Trust and socio-demographic influences on mobile banking adoption in South Africa: a longitudinal study Open Access

Frank Knight’s seminal concept of uncertainty differentiates between measurable risk, where outcomes are probabilistic, and true uncertainty, where the future cannot be predicted or quantified (Knight, 1921). In the context of financial services, trust becomes critical under conditions of Knightian uncertainty, as it reflects an individual’s decision to rely on another party in situations where outcomes are inherently unpredictable (Bachmann and Inkpen, 2011). Trust has long been a cornerstone of the financial services industry, mainly due to the fiduciary relationships embedded within the sector (Devlin et al., 2015). Since the mid-20th century, retail banking has undergone substantial automation and digitalisation, spurring the emergence of new business models and practices (Arner et al., 2016; Bajwa et al., 2022; Bátiz-Lazo and Boyns, 2004; Bátiz-Lazo and Woldesenbet, 2006; Bátiz-Lazo and Wood, 2002).

The adoption of internet and mobile banking has further amplified the role of trust in retail finance (Dimitriadis and Kyrezis, 2010; Kosiba et al., 2020; Wang et al., 2015; Yousafzai et al., 2009), as trust remains integral to consumer adoption of these digital channels (Gefen et al., 2003; Lee, 2009; Pavlou, 2003; Tsang et al., 2004). However, while technological advancements such as mobile banking have improved efficiency, they have also heightened consumer perceptions of risk ^[1] (Zhou 2012), particularly where human interaction is reduced or eliminated, thereby complicating problem resolution (Baden-Fuller and Haefliger, 2013). The erosion of trust in large financial institutions, as discussed by Hurley et al. (2014), underscores the importance of restoring consumer confidence to facilitate digital banking adoption. Furthermore, Mukherjee and Nath (2003) highlight that trust is a crucial determinant of relationship banking in online environments, reinforcing its relevance in the shift towards digital financial services.

The Technology Acceptance Model (TAM) by Davis (1989) provides a widely accepted framework for understanding how perceived usefulness and ease of use influence the adoption of banking technologies. While our study does not directly employ TAM, it aligns with its principles by examining how trust influences mobile banking adoption. Unlike many TAM-based studies, which often rely on cross-sectional or short-term longitudinal data (e.g. Davis, 1989; Pikkarainen et al., 2004; Venkatesh et al., 2003), our research takes a ten-year longitudinal perspective, similar to Venkatesh and Davis’s (2000) approach. Moreover, TAM-based studies frequently emphasise technological factors while overlooking socio-demographic elements such as ethnicity and Living Standards Measures (LSM), which significantly influence mobile banking adoption. Our study addresses this gap by offering a perspective on mobile banking in South Africa, an underrepresented region in the literature compared to research conducted in the United States, China, and Europe.

Our findings align with studies emphasising the necessity for financial services providers to understand consumer intentions towards mobile banking (Isaeva et al., 2020; Wang et al., 2015). Furthermore, our research reinforces the argument that retail finance providers should prioritise building consumer trust to encourage the adoption of mobile banking and related financial technologies (Dimitriadis and Kyrezis, 2010; Kosiba et al., 2020). Prior research suggests that demographic characteristics such as age, ethnicity, and gender influence mobile banking preferences (Laukkanen and Pasanen, 2008; Malaquias et al., 2018; Malaquias and Hwang, 2019; Park et al., 2007). Expanding on this body of work, our study identifies the primary determinants of mobile banking adoption over an extended period, specifically examining whether individuals with trust in banks are more likely to adopt mobile banking than those without such trust, while considering the mediating role of socio-demographic factors such as ethnicity and LSM.

Our ten-year longitudinal study reveals that income and education levels significantly influence mobile banking usage, while ethnicity also plays a role. Specifically, Black ethnic groups in South Africa were more likely to use mobile banking than other demographic segments. Our findings further suggest that while trust in banks positively influences mobile banking adoption, its economic impact is relatively modest compared to other socio-demographic factors. Specifically, individuals with trust in banks were 1.5% more likely to use mobile banking than those without such trust.

The reminder of this article is structured as follows: Section 2 explores mobile banking usage and the role of trust in banking. Section 3 outlines the empirical approach and rationale for focusing on South Africa. Section 4 presents the results, and Section 5 concludes with potential policy implications.

Socio-demographic characteristics such as age, income, and ethnicity can influence mobile banking adoption. Table 1 summarises key socio-demographic variables and their impact on electronic/mobile banking adoption, highlighting the diversity of findings in the literature.

Education and income levels are consistently identified as positive predictors of electronic banking adoption (Kolodinsky et al., 2004; Lee et al., 2002). Younger individuals are more likely to adopt mobile banking (Koenig-Lewis et al., 2010; Laukkanen, 2016; Tsai et al., 2013), while gender differences reveal mixed findings. Women are often less likely to trust and adopt mobile banking (Laukkanen and Pasanen, 2008; Malaquias and Hwang, 2016), though some studies report no significant gender differences (Chawla and Joshi, 2018; Malaquias and Hwang, 2019). Married individuals are more likely to use electronic banking (Chawla and Joshi, 2018; Kolodinsky et al., 2004), but overall, findings on socio-demographic impacts remain inconclusive (Souiden et al., 2021). Hence, we posit:

The influence of culture on mobile banking adoption is an area that needs more research, as it has been explored in only a few studies (Souiden et al., 2021). According to Goodenough (1971), culture is a set of beliefs or standards shared by a group of people which influence an individual’s behaviour. As a dimension of culture, ethnicity involves a sense of belonging to a specific group, separate from society at large, based on shared beliefs and habits (Usunier and Lee, 2005). Individuals’ beliefs, values and behaviour echo the habits and practices of the cultural group to which they belong (Legoherel et al., 2009). In their study, Ting et al. (2016) found significant differences in mobile payment adoption across ethnic groups, suggesting that shared beliefs and habits influence behaviour (Usunier and Lee, 2005). Their findings suggest that there are significant differences between certain ethnic groups. Hence, we posit:

According to Mayer et al. (1995), trust in another party is an intentional choice influenced by an individual’s inherent tendency to be confident and their perception of the other party’s reliability. This relational aspect of trust is critical in financial services, where the intangibility of products and the fiduciary nature of many interactions heighten both perceived risk and uncertainty. With the growing adoption of electronic and mobile technologies, understanding how trust, privacy, and security beliefs affect behavioural intentions has become essential (Gefen et al., 2003; McKnight et al., 2002). Early research focused on trust in online banking, but more recent studies address trust in mobile banking, recognising the specific risks and uncertainties it presents (Wang et al., 2015). Al-Somali et al. (2009) examined online banking adoption in Saudi Arabia, highlighting the role of cultural and contextual factors, such as trust and perceived risk, in influencing acceptance. Perceived risk and trust are therefore interlinked concepts with trust mitigating some of the barriers to adopting online and mobile banking services (Kim et al., 2018). Table 2 summarises the impact of trust on mobile banking adoption, highlighting the contrasting findings in the literature.

Table 2 suggests there are two types of trust when assessing barriers to mobile banking adoption – trust in the institution (the mobile network operator and their agents providing the service) and trust in the technology (Koenig-Lewis et al., 2010). Institutional trust means that a person believes the necessary impersonal structures are in place to anticipate a successful future endeavour (McKnight et al., 1998). Trust in technology refers to trust that the software will reliably fulfil one’s needs, securely, dependably and consistently over time (McKnight, 2005).

Institutional trust refers to the development of confidence based on the structures and safeguards that institutions provide, which help mitigate perceived uncertainty in the business environment (Bachmann and Inkpen, 2011; Isaeva et al., 2020). Studies have explored individual factors affecting trust in banks. Studies have shown that trust in banks is linked to perceived asset safety and tends to be higher among younger individuals and those with higher incomes (Knell and Stix, 2015). Additionally, higher trust levels are observed among women and individuals who generally display a stronger tendency to trust others (Fungáčová et al., 2019). Institutional trust, however, is vulnerable to external shocks, as evidenced by declining trust in banks during financial crises (Jansen et al., 2015).

For many consumers, a lack of trust in financial institutions acts as a significant barrier to adopting financial services (Allen et al., 2016). Due to the intangible, complex, and often long-term nature of financial products, perceived risk is particularly high, especially in relation to savings and investment offerings (Ennew and Sekhon, 2007). To adopt mobile financial services, consumers first need to trust the institution offering these services before extending that trust to the technology itself (Rotchanakitumnuai and Speece, 2003).

Some research suggests, however, that pre-existing trust in banks does not necessarily translate to higher mobile banking adoption. For instance, the findings of Luo et al. (2010) found that pre-existing trust in banks had a limited impact on mobile banking adoption, as customers viewed mobile banking as inherently riskier due to the wireless nature of the platform, which banks partially control. Chemingui and Iallouna (2013) similarly observed that pre-existing trust in banks had little influence on the intention to adopt mobile banking, with resistance to change cited as a key barrier.

Conversely, other studies report that pre-existing trust in banks positively impacts mobile banking adoption. For instance, (Wang et al., 2015) found that institutional trust positively affects trust in online and mobile services provided by banks. Gu et al. (2009) suggest that customers who had pre-existing trust in banks are more likely to see the advantages of mobile banking and to adopt mobile banking. Similarly, Arvidsson (2014) reported that trust in banks is positively associated with the adoption of mobile banking and that trust in banks is a significant factor in determining the likelihood of adoption.

These studies collectively underscore the centrality of trust in understanding the adoption of technology and highlight the importance of perceived usefulness, ease of use, trust, and contextual factors in driving the adoption of digital banking technologies. Extant research, therefore, lead us to anticipate that there is a significant relationship between having trust in banks and the use of mobile banking and hence posit:

In what follows, the above three hypothesis are tested against a data for a ten-year period around the adoption of mobile banking in South Africa.

In emerging economies, mobile money has been used as a replacement for formal financial institutions and in several countries, there are more mobile money accounts than bank accounts (Awanis et al., 2022). Emerging economies, particularly in Sub-Saharan Africa (SSA), have pioneered the use of mobile and technological innovations to address financial inclusion (Rouse et al., 2023; Rouse and Verhoef, 2016). The high growth and penetration of mobile phones has resulted in increasingly affordable financial services and cost-effective means of providing financial services to the previously unbanked (Asongu, 2013).

There has been a rise in mobile phones and improved access to the Internet in South Africa. Thus, the improved access to mobile phones and mobile banking offerings combined with nearly 60% of the population under the age of 35 (Edelman Trust Institute, 2016) provides a conducive landscape for mobile banking to flourish. South African consumers have a variety of opportunities to use mobile banking services, both from the major banks and mobile network operators since 2005, providing rich data for the current study.

The data was obtained from the FinScope Household Surveys in South Africa. The FinMark Trust is responsible for the annual FinScope Household Survey which provides country-specific information on the use of financial services and consists of face-to-face interviews with individuals across a country (FinMark Trust, 2018). The FinScope Surveys are nationally representative and are benchmarked to mid-year population estimates provided by government agency Statistics South Africa (StatsSA). In the first stage of sample weighting, the primary sampling units (the enumerator areas) were selected with probability proportional to size, with the number of persons aged 16 years and older as a measure of size, from the population-sampling frame. In the second stage, households were systematically selected in each primary sampling unit. In the third stage, the design weights of the respondents were adjusted to compensate for differential non-responses (TNS, 2016).

Fieldwork utilised annual data from the FinScope Household Surveys. Data was available for the years 2006–2016, providing a robust longitudinal framework to capture both the initial and longer-term impacts of mobile banking adoption. This timeframe enabled an analysis of adoption patterns and behavioural shifts as mobile banking evolved from its nascent stages to a more established financial service.

The datasets included sampling weights to account for the differences in the ratio of the sample size to population size. The regressions were run including these survey weights. A descriptive approach was taken and thus these weights were used when calculating the logistic regressions (Cameron and Trivedi, 2009). The sample selections were specifically made to explore the target population, and as a consequence of using the weighted samples, the results can be interpreted as estimating the census coefficients (Solon et al., 2015). After removing missing values from all independent and control variables, the final sample consists of 11,762 observations. The dependent variable is mobile banking status, which is a dummy variable that equals one if the respondent has a mobile bank account and zero otherwise. The explanatory variables are shown in column 1 of Table 3.

The explanatory variables, with the exception of age, were either dichotomous or categorical and dummy variables were included. The standard dummy variable technique was applied, which used the first-mentioned category as a reference in the final equation. This approach allowed for coefficients, standard errors and significance levels to be generated. These outputs result from the predicted logit transformation of the probability of the presence of the dependent variable (Wentzel et al., 2016). In general, for n levels of a variable, n-1 dummies are included.

A number of socio-demographic variables were included, as studies have found that these factors affect the use of mobile banking. Age was included as it is a significant indicator of electronic banking use among those who are younger and are more likely to use electronic banking channels (Laukkanen, 2016; Tsai et al., 2013). Education and income levels also have a significant positive influence on electronic banking adoption (Kolodinsky et al., 2004; Sohail and Al-Jabri, 2014). Gender has been explored in several studies (Chawla and Joshi, 2018; Park et al., 2007; Sohail and Al-Jabri, 2014). The model includes a binary variable for gender (male or female) as the questionnaire included only two categories for gender. Broader categories could not be interpreted and was a limitation of this study. The model also includes a variable ethnicity (Ting et al., 2016). The marital status of individuals was included as this variable has been explored in other studies (Chawla and Joshi, 2018; Kolodinsky et al., 2004).

A variable for the Living Standards Measure (LSM) was also included as it is frequently used in marketing research and provides a further indicator of the poverty levels of individuals (South African Audience Research Foundation, 2017). The aim of LSMs is to divide the population into different “wealth” or “poverty” groups without the use of income data (von Maltzahn and Durrheim, 2008). The study could thus explore the relationship between the LSM and the use of mobile banking. The relationship between financial inclusion and the geographical area had been explored in several studies (Honohan and King, 2012; Klapper and Singer, 2015). This study included a binary variable indicating the location of the individual as living in an urban or rural area to establish whether the use of mobile banking differed depending on where individuals lived. The variable bank trust was included which is a dummy variable that equals one if the respondent has a trust in banks and zero otherwise.

A logistic regression model was run on the pooled sample (2006, 2010, 2012 and 2016). The datasets included sampling weights to account for the differences in the ratio of the sample size to population size. The regressions were run including these survey weights. We employ the following model:

The datasets for the individual survey years were pooled into a cross-section and the model included year and region dummies and the error term for individuals i for t periods. The pooled cross-sectional analysis allowed us to increase the number of observations and to assess the relationship across time (Wooldridge, 2009). In order to address any concerns on the heteroscedasticity and serial correlation in the error terms, the estimation included heteroscedasticity-robust standard errors, in line with White (1980).

The descriptive statistics are provided in Table 4 based on the unweighted sample. Women approximated 58% of the sample size and the majority of individuals surveyed were unmarried/not living with a partner. The surveys were conducted in both urban and rural areas, with approximately 75% living in urban areas. The majority of individuals reported a monthly income in the second income category (R1 – R1,999) and had either commenced or completed a high school education. There was a high proportion of respondents indicating that they trusted banks (an estimated 68% of the sample). The majority of the sample did not use mobile banking, with only 9.8% having a mobile banking account. The uptake of mobile banking in South Africa has been hindered by strict banking regulations, established financial retail infrastructure and a preference to use cash (Rouse and Verhoef, 2017).

The main results are presented in Table 5. The following variables were found to be significant and to influence the use of mobile banking: age, bank trust, education, ethnicity and income.

Regarding the other individual determinants of mobile banking use, the most significant factors were income and education. Those with higher income and more years of education were more likely to use mobile banking which is in line with our expectations and prior research (Kolodinsky et al., 2004; Laforet and Li, 2005). The results suggest that individuals in the highest income category increased the probability of using mobile banking by 9.44% compared to those in the lowest income category. Thus, confirming the importance of the relationship between education and mobile banking. Our results suggest that individuals with a tertiary education are 4.77% more likely to use mobile banking than those with primary school education.

The age of an individual was found to be significantly negatively associated with mobile banking, with those in the older age groups less likely to use mobile banking. This finding is in line with previous studies which indicate that older age groups are less likely to use electronic banking channels (Laukkanen, 2016; Laukkanen et al., 2007; Tsai et al., 2013). Consumers in the older age groups tend to resist new technologies, whilst the younger age groups are more likely to adopt mobile technologies. To determine whether this association was linear, the square term of age was included in the model. See Table 6.

After introducing the squared term on mobile banking use, evidence in support of non-linearity was found. The findings indicated that there was a non-linear relationship between age and mobile banking, as suggested by Kikulwe et al. (2014). As the age term was positive and its squared term was negative, this suggested a convex relationship. Other studies have found that, contrary to the traditional views of adoption and innovation theory, it is not the youngest age group that is adopting mobile banking (Laukkanen and Pasanen, 2008). The results of this study confirm this expectation. These results are also broadly in line with Sohail and Al-Jabri (2014), who conclude that mobile banking users are in the age category of 36–40 years, and Laforet and Li (2005), who suggest that users aged 35–44 years are the most likely to use mobile banking.

With respect to gender in the baseline model, there was a negative relationship with mobile banking use but the association was not significant. This is in line with the findings of Malaquias and Hwang (2019). Refer to Table 5. However, the negative association is broadly in line with studies suggesting that men are more likely to use electronic banking than women (Laforet and Li, 2005; Laukkanen and Pasanen, 2008; Sohail and Al-Jabri, 2014). The Living Standards Measure was only significant at the 1% level, with those in the LSM 3–4 category more likely to adopt mobile banking than those in the LSM 1–2 category. Refer to Table 5. Location and marital status also had no significant influence on the use of mobile banking. This study also explored the effects of time on mobile banking use and included a year dummy variable for each year of the survey.

We found a significant association between ethnicity and the use of mobile banking, whilst controlling for other factors, confirming H2. This finding is in line with that of Ting et al. (2016) which found a significant association between ethnicity and the intention to use a mobile payment system. The findings of our study suggest that Black individuals are more likely to adopt mobile banking than other ethnicities. Our findings also corroborate the results of the Consumer and Mobile Banking Services 2016 survey conducted in the United States of America which indicated that Black individuals (and other minorities) are more likely to adopt mobile banking services than White individuals (Board of Governors of the Federal Reserve System, 2016). Our findings further suggest that Asian individuals were 1.4% less likely to use mobile banking than Black individuals, whilst controlling for other factors. The marginal effects for other ethnicities were less than 1%, whilst controlling for other factors. Thus, the relative importance of ethnicity as a factor influencing the use of mobile banking appears to be low and that income and education remain the principal drivers for mobile banking use in South Africa.

The model also included the year of the survey as a variable. The results indicate that the year of the survey had a significantly positive effect on the use of mobile banking. In line with our expectations, the results suggest that consumers are more likely to use mobile banking in all survey years, compared to consumers in 2006. This could possibly be explained by technological innovations and the penetration of mobile phones in the South African market.

To summarise results so far, income and education emerged as critical determinants of mobile banking adoption, with higher income levels and tertiary education significantly increasing the likelihood of use. Individuals in the highest income category were 9.44% more likely to adopt mobile banking compared to those in the lowest income group, while those with tertiary education were 4.77% more likely to use mobile banking than those with only primary education. These findings align with prior research and underscore the persistent digital divide, where socio-economic status acts as a barrier to accessing digital financial services. This highlights the need for targeted interventions to enhance financial inclusion among lower-income and less-educated populations.

Ethnicity also played a significant role, with Black individuals more likely to adopt mobile banking compared to other ethnic groups. This finding is particularly relevant in the South African context, where historical inequalities and cultural diversity influence financial behaviour. Shared cultural beliefs and practices within ethnic groups may foster greater trust and familiarity with mobile banking technologies, offering valuable insights for banks and policymakers aiming to design inclusive financial products.

Results also revealed a non-linear relationship between age and mobile banking adoption, challenging the traditional assumption that younger individuals are the primary adopters of new technologies. Adoption rates peaked at around 40 years of age before declining, suggesting that middle-aged individuals, who often have greater financial responsibilities and resources, are more likely to adopt mobile banking. In contrast, older individuals exhibit resistance due to lower technological familiarity and higher perceived risk. This finding aligns with studies that identify middle-aged groups as key adopters of mobile banking, offering a more nuanced perspective on the role of age in technology adoption.

These differences in mobile banking adoption can be attributed to a combination of socio-economic, cultural, and psychological factors. Higher income and education levels provide individuals with greater access to technology and the skills needed to navigate digital platforms, reducing barriers to adoption. Cultural norms and shared experiences within ethnic groups may influence trust and familiarity with mobile banking, particularly in contexts where certain groups have historically been excluded from formal financial systems. Middle-aged individuals may adopt mobile banking due to greater financial responsibilities, while older individuals may resist due to lower technological literacy and higher perceived risk.

For banks and fintech companies, these findings highlight the need to tailor strategies to specific consumer segments. For instance, targeted financial literacy programmes could help bridge the gap for lower-income and less-educated individuals, while culturally sensitive marketing campaigns could enhance trust and adoption among diverse ethnic groups. Policymakers can use these insights to design interventions that address socio-economic disparities and promote financial inclusion.

Trust in banks was found to be significantly positively associated with the use of mobile banking. These findings are consistent with the expectations of H3. The results suggest that when consumers trust banks, they will perceive mobile banking as useful and are willing to use it. These results confirm the contention that trust in banks has a positive relation to the use of electronic/mobile banking (Arvidsson, 2014; Gu et al., 2009; Montazemi and Qahri-Saremi, 2015). However, this finding is in contrast to studies suggesting that pre-existing trust in banks does not have a significant influence on the use of mobile banking (Chemingui and Iallouna, 2013; Luo et al., 2010). A possible explanation is that Luo et al. (2010) study was conducted on early adopters of mobile banking and resulting in a higher risk perception and the lack of transfer of pre-existing trust.

In order to explore the relationship between bank trust and the use of mobile banking further, the marginal effects were calculated. The economic magnitude of the effect of bank trust on the use of mobile banking was low. The relative importance of trust in banks as a factor influencing mobile banking use was lower than expected. A possible explanation may be due to differences in sampling strategy. In the current study, respondents were randomly selected from the total population whilst in contrast, the respondents in the Arvidsson’s (2014) and Luo et al.’s (2010) studies were sampled from the population of early adopters of mobile banking. Thus, the results from a study of the general population may differ, as prior studies have shown that the factors influencing early adopters differ from late adopters of mobile banking (Kim et al., 2010; Laukkanen and Pasanen, 2008).

To summarise, trust in banks reduces perceived risk and uncertainty, making individuals more willing to adopt mobile banking, though its impact may be moderated by other factors such as technological familiarity and cultural context. This aligns with studies highlighting the importance of institutional trust in mitigating perceived risks associated with digital financial services. However, the study also acknowledges the mixed findings in the literature, suggesting that trust alone may not be sufficient to drive adoption without addressing other barriers such as technological literacy and perceived risk.

The model specification was modified to assess whether the model was robust to changes in the sample specification. The samples from the baseline model were split into two samples – those who had trust in banks and those who did not. The main variables of interest remained significant, except for ethnicity. In the sub-sample of those who did not trust banks, the role of ethnicity was no longer significant whereas for those who did trust banks, the relationship was significant. There was also a change in the relationship between age and mobile banking. This relationship was negative for those who trusted banks and positive for those who did not. This finding suggests that if a person does not have trust in banks, and the older they get, the more likely they will be to use mobile banking. The opposite would be true for those who did trust banks. A possible explanation for this finding is that a person who does not trust banks may prefer using the technology to perform their banking transactions, as it does not require any interactions with bank staff or the need to go to a bank branch. Future research could explore this relationship further. The results are presented in Table 7 and support the baseline model.

In order to assess the robustness of the model to the region effects and the effect of the period of the samples, the interaction term of year and region was included. The primary relationships of interest remained significant with the appropriate signs. The results are presented in Table 8 and the results support the baseline model.

The model was run without applying the sampling weights and heteroscedasticity-robust standard errors (Solon et al., 2015) were reported. The results of the unweighted sample had a lower pseudo-r squared of 0.2842, indicating that this model had a lower explanatory power compared to the baseline model. However, the robust standard errors for several variables were lower, indicating that the unweighted estimation was more precise. The results are presented in Table 9 and are consistent with the baseline model.

This study underscores the significant role of socio-demographic factors and ethnicity in shaping mobile banking adoption, particularly in the context of financial inclusion. While digital banking innovations offer considerable potential for expanding financial access, their adoption is not uniform across different population segments. Our findings indicate that income and education levels are strongly associated with mobile banking use, with higher-income individuals and those with tertiary education exhibiting greater adoption rates. This highlights a persistent digital divide and the need for targeted interventions to support lower-income and less-educated groups.

A key contribution of this study is its focus on ethnicity, an often-overlooked factor in mobile banking research. The findings demonstrate that individuals from Black ethnic backgrounds are more likely to adopt mobile banking than other groups, a pattern that carries important implications in the South African context. Historical inequalities and cultural diversity shape financial behaviour, and mobile banking appears to offer a pathway to formal financial services for previously unbanked populations. However, differences in adoption rates across ethnicities highlight the necessity of culturally sensitive approaches to product design and marketing. Understanding community-specific trust dynamics and financial habits can enhance the effectiveness of mobile banking solutions, ensuring that they align with users’ expectations and needs.

The study further reveals a non-linear relationship between age and mobile banking adoption, with adoption rates peaking at around 40 years of age before declining. This finding challenges the assumption that younger consumers are the primary adopters of digital financial services. Instead, middle-aged individuals, who typically have greater financial responsibilities and resources, emerge as the most engaged users. Older individuals, by contrast, exhibit lower adoption rates due to limited technological familiarity and heightened perceptions of risk. These insights suggest that financial institutions should develop tailored strategies to engage younger users by leveraging their technological proficiency, while also addressing the concerns of older consumers through user-friendly interfaces and financial education initiatives.

Trust emerged as a critical factor in mobile banking adoption, with pre-existing confidence in banks positively influencing consumers’ willingness to engage with digital services. However, trust alone is insufficient; technological literacy and risk perceptions also play a mediating role. Financial institutions can build on established trust through awareness campaigns that highlight the reliability and security of digital banking, while also addressing concerns about usability and fraud.

These findings offer valuable insights for financial institutions and policymakers seeking to enhance financial inclusion through mobile banking. Banks and fintech companies should prioritise financial literacy initiatives tailored to lower-income and less-educated populations, while also incorporating culturally sensitive marketing strategies to reach diverse ethnic groups. Policymakers can support these efforts by designing interventions that address socio-economic disparities, ensuring equitable access to digital financial services. Partnerships between financial institutions, governments, and community organisations could further facilitate the delivery of tailored financial education programmes, enhancing financial confidence and digital capability among underserved populations.

By integrating socio-demographic and trust-related factors into the analysis, this study contributes to a deeper understanding of mobile banking adoption and its implications for financial inclusion. It highlights both the opportunities and challenges presented by digital financial services in emerging markets, where economic inequalities and cultural diversity shape banking behaviours. As financial institutions continue to expand digital offerings, adopting nuanced, consumer-centred strategies will be essential to maximising mobile banking’s potential as a tool for economic empowerment and financial accessibility.

The authors would like to thank the FinMark Trust for providing access to the FinScope Survey data. The usual disclaimers apply. Bernardo Batiz-Lazo appreciates financial support through the Spanish State Agency for Innovation (AEI) grant PID2022-139315OB-I00, funded by MICIU/AEI/10.13039/501100011033, FEDER/UE A way of making Europe. Santiago Carbo-Valverde gratefully acknowledges financial support from research grants numbered PGC2018-099415-B-00 and PID2020-118883GB-I00 from the Spanish Ministry of Science and Innovation.