The Resource-Based View (RBV) theory emphasizes that a company’s competitive advantage stems from unique internal capabilities that are difficult to imitate and replace by competitors. RBV argues that organizations can achieve and sustain competitive advantage through the strategic utilization of resources and capabilities (Barney, 2020). These resources are not only tangible—such as physical assets or technology—but also include intangible assets like employee expertise, organizational culture, business processes, and technological knowledge. In the context of SMEs, which often face resource constraints, the RBV approach provides a theoretical foundation to understand how digital resource utilization and other strategic capabilities can enhance competitiveness and business performance.

Digital capabilities reflect an organization’s ability to leverage digital technologies to enhance operational excellence, market responsiveness, and innovation. Freitas Junior (2018) defines digital capability as the ability to develop, mobilize, and utilize resources through digital technology platform support. This includes aspects such as sensing (the ability to detect environmental changes), responsiveness (responding to market changes), and digitalization (transforming business processes through technology). In the context of SMEs, these capabilities enable companies to act more quickly, innovate more effectively, and overcome their resource limitations.

Dynamic capabilities refer to an organization’s ability to systematically adapt to environmental changes through the flexible management and development of resources. Teece et al. (1997) suggest that dynamic capabilities consist of three core processes: sensing (detecting external opportunities and threats), seizing (capturing opportunities through strategic innovation), and reconfiguring (reorganizing resources to adjust strategies). In the highly dynamic modern business context, these capabilities are critical to the survival and competitiveness of companies, including SMEs.

Strategic entrepreneurial behavior represents a planned and long-term- oriented entrepreneurial approach to seizing market opportunities. Amankwah- Amoah et al. (2024) explain that this behavior includes proactive action, innovation, and calculated risk-taking—all directed at supporting the organization’s vision and strategy. In this framework, entrepreneurs not only act as economic agents but also as drivers of business transformation through the development of adaptive and innovative strategies.

Business model innovation is the structural transformation process of how companies create, deliver, and capture value. Amit & Zott (2012) explain that business model innovation encompasses various aspects, including changes in cost and revenue structures, partnerships with business partners, and shifts in value propositions to customers. An innovative business model can create unique sources of differentiation that are difficult for competitors to imitate— especially in rapidly evolving industries.

Business performance in this study is influenced by digital capabilities, dynamic capabilities, and strategic entrepreneurial behavior—both directly and indirectly through business model innovation. In other words, the combination of internal factors (capabilities and behaviors) and innovation management determines how effectively SMEs achieve optimal business results.

Therefore, understanding the relationships among these variables is crucial for formulating strategies to enhance business performance based on capabilities and innovation.

This study conducted measurements to test the validity and reliability of each variable, namely digital capabilities, dynamic capabilities, strategic entrepreneurial behavior, business model innovation, and business performance. The validity and reliability tests for all these variables were processed using SmartPLS with a total of 156 respondents.

The data processing involved 31 indicators across the 5 variables, including:

Figure 4.1 Outer Loading Test Results

Convergent validity is used in testing as a benchmark for describing item variation and demonstrating a positive relationship between construct indicators and their respective constructs. At this stage, validity testing is conducted to determine the extent to which the indicators measure the construct. The ideal value for a loading factor is >0.70, indicating that the indicator is valid in measuring the construct. Another study by Kundu (2024) found that a loading factor value >50 is acceptable. The table below displays the complete values for each loading factor for each indicator in this study.

Table 4. 11 Convergent Validity

Source: Processed Primary Data (2025)

All of the items measured in this study had factor loadings that satisfied the validity requirements, which are above 0.70, according to the table data. As a result, nothing else needed to be taken out. Additionally, the following figure displays the outcomes of the outer loadings:

Average Variance Extracted (AVE)

One of the elements taken into account in convergent validity is Average Variance Extracted (AVE); for each variable to be deemed to have sufficient consistency, its AVE value must be greater than or at least equal to 0.50. Every variable, including digital capacity, dynamic capability, strategic entrepreneurial conduct, business model innovation, and company performance, has AVE values larger than 0.50, as Table 4.12 demonstrates. Thus, it can be said that any variable is regarded as legitimate.

Table 4.12 Average Variance Extracted (AVE)

It is evident from the data in Table 4.12 above that each variable's AVE value is higher than 0.50. Therefore, it may be said that the indicators or items of the constructs have very few faults.

The degree to which indicators can measure the desired variable is ascertained using discriminant validity. If an indicator's loading value on its associated latent variable is higher than its loading values on other latent variables in the cross-loading matrix, the indicator is said to satisfy the discriminant validity criterion. Nonetheless, the result utilized in this test comes from the Fornell-Larcker Criterion, which shows the total evaluation of cross- loadings for every variable.

Making ensuring that each indication has a higher loading on the construct it is meant to evaluate than on other constructs is the goal of cross-loading evaluation. This is crucial for evaluating the model's discriminant validity. According to Heseler et al. (2015), the anticipated cross-loading value is higher than 0.7. The cross-loading values for every variable and item in this study are shown in Table 4.13.

Table 4.13 Cross Loading

Source: Processed Primary Data (2025)

Discriminant validity is tested using the Fornell-Larcker approach, which compares the correlation value between the constructs with the square root of each construct's AVE. Discriminant validity is satisfied if the square root of the AVE is higher than the correlation value. The test's findings are displayed in

Table 4.14 Fornell-Larcker Criterion

Source: Processed Primary Data (2025)

Based on the table above, the variables business model innovation, digital capability, dynamic capability, and strategic entrepreneurial behavior have higher average Fornell-Larcker correlations than their correlations with other variables. For the digital capability variable, the correlation value is 0.781, which is higher than its correlations with other variables such as business model innovation (0.705), dynamic capability (0.740), business performance (0.616), and strategic entrepreneurial behavior (0.648). This indicates that the items within these variables represent their own constructs more than the other constructs. However, for the dynamic capability and strategic entrepreneurial behavior variables, the opposite is true, where the correlation between the two (0.701) is higher than the value for the dynamic capability variable (0.707). This means that the items within the dynamic capability variable are more closely correlated with the strategic entrepreneurial behavior variable than with the construct itself. However, other variables such as business performance and strategic entrepreneurial behavior still show good discriminant validity.

The reliability test in SEM-PLS aims to ensure that the indicators within a construct are able to measure consistently in accordance with their intended purpose, by demonstrating the stability and consistency of the measurement results. The main measures used to evaluate reliability are Cronbach’s Alpha and Composite Reliability.

Cronbach’s Alpha is a measure of internal reliability used to assess the consistency of values within a construct. An instrument is considered reliable if the Cronbach’s Alpha value is greater than 0.60. Further details can be seen in Table 4.15 below:

Table 4.15 Cronbach’s Alpha

Based on the results in the table above, it can be seen that all variables meet the required value of > 0.60.

Composite reliability is a measure of internal consistency that indicates how well indicators of a construct consistently measure the same concept. If the Composite Reliability value is greater than 0.70, then the indicators are considered consistent and acceptable.

Table 4.16 Composite Reliability

Based on the table above, all variables in this study have Composite Reliability values greater than 0.70. This indicates that all variables meet the reliability requirements.

This study uses a structural (inner model) approach to identify the relationships or correlations between the studied variables. The analysis process was conducted to obtain the R-Square values related to the dependent variables, which serve as the basis for structural testing. The inner model assessment is based on the following:

The purpose of collinearity testing is to assess the presence of strong linear relationships among independent variables in the model. A VIF value of < 5 is the threshold used to indicate that no multicollinearity issues exist; otherwise, it suggests deviations in the model.

Table 4.17 Collinearity VIF

Based on the table above, this study meets the required threshold, as all VIF values are below the limit of 5, indicating no collinearity among the variables in the model.

The degree to which the model explains the variance in each dependent variable examined in the study is indicated by the R-Square and Adjusted R- Square values. R-Square values vary from 0 to 1, where 0 denotes no association and 1 denotes a perfect or extremely strong one. The R-Square values for this investigation are shown in Table 4.18.

Table 4.18 R-Square

According to Table 4.18, the business model innovation variable's R2 value is 0.619, meaning that digital capacity, dynamic capability, and strategic entrepreneurial behavior account for 61.9% of the variation in business model innovation. Other factors not included in the model account for the remaining 38.1%. This suggests that the model's ability to explain business model innovation is strong. company model innovation, on the other hand, can account for 38.0% of the variance in company success as a mediating variable, according to the R2 value for business performance, which is 0.380. This number, which is in the moderate range, suggests that changes in business performance may be mostly explained by business model innovation.

The degree to which a model can forecast the variance of dependent variables is measured by predictive relevance (Q-Square). This testing method has a rule: a predictive measure is deemed good if its Q-Square value is greater than 0; if it is less than 0, it is deemed poor. The model's predictive accuracy for the pertinent associations is demonstrated by the study's Q-Square value, which is larger than 0. Table 4.19 displays the analysis's specific findings.

Table 4.19 Q-Square

Source: Processed Primary Data (2025)

According to Table 4.19's computation results, the business model innovation variable's Q2 value is 0.330. Given that this value is higher than zero, the model's predictive power for the business model innovation variable is 33%. This shows that the independent factors can predict business model innovation with a fair degree of accuracy.

In contrast, the business performance Q2 value is 0.202, meaning that the model can forecast company performance with a 20.2% accuracy rate. This shows that the model has a rather good predictive relevance for business performance, even though it is lower than the Q2 for business model innovation.

A statistical technique called hypothesis testing is used to ascertain if a population parameter assumption can be accepted or rejected in light of the sample that was collected. According to Hair J (2014), the purpose of this test is to investigate the importance of the connections between the structural model's structures. Making ensuring that the relationships in the study model are statistically supported and not the result of chance is crucial.

Each independent variable's direct impact on the dependent variable was assessed using the direct hypothesis testing analysis results, which are displayed in Table 4.20. The T-Statistics and P-Values are used to evaluate the direct effect's significance. The table below displays all of the direct impact test's findings.

Table 4.20 Results of Direct Hypothesis Testing

Source: Processed Primary Data (2025)

Based on the table above, P-values are all below 0.05, and the interpretations are as follows:

Has a 0.342 coefficient. This indicates that business model innovation will rise by 0.342 units for every unit increase in digital capacity. It has been demonstrated that digital capabilities significantly boosts business model innovation, with a P- value of 0.000 < 0.05. This is in line with the theories of Westerman (2012) and Freitas Junior (2018), who contend that digital competence empowers businesses to exploit information, adapt to changes in the market, and foster business model innovation through the integration of digital systems.

Has a 0.242 coefficient. This indicates that business model innovation will rise by 0.242 units for every unit increase in dynamic capabilities. Dynamic capability has a strong positive impact on business model innovation, as demonstrated by the P-value of 0.001 < 0.05. This bolsters the argument put forth by Teece et al. (1997) that dynamic capability is essential for seizing opportunities and bringing about change via innovative business models.

Has a coefficient of 0.310. This means that if strategic entrepreneurial behavior increases by 1 unit, business model innovation will increase by 0.310 units. With a P-value of 0.001 < 0.05, it is proven that strategic entrepreneurial behavior has a significant positive effect on business model innovation. This finding reinforces the theory of Amankwah-Amoah et al. (2024) and Zahoor (2023), which suggests that strategic entrepreneurial behavior—comprising innovation, proactiveness, and risk-taking—plays an essential role in encouraging SMEs to create relevant and competitive new business models.

Has a coefficient of 0.603. This means that if business model innovation increases by 1 unit, business performance will increase by 0.603 units. With a P- value of 0.000 < 0.05, it is proven that business model innovation has a significant positive effect on business performance. This is in line with the findings of Amit & Zott (2012), who explain that business model innovation can create greater value for customers and improve both financial and non-financial business performance.

This study examines indirect hypotheses by testing the role of business model innovation as a mediator between digital capability, dynamic capability, and strategic entrepreneurial behavior and business performance. The researcher conducted the test using T-Statistics and P-Values to determine significance.

Table 4.21 Results of Indirect Hypothesis Testing

Source: Processed Primary Data (2025)

Based on the table above, the results of the mediation path coefficient analysis for each hypothesis are explained as follows:

The interpretation of these three results shows that SMEs need to develop business model innovation to achieve optimal business performance. Business model innovation acts as a transformation mechanism for capabilities and entrepreneurial behavior. This supports the argument of Amit & Zott (2012), who state that business model innovation serves as a mediator linking internal company capabilities with external value creation.

Figure 4.2 Bootstrapping Test Results

The results of this study indicate that digital capability, dynamic capability, and strategic entrepreneurial behavior significantly influence business model innovation in SMEs in Indonesia. This suggests that the integration of technology, adaptive capabilities, and strategic entrepreneurial behavior—such as being innovative, proactive, and willing to take risks—are crucial in driving business model renewal.

These findings support Wang et al. (2023), who state that digitalization enhances SME business performance through business model innovation, where digital capabilities accelerate value creation through experimentation with new business models. Moreover, this research is in line with Ramadan et al. (2024) and Saputra et al. (2024), who emphasize the strong relationship between organizational capabilities and the tendency toward innovation in business structures.

The results indicate that the influence of digital capabilities is more pronounced compared to several previous studies conducted in developed countries. This may be due to differences in the level of digital technology adoption, where Indonesian SMEs, previously lagging in technology, are currently experiencing rapid digitalization, making the effects of digital capabilities more noticeable.

Theoretically, this study enriches the literature on the integration of capabilities and strategic entrepreneurial behavior as a driver of business model innovation in the SME sector. Practically, SME actors need to allocate resources in a balanced manner between technological development and the cultivation of an entrepreneurial culture to produce business models that are relevant to the market.

The analysis results show that business model innovation significantly mediates the relationship between digital capability, dynamic capability, and strategic entrepreneurial behavior on SME business performance. This means that the organization's ability to develop business model innovation acts as an important bridge between internal capabilities and resulting performance.

These findings strengthen the study by Ramadan et al. (2024), which highlights that business model innovation is a key pathway in transforming organizational strategic potential into business performance. The research also aligns with Amaya et al. (2024), which emphasizes the importance of innovation in leveraging dynamic capabilities and digitalization to enhance efficiency and competitiveness.

This research differs from several other studies, such as Merín-Rodrigáñez et al. (2024), who emphasize strategic leadership as the primary mediator. In contrast, this study positions business model innovation as the central link. This difference may be due to geographic context and business mindset structure, where decision-making in Indonesian SMEs is more concentrated in the owner, rather than in operational managers.

Theoretically, this study strengthens the position of business model innovation as a strategic mediating variable within the framework of capabilities and behavior. In practice, SME actors and policymakers need to promote innovation-based business transformation as a primary strategy to enhance long- term competitiveness.

The results of this study indicate that business model innovation has a significantly positive effect on SME business performance. Innovations made in creating, delivering, and capturing value have been proven to improve the operational and financial performance of the company.

This finding supports the studies by Amit & Zott (2012) and Wang et al. (2023), which emphasize that business model innovation is not only a tool for adaptation but also a growth strategy that creates competitive advantages. Furthermore, this research highlights the importance of regularly updating business models in response to market and technological dynamics.

This result differs from studies such as Liu (2022), which focused on large manufacturing companies, where the effect of business model innovation in SMEs appears to be more direct and significant. This may be because SMEs have more flexible organizational structures and quicker decision-making, making innovation adoption have a more immediate impact on business performance.

Theoretically, these results reinforce the role of business model innovation as a key determinant of organizational performance, especially in small organizations. Practically, SMEs are encouraged to regularly redesign their business models, leverage technology, and create new sustainable value creation mechanisms.