1. INTRODUCTION

Digital transformation has permeated many aspects of contemporary society, and the banking sector is not exception (González Núñez & Mariné Osorio, 2021). The digitization of financial services is revolutionizing the way consumers interact with banking institutions, altering their consumption processes, as well as their habits and expectations (Kitsios et al., 2021). Digital transformation in the financial sphere has been characterized by the adoption of technology such as artificial intelligence, data analytics, cloud computing, and mobility (Camoiras-Rodríguez & Varela-Neira, 2020), innovations that have enabled financial institutions to offer more efficient, personalized, and accessible services (Tsindeliani et al., 2022); however, this change not only involves the modernization of technological infrastructure in banks but also a cultural and operational adaptation in their administrative processes (Alam et al., 2023).

Contemporary financial consumers seek effective, efficient, convenient, and secure banking experiences, scenarios in which digital transformation has responded to these demands with the introduction of online services, mobile applications, electronic payments, and virtual assistants (López-Rodríguez & Cardozo-Munar, 2023). Access to real-time financial information and the ability to conduct transactions anywhere has led to a shift not only in customer expectations but also in their consumption dynamics (Kaur et al., 2021).

Despite the evident benefits, Filotto et al. (2021) have stated that the adoption of digital transformation in the banking sector faces significant challenges. The digital divide, cybersecurity, and resistance to change are obstacles that must be overcome to ensure a successful transition (Cichosz et al., 2020). Consumer trust in the security of online transactions and data protection is a critical aspect that financial institutions must address to embrace the ever-growing demand for digital services (Kitsios et al., 2021). This process should not be entrusted to all market segments, as some may view digital transformation as a barrier, which would not only hinder the consumption of emerging banking services but also affect the brand value of banks (López-Rodríguez & Sandoval-Escobar, 2023; Calvo Dopico et al., 2021).

Digital transformation has influenced how consumers manage their finances, primarily due to the preference for online services, a trend that has reduced reliance on brick-and-mortar branches and has led to a shift towards digital banking models (Tsindeliani et al., 2022). The customization of supply and the ability to predict individual financial needs are aspects that have redefined the relationship between consumers and banking institutions (Louw & Nieuwenhuizen, 2020), a process that should be focused on customer experience as a key differentiator in the digital era (Monroy-Perdomo et al., 2022).

Digital transformation not only requires consumers to adapt to new technologies but also to understand how to use them safely and effectively by promoting financial education, an essential scenario to empower users and ensure that they make informed decisions in the digital environment (López-Rodríguez & López-Ordoñez, 2022). This consolidates technology acceptance, which is the ability to willingly receive and approve what is offered among all available forms of technology (Venkatesh et al., 2003); however, this acceptance has also revealed a generational gap when it comes to adopting digital services in the banking sector. While younger generations tend to quickly embrace new technology, older customers may be reluctant to do so (Parida et al., 2016; Laukkanen, 2016; Khanra et al., 2021), due to both psychological and functional barriers (Chaouali & Souiden, 2019).

For this reason, this study has been based on the Unified Theory of Acceptance and Use of Technology (UTAUT), developed by Venkatesh et al. (2003), which provides a solid framework that sheds light on which factors influence the adoption of new technology in the financial domain, particularly in the banking sector. Challenges to be tackled regarding technological acceptance for the banking sector are greater integration of advanced technology and a deeper understanding of the psychological and functional factors that can affect consumers when deciding on whether to accept it.

Based on the above, this study addresses the following research question: How do the elements of the UTAUT sway the behavioral intention of account holders for technology use and their actual consumption in the Colombian banking sector? By providing an answer to this question, this study aims to enhance the understanding of various key aspects that the banking sector must consider for its digital transformation to be consolidated considering consumer behavior, thereby strengthening the market orientation of these entities. The next two sections present a literature review accompanied by the formulation of relevant hypotheses that support the validity of the research. After these two sections, the results of the empirical exercise are presented, followed by the Conclusions section, which includes theoretical and practical implications, limitations, and future research directions emerging from this study.

2. LITERATURE REVIEW

It is important to acknowledge that technology has enabled greater efficiency and convenience in the provision of banking services (Bhatt, 2021; Chaouali & Souiden, 2019). From the automation of administrative processes to the development of online payment systems, technology has simplified and expedited many everyday banking operations (Laukkanen, 2016), allowing consumers to have better user experiences. However, empirical research has identified that there are barriers to technology acceptance among customers (Venkatesh et al., 2003).

The increasing reliance on technology in banking services also presents significant challenges, not only for institutions but also for consumers (Leesakul et al., 2022). In this regard, there is a risk of financial exclusion for those who lack access to technology or do not know how to use it (Chaouali & Souiden, 2019). As banks close physical branches in favor of online services, those with no internet access or who are not comfortable conducting transactions online may be left behind (Laukkanen, 2016). This can widen the digital and financial divide, leaving some segments of society without access to essential banking services (Chung & Liang, 2020).

Colombian financial consumers find themselves immersed in an environment of transformation, marked by the convergence of technology and financial services. Due to the diversity of profiles, ranging from individuals with limited access to financial services to those with a high degree of technological literacy and familiarity, banking institutions that incorporate customer relationship management and market orientation into their corporate philosophy must understand consumer behavior to adapt their offerings to actual demand (Schulze et al., 2022). With the increasing market penetration of the internet and the proliferation of mobile devices, more and more customers are turning to mobile applications and online platforms to make transfers, check balances, pay bills, and perform other transactions (Hossain et al., 2020).

Despite this trend toward digitization, certain preferences for human contact and personalized advice persist among Colombian financial consumers, such as in the case of the elderly segment, which is becoming increasingly important due to the investment in the population pyramid, not only regionally but also globally (López-Rodríguez et al., 2024). This duality in the preferences of these customers presents a challenge for banking institutions, which must balance digital service offers with personalized attention (Martínez Jiménez, 2021). For this reason, the acceptance and use of technology by the consumer in the Colombian banking sector needs to be analyzed.

UTAUT was proposed by Venkatesh et al. (2003) to integrate various prominent technology adoption theories. These include the Technology Acceptance Model (Davis, 1989), the Extended Technology Acceptance Model (Venkatesh & Davis, 2000), the Revised Technology Acceptance Model (Venkatesh & Bala, 2008), and the Socio-Cognitive Technology Acceptance Model (Thompson et al., 1994), among others.

The idea of the UTAUT theory is to explain and predict the acceptance and use behavior of technology in various contexts (Venkatesh et al., 2016). It consists of four key factors: performance expectancy, effort expectancy, social influence, and facilitating conditions. Additionally, it includes four moderating factors: age, gender, experience, and voluntariness, which are related to predicting the behavioral intention to use technology and actual technology use, mainly in organizational contexts (Venkatesh et al., 2016).

Performance expectancy, as defined by Fernández Morales et al. (2015), refers to the level of benefit that users believe that they will reap from using an information system. According to Puspitasari et al. (2019), it also corresponds the degree to which using certain types of technology offer benefits when certain activities are carried out, based on the perception of how the technology will perform with a probabilistic and subjective estimation, involving associating a specific action or performance with one or more outcomes (Yu et al., 2021). In this regard, it is proposed that:

Effort expectancy represents the subjective and probabilistic relationship between one`s effort and the likelihood of one`s successfully performing an action or task (Maillet et al., 2015). According to Morales et al. (2015), it is also elated to the usability of a particular system. This component supports the consumer's perception regarding how easy or difficult it is to use technology in consumption processes, or to learn how to use it, which, in turn, impacts a user’s intentions (Puspitasari et al., 2019). Based on the above, it is proposed that:

For Maillet et al. (2015), social influence is used to measure the impact of perceived influence from a user's social circle, whether it be friends or family, on the use of technology. According to Fernández Morales et al. (2015), it is also referred to the extent to which an individual perceives the use of an information system as important to those around him/her. It is one’s ability to develop or change one’s mind, feelings, attitudes or behavior because of interaction with others that is perceived as being similar (Yu et al., 2021). Therefore, it is proposed that:

Facilitating conditions represent consumers perceptions of available resources and customer to carry out the actions they wish to perform, focusing on access to the resources they need to use new technology (Al-Mamary, 2022). Puspitasari et al. (2019) have indicated that these are supportive criteria reflecting the extent to which an individual perceives that the organization is being able to handle new technology. They can be considered objective factors that allow consumers to perform the tasks that they need (Almogren, 2022). In this regard, it is proposed that:

Behavioral intention to use is a key component that represents the willingness of users to incorporate technology into their consumption processes, so it is a fundamental predictor of actual use of technology behavior (Chao, 2019); therefore, the higher the intention to use technology, the greater the likelihood that it will be coupled and used Venkatesh et al., 2016).

It is pertinent to highlight that while usage intention predicts users’ initial willingness to adopt technology, actual usage reflects the practical and sustained implementation of technology in their daily lives (Chao, 2019); hence, actual use o technology refers to when, in real situations, they have adopted it in their consumption processes (Williams et al., 2015), making it a measure of effective and continuous technology use in their everyday lives (Arfi et al., 2021).

The relationship between behavioral intention and consumers’ actual use of technology is a phenomenon that has been studied in various fields, including psychology, sociology, and marketing. It refers to the disparity between what people say they will do or think they will do regarding the coupled and use of a type of technology, and what they do (Chan et al., 2023; Leesakul et al., 2022). This is known as perceived complexity, which emerges when they intend to use a new kind of technology then ultimately perceive it to be too complex to use (Chung & Liang, 2020).

Another aspect related to behavioral intention and actual use of technology is framed within barriers of resistance to technological innovation (Laukkanen, 2016). The challenges of technology acceptance which the banking sector seeks to overcome are greater integration of advanced technology and a deeper understanding of the psychological and functional factors that convince consumers to adopt it (Chaouali & Souiden, 2019). While functional barriers are represented by various obstacles that may hinder the use of new technology by individuals in consumption processes (Antioco & Kleijnen, 2010), psychological barriers refer to cognitive and motivational processes that prevent specific action or behavior modification from occurring (Bercht, 2019).

Taking the above into account, Figure 1 presents the theoretical model addressed in this research.

Figure 1Proposed theoretical model 

3. MATERIALS AND METHODS

This body of research has been consolidated through an empirical study that adopted a quantitative, descriptive, and causal approach. To achieve the study objectives, an exploratory factor analysis (EFA) and a confirmatory factor analysis (CFA) were used, later complemented with a Structural Equation Modeling (SEM) framework, following the recommendations of Hair et al. (2021).

4. RESULTS

To identify the underlying structure of the variables in the instrument used in the study, an EFA was conducted, including both the dependent variables (BI, ATU) and the independent variables (PE, EE, SI, FC), so that each factor could be revealed separately. Bartlett's test was analyzed, which was observed to be statistically significant ( ${\chi }^2$ = 2783.254, p < 0.001), allowing for the consideration of factorial extraction. The KMO test was performed to whether the total sample was suitable enough to support the relevance of this analysis, yielding a value of 0.922, as shown in Table 2.

Regarding the RMSEA, which explains the amount of unexplained variance, a value of 0.071 was obtained, indicating a good fit for the data. The TLI, which compares the fit of the proposed model to a null model in degrees of freedom, reached a value of 0.921, identifying a very good fit for it. As for the CFI, which evaluates the goodness of fit of a model to the observed data, the result was 0.965, determining an excellent fit of the model.

Based on the solution provided by the EFA, which utilized principal axis factoring with a Varimax rotation, five latent factors were revealed, explaining 65.7% of the cumulative variance. This finding suggests a complex yet significant underlying structure in the data, providing a solid foundation for the relationships between dependent and independent variables to be better understood.

A combination of the two dependent variables was included, resulting in an initial factor termed "behavioral intention and actual use of technology (BIA)," which accounted for 19.2% of the explained variance, comprising six items. This factor focuses on behavioral intention and actual use of technology.

Regarding the independent variables, PE was the second factor, explaining 16.7% of the variance, composed of four items. The third factor, FC, accounted for 10.9% of the variance, comprising four items. EE was the fourth factor, justifying 9.1% of the variance, with three items. Finally, the fifth factor, SI, attributed 8.2% of the variance and consisted of three items. The factor loadings of the five emerging latent factors are presented in Table 3

To support the results obtained in the EFA, a CFA was conducted. This was done to address its weaknesses (Thompson, 2004) and maintain the phases of structural model construction (Escobedo et al., 2016). The maximum likelihood factorial method (ML) was employed, which allows for estimating the model parameters in such a way as to maximize the probability of observing the actual data based on those parameters. (Liang & Luo, 2020).

Table 4 presents the overall fit indices of the CFA, allowing the fit of the proposed model with the observed data to be evaluated and its underlying structure to be better understood, with a favorable outcome found in this study. The results indicate that the observed factors of the test confirmed the theoretical factorial structure, which subsequently enabled the development of the structural model.

The data from the latent variables and their respective items showed a general Omega Coefficient (ω) of 0.936, and an Alpha Coefficient (α) of 0.930. Table 5 presents the same indicators independently for each of the resulting latent variables from the EFA, along with the unstandardized factor loadings for each item, the KMO test, and the standard error, revealing satisfactory results for each subscale.

For the final stage of the study, and to construct and evaluate the structural model, the ML estimation method was used. The first element considered for the model fit to be evaluated was the chi-square to degrees of freedom ratio ( ${\chi }^2$/df) where the following results were obtained: ${\chi }^2$ (163.940) / df (94) = 1.74, meeting the parameters proposed by Escobedo et al. (2015), who stated that values below 3 are considered acceptable. Table 6 presents other fit indices for the proposed structural model to be analyzed, which were favorable for the empirical exercise, while Figure 2 displays the structural model of the UTAUT Theory in users’ behavioral intention and actual use of technology in the Colombian banking sector.

Figure 2Structural model of the UTAUT theory in users’ behavioral intention and actual use of technology in the Colombian banking sector. 

The integration of "behavioral intention" and "actual use of technology" as a specific construct (dependent variable) in this research implies having a grasp of the dynamic relationship between consumers’ willingness to adopt the technology and their actual behavior when interacting with it in relation with products in the Colombian banking sector. This relationship is fundamental for understanding the technology adoption process and provides valuable information for designing effective implementation and promotion strategies for technology in the banking sector.

Table 7 presents the hypothesis test, where two out of the four proposals were accepted. Regarding H1 and H3, it was inferred that performance expectancy and social influence did not predict users’ behavioral intention and actual use of technology in the Colombian banking sector and were, therefore, rejected. H2 and H4 were accepted because effort expectancy and facilitating conditions did predict them.

5. CONCLUSIONS

The objective of this article was to provide empirical evidence of the relationship between the elements of the UTAUT, with customers’ behavioral intention to use technology and their actual use of it in the banking sector. The results showed that both effort expectancy, referring to an individual's perception of the amount of effort required to adopt and use new technology, and facilitating conditions, representing the extent to which consumers perceive that resources exist to facilitate their experience with technology, predict this theory when it comes to banking. However, performance expectancy and social influence as relevant components of UTAUT neither had a significant impact as a predictor on the behavioral intention of consumers to real use technology.

These results have significant implications for the commercial management of banking entities, particularly in terms of strengthening their market orientation. This challenge for contemporary organizations requires the consolidation of integrated marketing, understood as the ability to integrate other management areas into the marketing function, primarily to collectively meet customer expectations. The findings align with previous research indicating that the UTAUT model is suitable for evaluating technology adoption, intention, and consumption in the banking sector (Patil et al., 2020; Samartha et al., 2022), as well as for studying customer satisfaction (Marinković et al., 2020; Bhatt, 2021).

Examining the links between the UTAUT theory and behavioral intention and actual use of technology reaffirms the importance not only of intention in consumer adoption studies in the banking sector but also of usage behavior. This is a distinguishing factor of this research since, unlike most studies that have focused on intention, it emphasizes these two elements as dependent variables, Called Behavioral Intention and Actual Use of Technology," when EFA was applied.