Financial technology, or FinTech, denotes a collection of digital advances employed to deliver, enhance, or transform financial services. A comprehensive definition is elusive (Schueffel 2017); however, consensus exists that FinTech amalgamates the financial sector with technologies including mobile platforms, artificial intelligence, cloud computing and blockchain (Leong & Sung 2018; Tritto, He & Junaedi 2020). Researchers characterise it as a multidisciplinary domain and an expeditiously expanding business, marked by advancements in loans, payments, wealth management and insurance (Iwashita 2022; Ryu 2018; Khahharovna & Jumaboevich 2022). Buckley, Arner and Barberis (2016) define FinTech as the convergence of finance and information technology, whereas Vijai (2019) emphasises its capacity to deliver innovative financial solutions resulting from technical advancements.

The expansion of FinTech was accelerated by the 2008 global financial crisis, which exposed the vulnerabilities of traditional financial institutions and necessitated the search for more adaptable, user-centric alternatives (Imerman & Fabozzi 2020; Serbulova 2021). The proliferation of smartphones and enhanced internet access, particularly in emerging nations, has accelerated this trend, hence promoting the dissemination of mobile financial services (Haddad & Hornuf 2019; Shin & Choi 2019). The M-Pesa scenario in Kenya illustrates that digital financial solutions can rapidly develop in response to unaddressed financial demands (Bateman 2020; Kheira 2021). The latter is characterised by contextual characteristics such as infrastructure quality, digital literacy, affordability and customer trust (Abdullah et al. 2018; Majib et al. 2021). The COVID-19 pandemic accelerated the development of FinTech as consumers and businesses sought contactless and remote financial services (Aysan & Nanaeva 2022; Hasan, Ashfaq & Shao 2021).

FinTech is globally categorised into several segments, including payments, lending, InsurTech, crowdfunding and digital advisory services (Imerman & Fabozzi 2020; National Treasury 2019). In developing economies, digital payments and alternative lending have become the most dynamic sectors, indicating a pressing demand for accessible, cost-effective financial instruments (Coffie & Hongjiang 2023; Kyari & Akinwale 2020). Academics contend that an effective FinTech ecosystem relies on cooperation among regulators, innovators, infrastructure providers and end-users (Lee & Shin 2018; Muthukannan, Priyadharshini & Arul 2021). Although FinTech has revolutionised access to financial services, it has simultaneously engendered worries of over-indebtedness, data privacy and digital exclusion, particularly in low-income countries (Bateman 2020; Branzoli & Supino 2020).

South Africa has solidified its status as a principal FinTech hub in Africa, alongside Kenya and Nigeria, characterised by a swiftly expanding start-up and digital financial service provider ecosystem. The national FinTech landscape encompasses several areas, including digital payments, mobile lending, savings and deposits, InsurTech and investment platforms (National Treasury 2019). Yoco, PayFast, Lulalend, TymeBank and Vault22 (formerly 22seven) are prominent providers of mobile-first, user-centric platforms tailored to the requirements of SMEs and underbanked consumers (Ndung’u 2022; Pollio & Cirolia 2022). Cape Town has been identified as a continental FinTech powerhouse, bolstered by a dynamic start-up ecosystem, a robust banking system and a growing information technology (IT) workforce.

Despite ongoing developments, the FinTech sector in South Africa is encountering structural and regulatory challenges. The absence of digital literacy, infrastructural deficiencies and persistent scepticism towards digital platforms hinder widespread adoption, especially among informal and rural enterprises (Arendse & Van Den Berg 2024; Molosiwa & Holland 2025). Furthermore, the legacy systems of traditional banks hinder the speed of digitalisation, prompting numerous institutions to form partnerships with FinTech start-ups to maintain competitiveness (KPMG 2021; Wilson 2017). Consequently, the SARB and the IFWG have developed policies aimed at fostering innovation while safeguarding financial stability. These pertain to Sandbox initiatives and regulatory structures for digital payments, crypto-assets and digital credit (National Treasury 2019).

South African banks are adopting FinTech, frequently through internal initiatives or collaborations with digitally focused firms (Coetzee 2018). While some FinTech entities are still perceived as disruptors, many are now regarded as complements to traditional financial institutions by providing customer-centric services, alternative credit assessments and expedited onboarding processes (Berg et al. 2022; Zalan & Toufaily 2017). Nonetheless, concerns persist that accessible financing via FinTech platforms may promote consumer over-indebtedness (Bateman 2020).

The South African FinTech sector is progressing towards more inclusion and innovation. The synergy of mobile penetration, a conducive regulatory framework and rising consumer demand enables the nation to spearhead digital financial change in the area. To guarantee equitable access and sustainable growth, stakeholders must persist in tackling digital exclusion, infrastructure constraints and consumer protection.

Despite the existence of numerous theories and methods for delineating and comprehending technology usage, the TAM is esteemed for its efficacy in examining technology adoption (Slazus 2022). This study employed the TAM to evaluate the influence of usefulness and ease of use of FinTech products and services on their adoption of FinTech. The TAM is extensively employed to assess technology adoption (Upadhyay et al. 2022). It is based on the two principles of perceived usefulness (PU) and perceived ease of use (PEOU), which are believed to affect attitude towards technology adoption (Ullah et al. 2021).

Perceived usefulness, characterised as the belief that technology can improve job performance (Tun-Pin et al. 2019; Venkatesh & Davis 2000), may be a crucial factor in FinTech adoption (FINAD) by SMEs. After all, customers, for instance, are more inclined to embrace FinTech solutions when they view these technologies as helpful in enhancing performance, conserving time and providing concrete advantages (Alwi et al. 2019; Owusu et al. 2021). The findings of studies undertaken by Alalwan et al. (2018) as well as Jantarakolica and Jantarakolica (2018) suggest that PU is a factor in the adoption of FinTech as it affects initial acceptance and sustained usage (Daragmeh, Sági & Zéman 2021). Empirical evidence from Indonesia and Malaysia also highlights the predictive ability of PU in facilitating FINAD (Mahmud, Joarder & Muheymin-Us-Sakib 2022; Nugraha et al. 2022).

However, alternate positions exist, as evidenced by research such as Rattanaburi and Vongurai (2021) as well as Brown (2002), which indicates that no significant correlation exists between PU and technology adoption, especially in developing economies or specialised applications. Interestingly, Wallace and Sheetz (2014) have challenged the unidimensional aspect of PU, contending that it may not comprehensively encompass users’ interpretations of ‘useful’ in various settings. Nonetheless, PU continues to be a frequently referenced concept in elucidating technology adoption, as it not only encourages adoption but also promotes ongoing engagement with FinTech solutions (Singh, Sahni & Kovid 2020; Slazus 2022). Chen, Xu and Arpan (2017) and Chang, Hajiyev and Su (2017) established the substantial influence of PU on behavioural intention (BI) and actual usage, whereas Jerene and Sharma (2020) and Majid (2021) highlighted that users are more inclined to embrace FinTech when they recognise concrete advantages in efficiency, cost reduction, or decision-making assistance. Mindful of this and other scholarly positions stemming from studies undertaken in different economic and geographical contexts, this study anticipates that:

Perceived ease of use, a fundamental element of the TAM, denotes the extent to which consumers view that the utilisation of a specific technology-based solution requires minimum effort (Venkatesh & Davis 2000). Perceived ease of use tends to affect the adoption of FinTech by diminishing the perceived effort necessary to utilise technological services, thereby motivating prospective users (Tun-Pin et al. 2019). Tiong (2020) notes that PEOU is a primary predictor of FINAD in Malaysia. Similarly, Jantarakolica and Jantarakolica (2018) identified PEOU as a crucial factor in FinTech adoption, particularly among consumers with minimal technological proficiency. This was further corroborated by Almashhadani et al. (2023), who asserted that PEOU is more critical to users who are inexperienced in using technology. Correspondingly, Nugraha et al. (2022) found that PEOU was a significant determinant of FinTech adoption. Moreover, PEOU frequently corresponds with other adoption characteristics, including trust and perceived complexity (Isiaka, Adeosun & Okewale 2022; Owusu et al. 2021). Streamlined FinTech solutions generally improve user experiences, thereby fostering adoption (Buckley & Webster 2016; Dinh, Nguyen & Nguyen 2018). Nonetheless, Alwi et al. (2019) contend that PEOU may not consistently demonstrate statistical significance with respect to having a relationship with technology adoption, as this is contingent upon the context. This notwithstanding, there is a degree of consensus in extant literature that user-friendly FinTech solutions enhance acceptance, particularly among inexperienced users, as simplicity diminishes cognitive load and stimulates interest in the technology (Al Nawayseh 2020; Matsepe & Van Der Lingen 2022). Furthermore, Khan, Yee and Gan (2021) and Abbasi et al. (2021) both established that PEOU and PU strongly influence the desire to utilise peer-to-peer (P2P) lending platforms. Chen et al. (2017) similarly discovered that both PEOU and PU strongly influence BI, with both parameters indirectly affecting actual usage through intention. Nonetheless, certain research contests the generality of these correlations. Granić and Marangunić (2019) discovered that PEOU did not significantly affect BI or actual usage, emphasising contextual factors such as user demographics, digital preparedness and platform complexity. In the South African SME setting, where numerous business owners may possess limited digital abilities, the simplicity and user-friendliness of FinTech systems are essential for uptake. Tun-Pin et al. (2019) and Noviyanti and Erawati (2021) observed that PEOU is most impactful when combined with personal innovativeness, perceived enjoyment and efficacy. It is against this backdrop that this study hypothesises that:

Since its inception, the TAM model has been revised and expanded by different academics (see Fedorko, Bacik & Gavurova 2018; Ooi & Tan 2016; Venkatesh, Thong & Xu 2016). However, this study is interested in interrogating the possible correlation between FINAD and organisational performance (OP). For SMEs, performance is frequently assessed by financial metrics such as profitability and operational efficiency and with non-financial factors such as decision-making quality and customer experience (Isiaka et al. 2022). FinTech adoption can be crucial for enhancing organisational performance through improved decision-making, operational efficiency and profitability (Chen et al. 2021; Matsepe & Van Der Lingen 2022).

FinTech adoption enables organisations to develop sustainable business models and achieve competitive advantages, therefore improving overall performance (Pizzi, Corbo & Caputo 2021; Subanidja et al. 2022). Research indicates that FINAD enhances financial performance, with Nigerian banks reporting a 5% rise in profitability (Kyari & Akinwale 2020). The advantages of FinTech, such as increased transaction speed, superior user experience and higher efficiency, are likely to engender greater productivity and operational effectiveness (AlBar & Hoque 2019; Marei et al. 2023). Moreover, FinTech also tends to broaden the accessibility of financial services, especially in developing nations, hence fostering organisational growth (Setiawan et al. 2021). Furthermore, research conducted by Kumar, Kohli and Kashyap (2023) and Darmansyah et al. (2020) demonstrates that SMEs utilising FinTech tools such as digital payments, cloud-based accounting, or online lending experience enhancements in operational efficiency, customer service and financial decision-making. These solutions not only optimise everyday processes but also enhance financial transparency and access to working capital, hence improving the overall agility and competitiveness of SMEs. It is noteworthy, though, that these literary positions stem from studies in different contexts, and so their findings are unlikely to accurately reflect what the nature of the relationship between FINAD and OP might be in the case of SMEs in South Africa. Nonetheless, positions canvassed in prior research encourage the study to deductively project that:

Reliant on the hypotheses of the study, which have been informed by due consideration for different scholarly positions as it pertains to plausible relationships between the study’s independent and dependent variables, the research model for the study was developed and is illustrated in Figure 1.

The study employed scales utilised in prior studies by Venkatesh (2000) as well as Gholami, Abdekhoda and Gvgani (2018) to evaluate the constructs of PU and PEOU. Similarly, the scale used to measure FINAD was adapted from instruments used for the same purpose by Dulcic, Pavlic and Silic (2012) as well as Shahbaz et al. (2020). Organisational performance was assessed with a scale sourced from a study by Elbashir, Collier and Davern (2008). All measurement items in the scales were accompanied by five-point Likert scale answer options on a range of strongly disagree (1) to strongly agree (5). Table 1 presents the measurement items together with the results obtained from a factor analysis conducted, which reflects the corresponding factor loadings and the t-stat for each item.

Table 1 indicates that the factor loadings for items in the PU scale varied from 0.699 to 0.787, while PEOU-related items exhibited factor loadings between 0.500 and 0.829. Moreover, items in the OP scale exhibited factor loadings ranging from 0.705 to 0.791. The factor loadings for FINAD varied from 0.320 to 0.822. Subsequent to the discovery of indicators with low loading items (e.g. FINAD1 with a factor loading of 0.320), a conservative methodology was implemented in accordance with Howard’s (2016) proposition. This technique involves the systematic elimination of items with loadings below 0.400 to assess potential improvements in the scale (Decius, Schaper & Seifert 2019). Interestingly, the elimination of FINAD1 did not enhance the scale; therefore, FINAD1 was retained. Based on the results of the factor analysis, the measurement items for PU, PEOU, FINAD and OP were considered to have achieved satisfactory indicator reliability.

The covariance-based structural equation modelling (CB-SEM) technique was utilised to investigate the proposed linear correlations between study constructs as expressed in the study’s hypotheses. Reinartz, Haenlein and Henseler (2009) assert that CB-SEM is suitable for validating theoretically posited associations. Aimran et al. (2017) also assert that CB-SEM outperformed Partial Least Square-Structural Equation Modelling (PLS-SEM) in accuracy consistency while estimating bigger samples (100 or more). The current study’s sample size exceeds 100, hence validating the application of CB-SEM for analysing the proposed correlations.

The CB-SEM process consists of two main stages: Firstly, it validates the measurement model by assessing the reliability and validity of the study constructs. Therefore, the internal consistency reliability of the constructs was assessed utilising both Cronbach’s alpha and composite reliability, with a threshold of 0.700. Convergent validity was assessed using average variance extracted (AVE), with values above 0.500 indicating adequate convergent validity. Furthermore, discriminant validity was evaluated using the heterotrait–monotrait (HTMT) criterion; an HTMT ratio below 0.850 signifies sufficient discriminant validity. Secondly, CB-SEM evaluates the structural model by analysing the linear relationships among the proposed relationships (De Carvalho & Chima 2014).

Ethical clearance to conduct this study was obtained from the University of Pretoria Faculty of Economic and Management Sciences Research Ethics Committee with reference number: EMS129/23.

The construct-related descriptive statistics results summarised in Table 2 offer insights into respondents’ average perceptions, data distribution normality and the likelihood of multicollinearity and common method bias (CMB).

The results reveal that PU possesses the highest mean of 4.276, signifying that respondents predominantly concurred on its importance. Furthermore, PEOU, FINAD and OP exhibited means of 3.992, 3.906 and 3.968, respectively, indicating that study respondents had a substantially elevated perception about these constructs. The skewness and kurtosis scores provide insights into the data normality. The skewness values for all constructs vary from −1.265 to −0.806, while the kurtosis values span from 0.367 to 2.318, all falling below the thresholds of 2 for skewness and 7 for kurtosis (Matore & Khairani 2020), which then signals that the data largely take the form of a normal distribution.

The tolerance and variance inflation factor (VIF) are employed to evaluate multicollinearity and CMB. Common method bias denotes the systematic measurement error that occurs when both dependent and independent variables are evaluated using the identical instrument, which may exaggerate the correlations between them (Desmaryani, Soleh & Wiarta 2024). A tolerance value under 0.1 and a VIF beyond 5 would signify significant multicollinearity problems and possible CMB (Kock 2015). The tolerance values for all constructs range from 0.521 to 0.614, while the VIF values were from 1.805 to 1.921, all of which are within acceptable ranges. The results indicate that multicollinearity is not an issue in this study and the data are free from CMB.

Prior to evaluating the hypothesised linear relationships between the study’s constructs, the measurement model was assessed for internal consistency reliability, convergent validity and discriminant validity. The results stemming from these assessments are shown in Table 3. The results signal that the study constructs achieved internal consistency reliability, as evidenced by Cronbach alpha values ranging from 0.799 to 0.880. This was additionally corroborated by composite reliability values that were above the 0.700 criterion (rho_c ranging from 0.826 to 0.881). The study’s findings also showed that the AVE values for the constructs ranged from 0.505 to 0.596, demonstrating the attainment of convergent validity. Furthermore, the HTMT values for PU, PEOU, FINAD and OP were below the 0.850 threshold, indicating sufficient discriminant validity among the study constructs.

To investigate the hypothesised linear relationships between PU, PEOU, FINAD and OP for the sampled SMEs in South Africa, a structural model was developed as illustrated in Figure 2. The model fit indices for the structural path analysis model shown in Figure 2 indicated that the structural model attained satisfactory model fit. This is shown by Chi-square Minimum Discrepancy Function (CMIN)/degrees of freedom (df) of 2.679, Comparitive Fit Index (CFI) of 0.977, Tucker-Lewis Index (TFL) of 0.972, NFI of 0.964, GFI of 0.963, AGFI of 0.949, RMSEA of 0.040 and standardised root mean square residual (SRMR) of 0.035, which are all above the cut-off values.

Figure 2 indicates that the path coefficient for the association between PU and FINAD is 0.40, for PEOU and FINAD is 0.45, and for the relationship between FINAD and OP is 0.80. The path coefficients indicate a positive correlation between PU and FINAD, PEOU and FINAD as well as between and FINAD and OP.

Table 4a and Table 4b encapsulates the path coefficients, t-values, p-values and conclusions about the direct relationships within the predictive model for PU, PEOU and FINAD, as well as the association between FINAD and OP for the sampled SMEs in South Africa.

The results in Table 4a and Table 4b indicate that the prediction model for the association between PU, PEOU and FINAD yielded an R² value of 0.639, implying that PU and PEOU account for 63.9% of the variance in FINAD. Correspondingly, the R² for the predictive model between FINAD and OP was 0.632, indicating that FINAD accounts for 63.2% of the variance in OP.

The results in Table 4 indicated that PU exerts a positive and statistically significant effect on FINAD (β = 0.403, t = 6.364, p = 0.000), hence corroborating H₁. This indicates that SMEs are more inclined to embrace FinTech solutions when they recognise the technology as advantageous and effective in enhancing business operations. Likewise, PEOU and FINAD exhibited a statistically significant and positive correlation (β = 0.445, t = 6.699, p = 0.000), which implies that H₂ was supported. These findings show that the cohort of respondents in the study perceive FinTech solutions as more adoptable when they are straightforward and user-friendly. Further, the findings reveal that FINAD has a statistically significant and positive correlation with the OP of SMEs (β = 0.795, t = 9.620, p = 0.000). Consequently, H₃ was statistically supported, meaning that FINAD contributes to improved OP for SMEs.

Firstly, the study utilised a quantitative cross-sectional methodology, which restricts the capacity to monitor temporal changes in FinTech uptake and its effects. A longitudinal study could offer more insight into the evolution of FinTech adoption and its effect on organisational performance over time. Secondly, the study relied on a purposive sample of respondents drawn from SMEs in South Africa, thus limiting the generalisability of the findings to other contexts, especially in diverse emerging markets or sectors. Subsequent research may broaden this investigation to more nations or sectors to ascertain if analogous trends arise. Furthermore, although the TAM was beneficial for this study, alternative frameworks such as the Unified Theory of Acceptance and Use of Technology (UTAUT) or the Diffusion of Innovations Theory may be examined to offer a more comprehensive understanding of technology uptake.