual accept or reject an information technology. For him, among the several variables that can influence the use of the systems, two are particularly important. First, people tend to use or not use an application insomuch as they believe that the application will help them to better perform their activities. He referred to this first variable as perceived utility. However, even if a potential user believes that a given application is useful, he may simultaneously believe that the system is too difficult to use, and thus, the benefits are outweighed by the effort required to use it.

The TAM is widely used today (Pando-Garcia et al., 2016; Wang and Sun, 2016; Abdullah et al., 2016; Agag and Masry, 2016). Although there is extensive empirical support for the TAM, it does not fully explain all aspects of technology adoption. Recent studies show that technology adoption can be highly motivated by other factors that are not part of the traditional components of the TAM, such as the perceived subjective adoption of monetary and non-monetary costs (Chen and Dubinsky, 2003; Kim et al., 2007) or social influence (SI) by others, such as friends or family members (Kwon and Chon, 2009). Thus, the following theories seek to fill some of the gaps left by the TAM.

Innovation Diffusion Theory (IDT)

Innovation Diffusion Theory (IDT) (Rogers, 1995) has been commonly applied to new technologies and services to predict their adoption. IDT proposes that within a population group, innovations are not adopted simultaneously by all individuals. The theory argues that the rate of innovation adoption of individuals is determined by a wide variety of personal (gender, ethnicity, age), social (education, social class) and technological (utility and perceived benefits) factors (Leung and Wei 1999).

IDT assumes that some people are more willing to test innovative ideas and technologies than others. Rogers (1995) argues that sociodemographic factors such as gender, age, education, and income can identify when individuals will adopt innovative ideas or technologies. He categorized individuals into five categories of adopters: innovators, initial adopters, initial majority, late majority, and latecomers (Kim et al., 2014).

Rogers (2003) defines the four main factors for the diffusion of new ideas: (1) innovation, (2) communication channels, (3) time, and (4) the social context. Of these four factors, we will address the first. Despite advancing in some directions, the IDT-based adoption model is still fragile because it takes into account only utilitarian reasons, leaving aside symbolic and social aspects.

Unified Theory of Acceptance and Use of Technology (UTAUT)

The Unified Theory of Acceptance and Use of Technology (UTAUT) was designed to unify the various existing theories about how users accept technology (Venkatesh and Morris, 2000; Venkatesh et al., 2003). The UTAUT is created from the following eight widely used theories: Theory of Rational Action (TRA) by Davis et al. (1989); the Technology Acceptance Model (TAM) by Davis (1989); the Motivation Model (MM) by Davis et al. (1992); Theory of Planned Behavior (TPB) and Combined TAM and TPB (C-TAM-TPB) by Taylor and Todd (1995); the Model of PC Utilization (MPCU) by Thompson et al. (1991); Innovation Diffusion Theory (IDT) by Moore and Benbasat (1991); and Social Cognitive Theory (SCT) by Compeau et al. (1999).

Based on the combination of these eight theories, the UTAUT model explains the behavioral intention to use or adopt the technology, proposing four predictive determinants (Venkatesh et al., 2003): expected performance, expected effort, social influence, and facilitating conditions. Venkatesh et al. (2003) also identified four major moderators that may affect the relationship between the main determinants and intent: gender, age, experience, and willingness.

Consumer Acceptance of Technology Model (CAT)

As the main objective of his work, Kulviwat et al. (2007) incorporated the PAD (Pleasure, Arousal, and Dominance) emotions paradigm (Mehrabian and Russell, 1974) for the Technology Acceptance Model (Davis, 1989), which is the most popular model used to predict the adoption of technology. An additional goal was to improve the concept of cognition by adding a key belief, relative advantage, which involves the extent to which an innovation is superior compared to alternatives. He considers that the CAT model is more powerful in describing and predicting adoption by the consumer.

Kulviwat et al. (2007) sought to improve the TAM with the proposal of the CAT model, which adds an affective dimension of emotion to the models of technology acceptance, thus conferring greater explanatory power.

Value-Based Adoption Model (VAM)

Value is a construct strongly emphasized in the field of economics, and it has its foundation in theories of exchange, utility, and value of work, and in marketing, accounting, and finance, with roots in psychology and social psychology as well (Kim et al., 2007). Researchers have already assigned several terms to define value, generally differentiating the context from the same basic concept: value of consumption (Sheth et al., 1991), value of acquisition and transaction (Thaler, 1985), service value and customer value (Woodruff, 1997), consumer value (Holbrook, 1999), and perceived value (Zeithaml, 1988).

For Kim et al. (2007), modeling the perceived value of a product solely based on its price is important but insufficient. Most of the time, consumers consider factors other than price, such as the perceived quality of the product. The most simplistic trade-off models ignore the multidimensionality of the decision-making process and do not fully represent perceived benefits and sacrifices. Thus, Kim et al. (2007) proposed the Value Based Adoption Model (VAM), which aims to fill gaps that still exist in previous models.

The study by Kim et al. (2007) was shown to be more effective using the VAM model to explain the adoption of the mobile internet by consumers (context researched in the article) than the traditional TAM. The next topic of this work will address the constructs that are part of the theories and models discussed here, which are better suited to the context of mobile commerce and which constitute antecedents of perceived value.

Table 1 shows a comparison among the theories discussed.

Table 1.
Comparison among theories.

Source: Prepared by the authors, based on the models (2017).

The studied theories have similarities with each other. However, over time, new constructs have been included in the theories of technology acceptance. In 1989, TAM established the constructs perceived utility and perceived ease of use as predictors of attitude, and this attitude reflects on behavioral intent that influences actual use. Six years later, IDT brought the constructs relative advantage, which at some level is related to the perceived utility from TAM; compatibility; complexity, which is similar to the perceived ease of use also from the TAM model; experimentability and observality as predictors of adoption.

Despite advancing in the field with new constructs, IDT still focused more on utilitarian aspects, leaving aside symbolic and sociodemographic ones. In 2003, the UTAUT brings as new constructs to the theories of technology acceptance the social aspect, through the construct social influence, and also for the first time takes into account demographic aspects as moderators. CAT is presented in 2007, and differs from previous ones by bringing an emotional dimension to the model, which is composed of three constructs (pleasure, excitement and domain).

Still, also in 2007, VAM is the first model that brings the construct perceived value to a technology acceptance model. It is presented as a predictor of the intention to adopt the technology. And it is in this way that the model of this current research is based. Perceived value is discussed more comprehensively in the following section.

In this section, the constructs that are part of the theories and models discussed in the past topic (TAM, IDT, UTAUT, CAT and VAM) will be approached. The ones which in some way are adequate to the reality of mobile commerce with regard to consumer acceptance and that are understood as antecedents, that have a direct influence on the perceived value.

The constructs discussed forward are perceived utility and perceived ease of use (TAM), relative advantage, compatibility and complexity (IDT), facilitating conditions (UTAUT), pleasure (CAT) and perceived cost (VAM).

Perceived utility

Perceived utility is one of the most studied variables when it comes to technology adoption (Jeyaraj et al., 2006). For Davis (1989), perceived utility is defined as the degree to which a person believes that using a particular system would increase their performance in their activities. Based on the exposed, the first research hypothesis is elaborated:

H1: Perceived utility positively influences perceived value.

Perceived ease of use

Perceived ease of use refers to the degree to which a person believes that using a particular system would be completely effortless (Davis, 1989). As perceived utility, perceived ease of use is one of the variables found in the original TAM model. In face of what had been discussed, the second research hypothesis is formulated:

H2: Perceived ease of use positively influences perceived value.

The relative advantage

The relative advantage is a construct of the IDT adoption model. For Rogers (2003), relative advantage is the degree to which a person perceives that the new technology is better than the one it replaces. From this discussion, the third hypothesis of research is drawn:

H3: Relative advantage positively influences perceived value.

Compatibility

As the relative advantage, compatibility is also a construct borrowed from IDT. Rogers (2003) defines it as the degree to which an innovation is perceived to be consistent with the values, past experiences, and needs of potential adopters. The lack of compatibility in information technology with the needs of the individual can negatively affect the use of technology (Mckenzie, 2001). Considering what was discussed, the fourth research hypothesis is elaborated:

H4: Compatibility positively influences perceived value.

Complexity

Complexity is also an IDT construct. Rogers (2003) defines it as the degree to which an innovation is perceived relatively difficult to understand and use. In contrast to the other constructs, complexity is negatively related to the adoption rate. Thus, the fifth research hypothesis is formulated:

H5: Complexity negatively influences perceived value.

Facilitating conditions

Facilitating conditions is a construct that comes from the UTAUT model (Venkatesh et al., 2003). They refer to the degree to which the individual considers that there are technical and organizational conditions that help facilitate the use of technology (Konrad et al., 2015). In view of the discussion, the sixth research hypothesis is drawn:

H6: Facilitating conditions positively influence perceived value.

Pleasure

Pleasure is a construct that constitutes the emotional dimension in CAT (Kulviwat et al., 2007). The pleasure was discovered as a strong factor of positive influence on the attitude towards shopping on the internet (Lee et al., 2003) and when operationalized as fun, it has a direct positive effect on the attitude toward the use of mobile devices (Bruner and Kumar, 2005). Thus, the seventh research hypothesis is elaborated:

H7: Pleasure positively influences perceived value.

Perceived security

According to Vatanasombut et al. (2008), perceived security is defined as the perception that people have that interactions on any platform are secure. Woodwall (2003) identified the perception of risk as a determining factor for the perception of value and the identification of benefits in the intentions of purchase. Thus, the eighth research hypothesis is formulated:

H8: Perceived security positively influences perceived value.

Perceived Cost

The perceived cost is a construct derived from the sacrifice dimension of the VAM model (Kim et al., 2007). It symbolizes the encoding or internalization of the target price of a product or service (Jacoby and Olson, 1976). And marketing studies show that perceived monetary value and perceived value are negatively related (Chang and Wildt, 1994). Thus, the ninth hypothesis of research is drawn:

H9: Perceived cost negatively influences perceived value.

From the consumer perspective, obtaining value is a substantial consumption goal in a successful buying experience (Davis and Hodges, 2012). For Schechter (1984), perceived value is composed of quantitative and qualitative factors, objective and subjective, which together form the buyer’s experience. Dodds et al. (1991) defines perceived value as the ratio between perceived benefits and perceived sacrifices. Woodruff and Gardial (1996) define perceived value as a trade-off of desirable attributes and attributes of sacrifice.

However, the widely accepted definition of perceived value is found in Zeithaml (1988), who stated that consumers define value in four ways: value is (1) low price, (2) what I want in a product, (3) quality proportional to the price paid, and (4) what I get for what I give. He still synthetically conceptualized the perceived value as the overall appreciation of the consumer for the utility of a product or service, determined by the consumer’s perception of what is given and what is received.

This value can be increased both by improving benefits for the individual and by reducing the costs of buying and using the good or service (Lovelock and Wirtz, 2011). Perceived value is commonly assessed by the relationship between quality and price, where it increases if the quality increases or if the price falls. However, the valuation of the total value is not based solely on price and quality. Previous studies have suggested several types of value: functional, social, emotional, epistemic, and conditional (Hsiao and Chen, 2016).

However, apparently only two types of value are more applicable to consumer behavior studies. The first refers to the functional motives, which includes more tangible needs such as price, convenience, and quality. The second is related to non-functional motives, which are more linked to intangible desires, such as emotional and social needs (Chen and Hu, 2010). Perceived value still influences the intention to buy (Li and Petrick, 2008; Tuškej et al., 2013).

Therefore, the last research hypothesis is proposed:

H10: Perceived value positively influences the adoption of m-commerce.

From the triad ‘m-commerce’, ‘technology acceptance theories’, and the pillars of ‘perceived value’, we constitute the theoretical principles guiding this article.

Regarding the sample, in terms of the sex of the individuals, a greater female (64.8%) representation was observed compared to males (35.2%), and for age group, the majority of the respondents were concentrated between 22 and 30 years of age (60%), with higher (55.1%) and postgraduate (40.9%) education levels. Most of them belonged to class B (IBGE, 2015), with 77% of the respondents being single, followed by married couples (21.2%).

Although there was not large heterogeneity in most sociodemographic variables, this does not affect the subsequent phases of analysis because the present research does not aim to generalize its results, but only to identify evidence of relationships among the studied constructs. To that end, the sample is adequate.

An exploratory analysis of the data noted that people who had previously made some purchase through a mobile phone presented averages that were different for the assertions that compose the constructs compared to those who never performed such purchases. Therefore, it was pertinent to investigate this behavior of the sample to verify whether the difference was only apparent or whether it was statistically significant. Table 4 shows the extractions of the tests performed.

Table 4.
Analysis of the difference between means.

Source: Study data (2017).

The data in Table 4 show that the only constructs in which those who had previously made some purchase through the mobile phone had lower average values ​​than those who had never previously bought anything were perceived complexity and perceived cost. In the remaining nine constructs, those who had previously made some purchase via mobile phones had higher averages than those who had never bought anything. Next, the group of people who declared that they had previously made e-commerce purchases via mobile phones was analyzed, which included 52.7% (238) of the respondents. This analysis was chosen due to the inherent limit of the study structure because it would be impracticable to offer an extended discussion and because in directing, one can better point out the particularities of a specific group that tends to increase daily, which are Brazilian e-consumers who are potentially m-consumers.

Two models were tested. The first one with all the data and the second only with the respondents who have already bought via m-commerce.

Hypotheses testing for the general model

The hypotheses were tested using Structural Equation Modeling. Table 5 shows the statistics for the initial general model.

Table 5.
Initial general model.

Source: Study data (2017).

These statistical data reveal that seven of the ten hypotheses traced were confirmed. Only the constructs complexity, facilitating conditions, and perceived cost had no influence on perceived value. This model obtained a R-squared coefficient (R²), that is, percentage of the response construct variation that is explained by the antecedent constructs, of 61% for the influence of the constructs derived from the theories and models of technology acceptance in the perceived value, and of 50% for the influence of perceived value on the intention to adopt m-commerce. Although it was a good result, this model is not ideal, because it has constructs which hypotheses have been refuted. Thus, a final model (Table 6) was generated only with accepted hypotheses.

In this final model we can observe an increase in the betas of some constructs, such as the perceived ease of use, which now has a greater contribution to the perceived value. The R² remained at 61% and 50%.

Table 6.
Final general model.

Source: Study data (2017).

Hypotheses testing for the model of those who have already bought

As in the previous model, a structural equation modeling was also performed here to test the hypotheses. However, this model contains responses only from people who have already made purchases via mobile devices.

The results of the initial structural equation modeling reveal that perceived ease of use (p-value=0.357), complexity (p-value=0.312), facilitating conditions (p-value=0.455) and perceived cost (p-value=0.132) are constructs that have no influence on perceived value. This initial model obtained R² of 58.5% for the influence of the constructs derived from the theories and models of technology acceptance in the perceived value, and 42% for the influence of perceived value on the intention to adopt m-commerce.

A final model was generated, with the exclusion of constructs that did not show influence on the previous analysis. Statistical data can be found in Table 7.

In this final model, the R² for the influence of the constructs derived from the theories and models of technology acceptance in the perceived value was 56%, and 42% for the influence of perceived value on the intention to adopt m-commerce.

In the following section the results of hypotheses tests are discussed.

Table 7.
Final model of those who have already bought.

Source: Study data (2017).

Table 8 presents a summary of the hypotheses test results, with ten research hypotheses, a description of each, their influence on the response variable (positive or negative), and the test results.

Table 8.
Summary of hypotheses test results.

Source: Study data (2017).

H1 to H9 are nine predictors of perceived value, which is the response variable. The first hypothesis (H1) to be formulated was that ‘perceived utility positively influenced the perceived value’ that people have about m-commerce. Corroborating with the works of Chong et al. (2012) and Faqih et al. (2015), the hypothesis was accepted, showing that it is a construct of great importance in the study of the perceived value of individuals regarding the use of mobile phones to make purchases and revealing that people believe that the use of this technology will yield better performance of their activities.

The second research hypothesis (H2), ‘perceived ease of use positively influences perceived value’, was rejected in the second model. This is not to say that people who have previously bought do not believe that m-commerce is easy to use. In the descriptive phase of the study, those who have previously bought something via a mobile phone have a significant mean in the perceived ease of use construct. One can interpret this result as follows: people who have had some experience with m-commerce perceive technology as being easy to use, but however much they find it easy to use, it is not this feature that makes them pursue value in m-commerce. Because they are already familiar with the technology, the simple fact that they do not have much difficulty in using it does not add to the perceived value because the use of technology is something very natural, and its ease does not become a factor of weight in the formation of perceived value.

The relative advantage construct contemplates the third research hypothesis (H3), which assumes that people will have a greater perception of perceived value, in that they believe m-commerce is a superior technology and better than its predecessors. The hypothesis was accepted. This result adds to the studies of Rogers (2003) and his theory of the diffusion of innovations.

The fourth research hypothesis to be launched (H4) was that ‘compatibility influences perceived value’. This hypothesis was accepted by showing that the degree of compatibility with the values, past experiences, and lifestyle that an innovation has, directly affects, in a positive way, the perceived value of individuals toward m-commerce (Agag and Masry, 2016).

The complexity construct is the fifth research hypothesis (H5), in which it is assumed that ‘the greater the complexity of a technology, the lower the perceived value’ of people (Ittersum et al., 2006). The hypothesis was rejected by those who have already bought via a mobile phone. It is understood that because one who has never used technology tends to perceive it as more complex, it is expected that the more difficult and complicated it is to use technology, the less value people who have never used the technology will perceive. The opposite is likely to occur with those who have already purchased, as is the case in this study sample.

The sixth research hypothesis (H6) was that ‘facilitating conditions positively influenced the perceived value’ of individuals. Although this construct presented moderate means in the descriptive phase of the research, in the hypothesis tests, there was no acceptance of the established hypothesis. The facilitating conditions are characterized as a construct of great importance for the elderly population (Zaremohzzabieh et al., 2014). However, the sample collected was predominantly composed of young adults. This would explain why there is no relation between the importance that people attribute to having means of support when buying in m-commerce with the value that they perceive in purchases via mobile phones.

Pleasure was the construct used to formulate the seventh research hypothesis (H7). With pleasure as a predictor of perceived value, the test results supported the hypothesis, indicating a positive influence of the construct on the response variable. This result is in agreement with the study of Lee et al. (2003), in which a strong relationship was found between the construct and the purchases made on the internet.

The eighth research hypothesis (H8) was that ‘perceived security positively influenced perceived value’. The findings of the research converge with that of studies such as Oliveira et al. (2016) because the hypothesis was accepted. Thus, it is assumed that the greater the sense of security that people have in transmitting their information and banking data through mobile phones, the greater the perceived value.

The perceived cost is the ninth hypothesis of the research (H9) and the second to assume a relation of negative influence. That is, ‘the higher the perceived cost, the lower the value that the person perceives’ in m-commerce. The hypothesis was rejected. It is probable that because they knew the m-commerce, these people have a greater notion of the diversity of direct and indirect prices for the acquisition of goods and services in this environment; therefore, a negative influence between the construct and the response variable was reported.

The last hypothesis to be formulated (H10) was that the perceived value, which was being treated as predicted in previous hypotheses, now becomes a predictor of another variable, the intention to adopt. This hypothesis was confirmed, showing that the relationship between the constructs is solid and that it does not dissolve easily through sample differences.

Thus, seven out of the ten hypotheses were accepted in de general model. Figure 2 represents the final model of the research, based on the tests performed on the variables and the data collected.

Figure 2.
Final model proposed from the empirical tests of the research.
Source: Prepared by the authors, based on statistical tests (2017).