Strategic planning and management abilities as internal factors that influence the corporate competitiveness of the SMEs

In recent years, SMEs have been an important matter for public policies, which has allowed an improvement of certain aspects directly linked to the competitiveness of these types of enterprises. However, the challenges being faced by the SMEs and by micro-enterprises require reinforcing the sensibility towards the problematics of said group, as well as intensifying their support.

To promote the national economic development through the boost of SMEs, the Mexican government approved the Law for the Development of Competitiveness of Micro, Small and Medium-sized Enterprises (^{DOF, 2009}), where competitiveness is defined as the capacity to maintain and strengthen the profitability of enterprises and their participation in the markets based on advantages associated to their products or services.

A study regarding SMEs projects, based on the strategic analysis for the development of the SME in the State of Veracruz, Mexico (^{Aragón and Rubio, 2006}), found that the development of the small and medium-sized enterprises requires five key factors: 1) having a manager with a university degree and being clearly oriented towards sales; 2) increasing formal strategic planning and alliances and cooperation agreements, aiming for a more innovative, flexible and daring behavior; 3) having a more developed organizational structure, mainly in the areas of accounting and finances; 4) use new information technologies, make the correct use of communications equipment and IT resources, and use software applications in the management of their enterprise; and 5) using systems for cost accounting, annual budget, and the analysis of their economic and financial situation for the making of decisions.

According to the Scientific and Technological Advisory Forum (cited in ^{Góngora and Madrid, 2010}), Mexico loses competitiveness to other countries mainly due to its structural issues and the inefficient functioning of its national markets. The Mexican scientific and technological policy of recent years has not managed to motivate an improvement in the national competitive levels. One way to achieve a sustainable competitive advantage is through the formation of a central nucleus via the set of intangible assets of the enterprise, also known as intellectual capital, which is comprised by the people in the enterprise, that is, its set of human capital and, more concretely, the knowledge, skills, values, capabilities and competences brought together individually and collectively (^{Araujo et al., 2006}).

Although it is true that the majority of people that comprise the human structure of an organization are important for the effective and efficient development of its economic activity, not all human groups within the organization are equally important and strategic. The responsibility to choose the path of an enterprise falls on the management, as well as the combination of resources that it requires for the achievement of its objectives and goals, and the market(s) in which it will participate; it is because of this that the management resources comprise a collective whose relevance in the generation and preservation of corporate success has been made manifest by several authors (e.g., ^{Castanias and Helfat, 1991}, ²⁰⁰¹; ^{Lado and Wilson, 1994}; ^{Pickett, 1998}; Landeta et al., 2007). Presently, Mexican enterprises, specially SMEs, lack management skills in their corporate management that will help them achieve a better behavior and more effective abilities to make changes in management strategies, programs, structures, etc. to allow the enterprise to adapt and thus be able to anticipate the changes that could affect it (^{Longenecker et al., 2009}).

As indicated by ^{Weik (2009)}, managers are now considered agents of change who, by establishing a mutually beneficial relationship with the different types of public that expect something from the enterprise, allow a progressive and continuous improvement of the organization. Thus, the performance of the management role for the achievement of corporate competitiveness is based on teamwork, having a network within and outside of the enterprise, and the execution of good strategic planning (^{Cantzler and Leijón, 2007}). Management skills are important for the efficient management of innovation, where technical capabilities are not sufficient for efficiency. The differences in learning styles are important in the acquisition of interpersonal skills, which can be used to develop leadership skills that help create corporate competitiveness strategies (^{Dreyfus, 2008}; ^{De Meuse et al., 2011}; ^{Koenigsfeld et al., 2012}; ^{Thorn, 2012}; ^{Tonidandel et al., 2012}; ^{Zhang et al., 2013}). In this sense, the following work hypotheses are considered:

H ₁ . The corporate competitiveness of SMEs is influenced by the Management Skills of the enterprise.

H ₂ . The corporate competitiveness of SMEs is influenced by the Strategic Planning of the enterprise.

Once the hypotheses have been presented, the proposed explicative model is reflected as shown in Figure 1. This model establishes the influence of Strategic Planning and Management Skills on the corporate competitiveness manifested in the comparative data of theoretical foundation and the exploratory discussion done through a panel of experts.

Figure 1 Modelo Conceptual 

After discussing and theoretically supporting the proposed relations, we detail the analysis methodology in the following section in order to contrast the formulated work hypotheses.

Data collection

To contrast the formulated hypotheses, a series of actions were developed under two mixed research phases. In a first qualitative phase, an exploratory analysis was done, supported by a panel of experts (also known as the Delphi method) in order to validate the measuring instrument (apparent validity). The panel was comprised of six experts from three sectors related to SMEs: Government, Industry and Education, which were selected according to the pre-fixed objective and catering to the criteria of experience, position, responsibility, access to information, availability, and who had characteristics relating to SMEs. In this context, the prevalence of the so called Triple Helix is key (^{Mejía, 2004}; ^{González, 2009}): Government-Industry-University.

The work was done in two phases: the first is an open discussion panel, and the second, the survey applied to experts with the help of a semi-structured questionnaire. The objective of the questionnaire is to identify (based on the opinion of the experts) the influence of management skills as an internal factor of the corporate competitiveness of SMEs in Hermosillo, Sonora, Mexico.

The collection of information was carried out by the members of the research project, using a structured questionnaire as support. This questionnaire was applied to the managers of the SMEs affiliated to the National Chamber of Commerce, Services and Tourism (Canaco-Servytur, for its acronym in Spanish), which has over 390 enterprises, and the National Chamber of Industry Transformation (Canacintra, for its acronym in Spanish), which has 190 affiliated enterprises; both are located in the city of Hermosillo, Sonora. The work was executed in three phases: the first phase was done through the e-mail addresses obtained through the information provided by the industries; though given that the response rate was low, a second phase then took place, which consisted on phone communication; and the third phase was done on-site to finalize the results. The size of the sample obtained, after the filtering process, was of 108 valid questionnaires (see Table 1), 80 manifested being from the commerce sector and 27 from the industrial sector. Furthermore, 72 enterprises declared being small and 34 medium-sized.

For the information collection process, a semi-structured questionnaire was used. It contained some closed questions that were previously validated through qualitative techniques. For this purpose, the respondents had to be in the role of enterprise manager, and show their level of agreement or disagreement with a series of statements using a seven-point Likert measurement scale. Specifically, the information referred to the influence of internal factors, such as strategic planning and management skills in the corporate competitiveness of SMEs in Hermosillo, Sonora, perceived by the SME manager, and finally the social and demographic characteristics of the businessperson.

Structural equation modeling

To carry out the statistical analysis of the collected data, the methodology of Structural Equation Modeling (SEM) was used along with the Partial Least Squares (PLS) technique, which is calculated through the variance components matrix. To validate the measuring model, the following methodological actions were executed: (1) analyzing the content and apparent validity, (2) calculating the individual reliability of the reflective indicators through factor loading, (3) examining the construct validity: convergent validity and divergent validity. Regarding the validity of the structural model, the following were analyzed: (1) explained variances (R²) and (2) the path coefficients or standardized regression weights (β). Each one of the statistical criteria are proven in the following section by applying the SmartPLS 3.0 statistical information package (^{Ringle et al., 2005}).

Content and apparent validity

To validate the statistical instrument and its corresponding measurement scales, the following qualitative tests were applied:

Content validity, an exhaustive revision of the literature specialized in strategic planning, management skills, and corporate competitiveness was carried out in this test in order to theoretically support the measurement scales.

Apparent validity, this analysis determines that the measurement scales reflect what they are intended to measure, therefore the measurement scales, which were initially proposed based on the result of the previous qualitative studies, were modulated and adapted. For this, the research instrument was submitted to a Panel of Experts (Delphi method) on SMEs, in order to filter the indicators that did not show any relation to the research, which in turn made it possible for us to guarantee the achievement of satisfactory results (^{Zaichkowsky, 1985}).

Reliability of the indicators

To verify the individual reliability of the indicators as a part of a reflective construct, they need to have a factor load (λ) equal to or greater than 0.707 (^{Carmines and Zeller, 1979}). The factor loading shows that the variance shared between the construct and its corresponding indicators is superior to the error variance. Considering the aforementioned statistical acceptance criterion (λ ≥ 0.707), the following indicators were eliminated in the first phase: OG1: “Ability for the identification and resolution of problems” (λ = 0.666), OG3: “Willingness to act, feel and/or think about the needs of the clients, directing all their actions towards a strategy to guarantee the satisfaction of the same” (λ = 0.661), OG4: “Ability to listen and correctly understand the thoughts, feelings or concerns of the others even when they are not verbally expressed but need to be perceived by the rest” (λ = 0.626), OG5: “Ability to efficiently resolve situations, events or conflicts in which interests that could affect the relations between people are at play, and which could also present a risk to the objectives, interests or the image of the organization” (λ = 0.684), PE8: “Ability to improve the environmental behavior of the activities of the organization” (λ = 0.670), RH1: “Internalization of norms and moral principles that make them responsible of their own well-being and that of others, through a behavior based on socially accepted conducts” (λ = 0.689). After re-calculating the PLS algorithm in a second stage, the indicator OG2 was eliminated: “Ability to present solutions and resolve the differences of ideas or opinions of the parties” (λ = 0.672), which makes it evident that management does not place importance on the ability of Operations and Management.

Similarly, the communality test (λ²) of the manifest variables was calculated, which is part of the variance that is explained by the theoretical construct (Bollen, 1989). To determine the communality, the square of the correlations between the manifest variables and its own latent variable was calculated. For example, for the AE2 indicator “Ability to communicate verbally and in writing with other people”, there is a factor load of λ = 0.798, which presents a communality of λ² = 0.6368, indicating that 63.68% of the variance of the manifest variable is related to the construct of management skills, a criterion that is above 50%, which is statistically acceptable (see Table 2).

Reliability of the construct

To determine the internal consistency of the indicators that measure the reflective constructs, the reliability of the construct was analyzed through Cronbach’s Alpha (α) and the Coefficient of Composite Reliability (ρ_c). Although 0.700 is an acceptable value for Cronbach’s Alpha and for composite reliability in the first stages of the research, the accepted values for more advanced stages are between 0.800 and 0.900. Values of 0.600 or less indicate a lack of internal reliability (^{Henseler et al., 2009}). Table 4 shows that the Cronbach’s Alpha Coefficient is greater than 0.700 in all cases, as recommended by ^{Nunnally (1978)} and ^{Sanz et al. (2008)}.

Regarding the Composite Reliability or Spearman's Rho coefficient, all reflective constructs have values greater than 0.600 (^{Bagozzi and Yi, 1988}; ^{Chin, 1998}; ^{Steenkamp and Geyskens, 2006}). Similarly, all Composite Reliability Coefficients (ρ_c) are greater than the values of Cronbach’s Alpha for each of the proposed constructs (^{Fornell and Larcker, 1981}), so that the internal reliability of the theoretical concepts is guaranteed. In the same manner, the Average Variance Extracted (AVE) should be greater than 0.500 (^{Bagozzi, 1991}; ^{Fornell and Larcker, 1981}) and significant at the level of 0.01 (^{Sanzo et al., 2003}), which indicates that more than 50% of the construct variance must be explained by its indicators. Therefore, it could be said that the constructs proposed in the model have a satisfactory internal consistency in statistical terms (see Table 3).

Convergent validity and divergent validity

After the validity of the construct in terms of the internal reliability of the indicators, we proceeded to determine the validity of the construct through the analysis of the following statistical tests:

Convergent validity, for the realization of this test the Average Variance Extracted (AVE) was calculated for the reflective constructs, as is suggested by ^{Fornell and Larcker (1981)}. The AVE Reliability Coefficient provides the amount of variance that a reflective construct obtains from its indicators in relation to the amount of variance due to the measuring error of the scales. As shown in Table 4, the AVE Coefficient for the constructs with reflective indicators must be greater than 0.500 (^{Bagozzi, 1991}; ^{Fornell and Larcker, 1981}), which indicates that more than 50% of the variance of each of the theoretical dimensions must explain and measure the indicators. The previous analysis allows us to demonstrate the convergent validity for the proposed model.

Divergent Validity, this statistical test determines if the proposed construct is significantly removed from other constructs which it is theoretically related with (^{Roldán, 2000}). In this sense, the values of the Correlation Matrix between Constructs were examined-which is comprised by the square root of the AVE Coefficient-, and which must be superior to the rest in their same column. This means that the coefficient of internal reliability of the constructs may be greater than the square of the correlations between the latent variables, indicating that with a greater portion of variance between the components of the latent variables, the more differences will exist between the blocks of the measuring indicators (^{Chin, 2000}; ^{Sánchez and Roldán; 2005}; and ^{Real et al., 2006}). According to ^{Sánchez and Roldán (2005)}, to satisfy divergent validity, the indicators on the diagonal (√AVE) must be greater than the indicators below the diagonal (see Table 5). As can be observed in Table 4, not all dimensions comply with the statistical criterion due to the augmentation of indicators by construct, so the divergence between the measurement scales that represent the concepts object of the analysis is theoretically justified (^{Martínez and Martínez, 2009}). The divergent validity of the measurement scales is guaranteed through the aforementioned empirical procedure, based on the theoretical and methodological evidences (content and apparent validity), thus confirming the discriminatory validity of the different constructs that comprised the proposed casual model.

Validation of the structural model

The validation of the structural model is analyzed through two basic indexes (Johnson, Herrmann and Huber, 2006) and the Q² Parameter:

Explained Variance or Coefficient of Determination (R²), this measurement indicator must be equal to or greater than 0.100 (^{Falk and Miller, 1992}). Based on the aforementioned empirical criterion, all constructs possess a satisfactory predictive power for the proposed structural model, R² = 0.948 (see Table 6).

Standardized Regression Coefficients (β), these weights or paths must reach a value of at least 0.200 to be considered significant, as is established by ^{Chin (1998)}. Therefore, the causal relations proposed in the hypotheses of the conceptual model do not comply with the acceptance criterion (H1: β = -0.133; and H2: β = 1.084), but it does comply with the significant criterion (H1: t = 2.461; p < 0.05; and H2: t = 22.498; p < 0.01). The standardized regression coefficients (beta) of the latent variables are levels below the statistical criterion, but given the significance of the relation, it is considered a partial regression (^{Barrera and Vargas, 2005}).

Stone-Geisser test or Q² parameter (Cross Validated Redundancy), determines the predictive power of the model through the blindfolding technique, which indicates that an indeterminate construct is a combination of its indicators plus an error term (^{Stone, 1954}; ^{Barroso et al., 2005}). Furthermore, this indicator must be greater than zero for the construct to have predictive validity (^{Chin, 1998}). Consequently, the proposed structural model complies with the aforementioned criteria, as the Q² parameter is above zero (Q² = 0.656), confirming that the observed values were reconstructed and the structural model has predictive power (see Table 5).

Goodness of Fit Index (GoF)

Finally, the Goodness of Fit index of the structural model was calculated (GoF = 0.7941), which determined that there is a fit of 79.41% to extrapolate the results of the study population to that of the analysis (see Table 6). This indicates that there is a good fit, complying with the empirical criterion that states that the goodness of fit must vary between 0 and 1. The greater the value the better the index (^{Tenenhaus, 2008}).