Digital transformation refers to an organisation-wide reconfiguration enabled by digital technologies consisting of artificial intelligence, cloud computing and analytics, which fundamentally modify how businesses create and attain value (Saeedikiya et al., 2025; Warner & Wäger, 2019). More than the adoption of isolated tools, DT reflects a holistic shift in strategy, culture and capabilities that reshape business models, customer engagement and decision-making. Scholars distinguish between digitisation, digitalisation and DT. Digitisation denotes to the transformation of analogue data into digital formats, whereas digitalisation includes automating and optimising processes using digital tools (Chisita et al., 2021). Digital transformation, in contrast, entails deliberate and strategic changes across the organisation that extend to business models, governance and organisational culture (Nasiri et al., 2022; Ranawaka & Said, 2024). In this sense, DT reflects not only the utilisation of technology but the integration of organisational and managerial capabilities to achieve sustainable competitive advantage. In the setting of this study, DT is evaluated through DI and TMI.

In this study, sustainability is understood as the capacity to maintain performance at a consistent level over time, extending beyond financial outcomes to encompass social and environmental responsibilities (Wentzel et al., 2022). Sustainable business performance reflects how effectively firms organise their functions, processes, and resources to achieve long-term viability, environmental stewardship and social well-being. Central to this perspective is the strengthening of stakeholder relationships, particularly with employees, customers and regulators, whose expectations increasingly shape organisational legitimacy (Liu et al., 2023). The triple bottom line framework (Elkington & Rowlands, 1999) remains a dominant approach for conceptualising sustainability, requiring firms to balance financial performance (FP) with environmental integrity and social value. Recent studies highlight that SMEs’ ability to meet these three dimensions is directly linked to their competitiveness, innovation capacity and social legitimacy (Belhadi et al., 2022; Kwiotkowska & Gębczyńska, 2021).

The RBV suggests that firms sustain a competitive edge through deploying internal assets and capabilities that competitors cannot easily imitate (Barney, 1991). These internal assets are tangible, such as infrastructure, or intangible, like skills, knowledge and organisational ethos, but their strategic value lies in how they are configured to deliver long-term performance (Barney et al., 2011). The RBV suggests that DT helps firms use technology and information more effectively, improving efficiency, adaptability and long-term competitiveness (Wu et al., 2025). Recent research continues to apply RBV to explain how organisations leverage digital resources and organisational capabilities to achieve resilience and sustained performance in dynamic environments (Khurana et al., 2022; Yi et al., 2023; You et al., 2023). Resource-Based View, therefore, provides an appropriate theoretical lens for examining how technological resources and organisational capabilities associated with DT contribute to sustainable business performance in SMEs.

Within this study, DI is conceptualised as a technological resource. It captures the breadth and depth of digital tools embedded into operations, ranging from cloud computing and analytics to artificial intelligence and Internet of Things (IoT). As a resource, DI equips firms with the infrastructure to streamline processes, generate efficiencies and respond to customer needs more effectively (Burinskienė & Seržantė, 2022; Criveanu, 2023). Higher levels of DI are associated with innovation and competitiveness at both firm and economy-wide levels, as digital adoption enables more scalable and adaptable models of value creation (Chiappini & Gaglio, 2024; Li et al., 2023). Recent studies further demonstrate that DT can strengthen enterprise innovation resilience and organisational adaptability (Peng & Jia, 2024), while effective digital data management systems support SME scalability and growth in the South African context (Gaaje et al., 2025).

However, without complementary organisational capabilities, these technological resources often remain underutilised or fragmented, limiting their transformative potential. Transformation Management Intensity, by contrast, represents an organisational capability. It reflects leadership commitment, cultural readiness, and managerial effort directed towards aligning digital initiatives with broader strategic goals. Transformation Management Intensity determines whether digital resources are applied in ways that generate long-term value, facilitating cross-functional collaboration, organisational learning and the reconfiguration of existing routines (Belhadi et al., 2022; Mwangi & Mang’ana, 2024). Research increasingly highlights that successful DT in SMEs depends on the development of organisational competencies capable of guiding and coordinating transformation initiatives (Gonzalez-Varona et al., 2024). Similarly, studies on sustainable DT emphasise the importance of structured transformation management approaches for integrating technological innovation with organisational change (Mick et al., 2024).

Prior studies suggest that strong transformation management is essential to embedding digital solutions beyond compliance functions, enabling firms to integrate sustainability goals, improve supply chain resilience and foster inclusive practices (Kargas et al., 2024). The RBV offers a useful lens through which to look at how SMEs combine technological resources and organisational capabilities to achieve sustainable performance. Recent DT research supports this view, demonstrating that firms derive value from digital technologies only when they combine technological adoption with organisational capabilities that enable learning, coordination and strategic alignment (Egala et al., 2024; Scuotto et al., 2021). Grounded in this theoretical framing and the reviewed literature, the next section sets out the study’s hypotheses.

The hypotheses for the study are based on how DI and TMI, as conceptualised within the RBV, influence financial, environmental and reviewed in SMEs.

Ethical clearance to conduct this study was obtained from the University of Johannesburg and the Johannesburg Business School Research Ethics Committee (No. JBSREC2024223). All respondents gave their informed consent, and participation was entirely voluntary. Confidentiality and anonymity were guaranteed, and the data were used solely for academic purposes.

The analysis of educational levels among respondents revealed that most individuals (51.6%) reported that high school was their highest level of education. Followed by 25.8% who had obtained a diploma qualification. A smaller portion of the sample, 12.2%, held an undergraduate degree, while only 10.3% had completed postgraduate studies. These statistics indicate a workforce that is primarily composed of individuals with basic to intermediate educational qualifications. Regarding the size of businesses represented in the study, an overwhelming 80.7% of enterprises reported employing fewer than five people. An additional 19.2% employed fewer than 20 people, while no respondents indicated having businesses that employed 50 or more individuals. This skew towards micro-enterprises reflects a characteristic trend in many emerging economies where entrepreneurship is often necessity-driven and concentrated in the informal or semi-formal sector. When considering the duration of business operation, the data indicated a relatively even distribution across experience levels. Approximately 18.3% of the businesses had been operational for less than 1 year, and 23.9% had been in existence for 1–2 years. A slightly higher proportion (24.8%) had operated for 3–4 years, while the largest group (32.8%) reported business operations extending beyond 5 years. This distribution suggests a vibrant entrepreneurial environment with a healthy mix of both new entrants and more seasoned operators. However, the significant number of businesses in their early stages may also imply potential volatility and higher business failure rates, necessitating tailored support mechanisms to enhance sustainability and long-term success. The sectoral distribution of respondents highlighted a predominance of service-oriented and primary sector activities. The highest representation was found in the community, Social and Personal Services, comprising 17.3% of the sample. This was closely followed by Agriculture and Catering, Accommodation and Other Trade sectors, each contributing 15.4%. Retail and Motor Trade and Repair Services accounted for 12.2%, while construction followed at 11.7%. Other notable but less represented sectors included Finance and Business Services (11.2%) and Transport, Storage and Communications (4.6%). In contrast, capital-intensive and heavily regulated sectors such as Mining and Quarrying (2.8%) and Electricity, Gas and Water (1.4%) were minimally represented. This trend suggests a local economy driven largely by accessible and lower-barrier sectors, which typically require modest capital investment and offer quicker entry. These sectors often absorb a significant portion of informal or semi-formal labour, reflecting the structural composition of entrepreneurship in the region. Table 1 illustrates the demographic characteristics.

In structural equation modelling (SEM) frameworks, reliability analysis is a critical component of ensuring that measurement instruments are consistent with one another. The study’s reliability was assessed using Cronbach’s Alpha (α) and Composite Reliability (CR or ρc). Table 2 illustrates the results. These are two frequently recommended metrics for reflective measurement models (Hair et al., 2021; Taber, 2018).

Table 2 shows that all five latent components in the investigation had high internal consistency. The figures for Cronbach’s alpha increased from 0.911 for FP to 0.968 for SP. The CR values increased from 0.933 to 0.975. Awang et al. (2018) and Purwanto and Sudargini (2021) demonstrate that the measurement model is robust, as the reliability threshold of 0.70 for both indices is substantially lower than these figures. Digital Intensity was a dependable method for evaluating DT abilities, as evidenced by its α value of 0.960 and CR value of 0.965. The reliability of the EP and FP constructs was demonstrated by their α values of 0.930 and 0.911, respectively, and their CRs of 0.950 and 0.933. The most reliable values were achieved by TMI and SP (α = 0.967 and 0.968, respectively). These values may indicate that the model is accurate; however, they may also indicate that certain items require verification for redundancy (Henseler et al., 2016; Nunnally & Bernstein, 1994).

Table 3 shows the assessment for convergent validity. Fornell and Larcker (1981) assert that Average Variance Extracted (AVE) values should exceed 0.50 to demonstrate that the latent construct, rather than error, accounts for more than 50% of the variance in the indicators. In addition, standardised factor loadings should exceed 0.70; however, loadings between 0.60 and 0.70 may be acceptable for exploratory research (Henseler et al., 2016; Saeed et al., 2022). The item’s high factor loadings indicate that it has a significant amount of variance in relation to the construct it is intended to evaluate.

Table 3 demonstrates that the AVE values of all five components exceed 0.736, which is significantly higher than the standard 0.50 threshold. This demonstrates that the measurement model has a high degree of convergent validity. The AVE of DI is 0.736, with factor loadings ranging from 0.802 to 0.906. This indicates that DI is internally consistent. Environmental Performance exhibited an even higher AVE of 0.826, with loadings ranging from 0.900 to 0.925, which further enhanced the construct’s coherence. The FP score was 0.736, which was equivalent to the criterion. The construct’s validity was confirmed by loadings such as 0.891. The item loadings were greater than 0.940, which contributed to the highest AVE of 0.888 for SP. This demonstrates that the items are highly comparable; however, it also raises concerns about their potential redundancy. The TMI also achieved an AVE of 0.774, with loadings ranging from 0.815 to 0.924, indicating its reliability.

In this study, the Heterotrait-Monotrait Ratio (HTMT) and the Fornell and Larcker Criterion were two widely recognised methods for evaluating discriminant validity.

Means and standard deviations were examined as scale statistics. The standard deviations were less than 1.4, and most of the mean values fell within the range of 2.7 to 3.3. This implies that the respondents primarily concurred, and there was minimal variation. The indicators employed in the model are corroborated by these figures.

Path modelling illustrates the relationship between seen or measured variables and a theoretical construct, in addition to examining the structural courses of a hypothesised research model (Guenther et al., 2023). To assess the effectiveness of the study’s structural model, p-values and regression coefficients were both examined. Figure 2 illustrates one potential result of the structural model for the theory that is being presented.

The measurement and structural model evaluations’ overall findings support the validity and reliability of the constructs employed in this investigation. All indicators met or exceeded the recommended thresholds for internal consistency, convergent validity and discriminant validity, while model fit indices indicated an acceptable level of fit. These results provide confidence that the measurement model is robust and suitable for testing the proposed research hypotheses.

With the validity and reliability of the measurement model established, the next stage of the analysis involved testing the hypothesised relationships between DT (DI and transformational management intensity) and sustainable business performance (financial, environmental and social dimensions). The hypothesis testing results for this study’s estimation are presented in Table 7. The table displays the path coefficients, the t-statistics, the proposed hypothesis and whether the hypothesis is supported or rejected. Path coefficients greater than one indicate a strong relationship between latent variables, while t > 1.96 indicates a robust association (Zhang, 2022):

The structural model results show that FP is significantly and positively influenced by DI, with a t-statistic of 7.533, a path coefficient (β) of 0.762 and a p-value of 0.000. This indicates that SMEs with higher DI achieve stronger financial outcomes. In practice, firms that allocate substantial resources to digital technologies, systems and capabilities are more likely to experience cost efficiencies, revenue growth and financial resilience. These results provide support for Hypothesis 1 (H1) and are consistent with prior studies. Moro-Visconti et al. (2025) highlight how digital platforms improve revenue streams, reduce costs and enhance financial viability. Criveanu (2023) similarly finds that DI contributes to competitiveness and economic performance, while Nasiri et al. (2022) emphasise the role of DI in enabling adaptability and long-term financial advantage. More recent applied studies also reinforce this link, showing how digital investments reduce transaction costs and enhance decision-making efficiency (Bindeeba et al., 2025; Gertzen et al., 2022):

The structural model results show that EP is significantly and positively influenced by DI, with a path coefficient of β = 0.777, a t-statistic of 7.389 and a p-value of 0.000. This indicates that higher levels of DI enable SMEs to achieve stronger environmental outcomes. In practice, firms that embed digital technologies into their operations are better able to optimise resource use, reduce waste, and adopt greener practices.

These results provide support for Hypothesis 2 (H2) and align with prior studies. Zhang and Zhao (2023) demonstrate that DI enhances firms’ capacity to improve EP by integrating sustainability goals with digital processes. Table 6 illustrates the scale accuracy analysis of the measurement items. Similarly, Chen and Ren (2025) find that DT significantly contributes to ESG outcomes, with particularly strong effects on environmental dimensions. Burinskienė and Nalivaikė (2024) further argue that SMEs adopting digital solutions are better positioned to meet sustainability objectives, particularly in the context of the EU Green Deal:

The structural path analysis indicates a statistically significant and robust correlation between DI and SP. This is demonstrated by a route coefficient of β = 0.616, t = 6.258 and p = 0.000. This is consistent with Hypothesis 3, which posits that companies’ SP, as measured by ethical governance, community involvement and stakeholder engagement, improves when they exert more effort to become digital. However, the correlation is not as robust as DI’s influence on environmental and FP. This implies that the social advantages of DT may be contingent upon the circumstances or may accrue gradually. The results of hypothesis 3 support Meng et al. (2022) finding that enhanced DI improves SP:

The results of the structural model for H4 indicate that there is no significant correlation between FP and TMI, with a path coefficient of β = –0.033, a t-value of 0.282 and a p-value of 0.778. This suggests that, within the SME context, TMI has little to no direct impact on financial outcomes. As a result, Hypothesis 4 (H4) is not supported. This finding contrasts with prior research, which generally identifies a positive relationship between transformation management and FP. For example, Mwangi and Mang’ana (2024) show that dynamic managerial capabilities strengthen financial outcomes in logistics firms, while Liang et al. (2022) argue that adaptive capabilities embedded in transformation strategies generate long-term financial advantage. Belhadi et al. (2022) also find that transformation initiatives integrating Industry 4.0 practices improve operational efficiency and economic resilience. The SME context may be the reason for the discrepancy between these studies and the current findings. Unlike larger firms, SMEs often face resource constraints and shorter planning horizons, which may limit the translation of transformation management efforts into immediate financial returns:

The structural model demonstrated a statistically insignificant and weak negative correlation between TMI and EP, with a path coefficient of β = −0.043, a t-value of 0.335, and a p-value of 0.715. These results indicate that TMI does not significantly influence EP in the context examined. Consequently, Hypothesis 5 (H5) is not supported. This finding contrasts with much of the literature, which suggests that transformation management can enhance environmental outcomes. Studies such as Wang and Teng (2022) and Yu and Ramanathan (2016) emphasise that managerial capabilities strengthen supply chain and operational practices that improve EP, while Belhadi et al. (2022) show that Industry 4.0 transformation fosters circular economy models that support emissions reduction and sustainable resource use. However, the present results are consistent with Kargas et al. (2024), who argue that transformation effort or management intensity alone is insufficient to produce meaningful environmental improvements:

Although the hypothesis proposed a positive relationship between TMI and SP, the statistical results do not support this claim. The path coefficient (β = 0.162) suggests a slight positive trend; however, the t-value (1.485) and p-value (0.138) fall below the commonly accepted thresholds for significance (t ≥ 1.96; p < 0.05). These results indicate that TMI does not have a statistically significant impact on SP in the context examined. Consequently, Hypothesis 6 (H6) is not supported. This finding diverges from prior literature, which has emphasised the social benefits of transformation management. For example, Kwiotkowska and Gębczyńska (2021) highlight the importance of leadership and sustainable transformation pathways in improving social outcomes, while Belhadi et al. (2022) suggest that transformation efforts can foster inclusive practices through circular economy models. However, the present results align with Chen and Wang (2024), who argue that improvements in ESG performance are significantly mediated by factors such as innovation capability and service orientation, rather than being directly attributable to transformation intensity alone.

This study advances the RBV by showing that digital resources are essential enablers of sustainable competitiveness in SMEs and extends sustainability research by highlighting the financial, environmental and social benefits of DT. For practice, SME leaders should prioritise investment in digital platforms, analytics and automation to strengthen DI, while support agencies must provide funding, training and advisory services to enhance SMEs’ digital capabilities. Embedding triple bottom line objectives into digital strategies is vital to ensure integrated and sustainable outcomes.

For policymakers and institutions that support SMEs, the findings emphasise the need to strengthen the wider digital ecosystem that enables SME transformation. This could involve expanding programmes that develop digital leadership and skills, providing financial incentives or grants to encourage technology adoption, and improving access to reliable digital infrastructure. Policymakers could therefore enhance these efforts by introducing sustainability-linked incentives, such as funding programmes or tax benefits that motivate SMEs to align digital adoption with environmental and social goals. In addition, support agencies could encourage SMEs to use organisational readiness or culture assessment tools to evaluate their preparedness for DT and identify capability gaps. Collectively, these measures can assist SMEs in turning DT initiatives into tangible improvements in financial, environmental and SP.

This study is limited by its focus on SMEs within a single municipality, its cross-sectional design, reliance on self-reported data, and the predominance of service-oriented firms in the sample, which constrain generalisability and causal inference. Future research should adopt longitudinal or mixed methods approaches, expand samples across regions and sectors, and include under-represented industries such as manufacturing and high-tech. Further studies could also explore mediating and moderating factors, such as innovation capability, leadership style, government support, digital literacy and cultural influences, in shaping the DT–sustainability relationship.