Entrepreneurial Orientation, Firm Growth and Performance in SMEs: Testing the Scale of EO in SME Context

1.1 The Theoretical Background and Hypothesis

1.2 Entrepreneurial Orientation and Firm Growth

Firm growth is a core element underlying the resource-based view of the firm (Barney 1991) and as such important inside-out (within firm to environment) perspective (Poudel, Carter, and Lonial 2019). Firm growth is combined with a firm’s ability to adapt and survive in an unpredictable business environment. Thus, growth does not happen in a vacuum, but is affected by dynamism, uncertainty, and unpredictable changes in markets (Craig and Moores 2006). The strategic management literature has long acknowledged that the attitudes, managerial philosophies, and personal proclivities among senior managers reflects in firm-level strategic choices (Hambrick 2007). Growth capabilities, in turn, are an outcome of leadership behaviors and management activities combined, developed through the results of interactions and complementarities among individuals, processes, and structures (Felin et al. 2012).

Growth can solve number of problems for SMEs. Frequently, management expresses growth as an important organizational performance goal (Brush, Bromiley, and Hendrickx 2000), because it is generally assumed that the positive development is the result from firm growth. The resource-based view, too, recognizes, the problematic nature of the resource challenges faced by SMEs. For small firms in particular the best manner by which they may improve their competitive situation is to grow (Wolff, Pett, and Ring 2015). So, when it comes to entrepreneurship research, firm growth has turned into the most significant indicator for overall firm success. Thus, we may assume that firm growth is an important dependent variable, especially in SMEs.

Rauch et al. (2009) stated that consistent findings show that entrepreneurial firms grow faster. However, there are ample constraints exist on a firm’s ability to maintain growth, including managerial desire, competitive pressure, resource acquisition, and market changes (Eshima and Anderson 2017). According to Penrose (1959) once growing, firms tend to seek ways to perpetuate growth. Further, prior growth reinforces growth intention (Sorama et al. 2015). According to Miller (1983) as a mechanism is EO. He argued that the dimensions of EO are strongly linked to growth orientation. In a more recent study, Soininen et al. (2013) found that a growth-oriented attitude is positively related especially to innovativeness and proactiveness, while more survival-oriented managers exhibit less risk-taking behavior. Therefore, we assume that the growth-oriented attitude relates success in the long-term and is aimed at pursuing growth in the future based on the time-consuming nature of proactiveness and innovations, while risk-taking is more related to the need to respond quickly to environmental shocks.

However, despite intensive research, many studies have failed to verify the link between growth and EO. This may be because growth is considered a dimension of performance (e.g. Wolff, Pett, and Ring 2015). Especially, when growth is linked to profit performance the connection is not demonstrated. As Gartner (1997) suggested growth may have a negative relationship to profitability. Covin, Green, and Slevin (2006) examined sales growth rate as a performance variable. In their sample of manufacturing SMEs employing over 50 employees, the result was EO has a positive effect on sales growth rate. However, the magnitude of this effect was only marginally significant. Their analyses indicated that risk taking and proactiveness have positive main effects on firm sales growth rate. Interestingly, the study of Soininen et al. (2013) found that firms’ turnover growth rate was positively related to EO.

High growth has often been seen as an indicator of performance, which it may not be, as high growth is rarely a prerequisite for profitability. SMEs lack the necessary resources and suffer from the liability of smallness and thus are more constrained in terms of growth sources (Anderson and Eshima 2013). Despite these shortcomings SMEs can benefit from the active development of EO to develop the capacity to use resources efficiently and to make available those that are not currently available (Wiklund and Shepherd 2003).

Based on the literature, we can assume that proactiveness and innovation are related to the firm’s development and long-term goals. Thus, their effects are reflected in the growth of turnover in the long run. The investments on development also have a negative impact on profitability. However, they are reflected in the development of the firm in relation to competitors. Instead, risk-taking often has a negative impact on profitability, but allows for rapid turnover growth, for example, by establishing new markets or through acquisitions. Therefore, we propose the following hypotheses:

1.3 Entrepreneurial Orientation and Firm Performance

The relationship between EO and performance is one of the most studied topics in the field of entrepreneurship (Saeed, Yousafzai, and Engelen 2014). Despite the impressive number of studies about EO-performance relationship, the results remain inconclusive and contradictory (Su, Xie, and Wang 2015). Many researchers argue that EO-performance relationship is positive (Wiklund and Shepherd 2003), while others claim that the effect of EO on performance is either negative or not significant (Renko, Carsrud, and Brännback 2009; Slater and Narver 2000). Hence, there is no consensus on this matter. Various results may be due to methodological moderators related to measurement and sample characteristics that may affect the relationship between EO and organizational performance. Some differences could be found in the operationalization of EO (Wales, Gupta, and Mousa 2013). In the literature there are two main operationalizations of EO; some authors have addressed the phenomenon based on Miller (1983), while others supported the definition of Lumpkin and Dess (1996).

The predominant argument is that EO plays an important role in enhancing a firm’s performance (Rauch et al. 2009). Wong (2014) noted that EO contributes to the success of new products by allowing firms to identify and proactively take advantage of new business opportunities. However, Renko, Carsrud, and Brännback (2009) found that the EO construct is not a significant predictor of product innovation. According to Lumpkin and Dess (1996) innovativeness reflects the important means by which firms pursue new opportunities. Thus, innovativeness can also mediate the impact of other factors – such as environmental changes – on performance. However, de Oliveira et al. (2018) stated that innovations may not have positive impact on financial gains in the short term. Various researchers suggest that high levels of innovativeness lead to success, while others conclude that proactiveness is the main driver for performance. Furthermore, some researchers suggest that risk-taking leads to success, although an over-exposure of risk-taking is contradictory (Filser et al. 2014). Naldi et al. (2007) found a negative effect of the risk-taking dimension of EO on performance in family-owned business. Especially, the negative effect related to the financial performance of the firm.

Performance can be measured with archival financial performance measures, perceived financial and nonfinancial measures, and some studies combined all three aspects. In studies, EO is positively related especially to a firm’s sales performance (Spillecke and Brettel 2014), profitability in both the short and the long run (Gupta and Gupta 2015), speed to the market (Clausen and Korneliussen 2012), and growth pace creating better chances to mitigate the repercussions of economic recession (Soininen et al. 2012). Also, Rauch et al. (2009) concluded that there are similar relationships with the perceived financial performance and non-financial performance indicators perceived. Among financial indicators, the widely used measures of firm performance are sales growth rate and profitability measures (Soininen et al. 2013). Non-financial performance measures include goal achievement, customer and owner satisfaction, global success ratings and other indicators (Rauch et al. 2009). Additionally, firm performance can be assessed with subjective measures, which are self-reported and reflect managers’ perceptions of their firm’s market performance. Investigations in literature reports that different performance measures (objective vs. subjective) can produce different correlations with independent variables (Evanschitzky et al. 2012). Subjective measures of performance, made by judgmental assessment of respondents appear to be more flexible and reliable than objective ones, facilitating comparisons between firms (González-Benito and González-Benito 2005). Therefore, it can be hypothesized that the strength of the relationship between EO and performance is greater when subjective performance criteria is used (Soares and Perin 2020). Prior research demonstrated the importance of multi-item measures of performance. It is claimed that the constructs’ measurement tends to be better and more precise when more than one indicator is used (Murphy, Trailer, and Hill 1996). Thus, stronger correlations between EO and business performance are associated with the use of multi-item measures of performance.

In many studies investigating the relationship between EO and performance the dimensions of EO show high intercorrelations with one another. Rauch et al. (2009) reported that the correlation between EO and performance is moderately large in their meta-analysis of 51 publications. In cases where the individual dimensions of EO were included and appropriate statistics exist, the strongest correlation was for the innovativeness dimension and the weakest was for risk-taking, but the differences were not statistically significant. However, most studies combined these dimensions into single factor (e.g. Wiklund and Shepherd 2003). Some studies suggested the usefulness of considering EO as a multidimensional construct as a contingency framework for exploring the relationship. Lumpkin and Dess (1996, 2001 stated that the basic premise underlying these arguments is each sub-dimension of EO may have a different relationship with key outcome variables, such as firm performance. Hughes and Morgan (2007) and later Soininen et al. (2013) found that only proactiveness and innovativeness have a positive influence on business performance while risk-taking has a negative relationship. Competitive aggressiveness and autonomy appear not to be influential in business performance or firm growth.

In summary, it can be assumed that innovativeness and proactiveness will increase firm’s performance even in the short term, although their impact will increase over time. Instead, risk-taking is often associated with actions whose effects on the development of the firm will only become apparent later. In addition, performance should be measured by subjective multi-item measures that describe how management perceives firm’s success. Defining success is related to the goals set by the firm’s management. Based on these premises, the following hypotheses are proposed:

2.1 Data Collection

We selected all limited companies from South Ostrobothnia region of Finland identified in the Finnish Voitto+ database and then excluded agriculture, forestry, and construction companies to arrive at an initial sample of 1005 companies. From that sample we targeted obtaining 300 complete survey responses mainly from CEOs and firm owners to assemble a representative sample. The survey instrument was distributed and collated between September 2019 and January 2020. The initial email round produced 102 answers, which prompted a team of six researches and three research assistants to embark on a round of telephone canvassing and personal visits to remind managers about the survey. That effort raised the total number of responses to 306 with a response rate of 30.5%.

To facilitate bootstrapping we removed 46 observations from the dataset because either financial information was not available for three years or the absence of other necessary information, such as name, or other values were missing or because the sample contained duplicate answers from the same respondent. Therefore, the final valid sample was 260 answers.

The respondents were from various industries, including manufacturing (34.1%), services (45.1%), and other related industries (20.8%). The level of analysis in this research is the company level. The firms studied employed between one and 209 employees, with the average staff size being 35 employees. The annual turnover ranged from EUR 200,000 to EUR 85 million. Sales growth in three years varied between a negative growth of EUR 100,000 and positive growth of EUR 1.6 million. Thirty percent of the companies had at least 30% growth in sales, 11% had a 20–29% growth in sales, 11% had a 10–19% growth in sales. Other companies had grown less than 10% in sales or the growth was negative.

2.2 Measurement Constructs

Firm growth is often taken as synonymous of success signaling entrepreneurial behavior (Davidsson, Delmat and Wiklund 2002). However, in literature there are still many different constructs in defining growth. In both theory and practice, turnover growth is the most accurate measure of growth. The increase in the number of employees is often used to measure growth, too. However, in research, employee growth is more difficult to measure accurately. The turnover is publicly available data, but information on the number of employees is based on the firm’s voluntary announcement. In this study, SME growth was operationalized with the financial indicator of growth rate, measured as the percentage change in a firm’s turnover from 2016 to 2018. The measure of turnover growth rate was obtained from Voitto+ the financial statement database. The period from 2016 to 2018 is considered as a short-term in this study.

Performance can be measured by a variety of factors. It is commonly evaluated relative to competitors, such as development of new products and services or market developments. These are usually based on self-evaluation, as accurate data on competitors may not be available. In this study, SME performance was measured with an 11-item scale from Länsiluoto et al. (2019). That scale was in turn an adaption from the scale by Chapman and Kihn (2009) (see also Hyvönen 2007), whose instrument was originally developed by Govindarajan and Fisher (1990) and Chenhall and Langfield-Smith (1998). Länsiluoto et al.’s (2019) modification adds two solvency items (solvency: equity ratio, gearing and cost control) to the instrument of Chapman and Kihn (2009) because they are also key aspects of financial performance. Following Artz, Homburg, and Rajab (2012), respondents were asked to rate their firm performance relative to competitors over the past three years (considered as a short-term in this study). A 7-point Likert scale was anchored with unsatisfactory (1) and excellent (7).

The final scale consists the following items: 1) Return on investment, 2) Profit, 3) Cash flow from operations, 4) Solvency (equity ratio, gearing), 5) Cost control, 6) Development of new products, 7) Sales volume, 8) Market share, 9) Market development, 10) Personnel developments, and 11) Political-public affairs. The composite reliability (0.90) and Cronbach’s alpha (0.87) indicated good reliability for the scale based on the recommendations of Hair et al. (1997) and Nunnally and Bernstein (1994). Thus, the performance scale measures both financial and non-financial success, and is different from our measurement of firm growth, which was measured with an objective measure of change in turnover.

2.3 Scale Development for EO

In our scale development for EO, we combined scale items developed by Soininen et al. (2013), Sok et al. (2017), and Wolff, Pett, and Ring (2015). The original items and the final scale are presented in Table 1. Soininen et al. (2013) stated that the operationalization of the original Covin and Slevin (1989) measure is not well suited to the context of Finnish SMEs. For this reason, in this study to investigate entrepreneurial orientation, we modified the measurement instrument, however retaining the three dimensions of the original Miller/Covin and Slevin EO structure. Also, Soininen et al. (2013) utilized Miller’s (1983) three-dimensional EO conceptualization and Covin and Slevin’s (1991) operationalization of innovativeness, proactivity, and risk-taking while adapting them to the context of Finnish small businesses.

Sok et al. (2017) measured EO using a 18-item scale adapted and refined from Covin and Slevin (1991), George, Wood, and Khan (2001), and Lumpkin, Cogliser, and Schneider (2009). The items scale was anchored with strongly disagree (1) and strongly agree (5). Sok et al. (2017) operationalized EO as a gestalt unidimensional construct capturing the aspects of innovativeness, risk-taking, proactiveness, competitive aggressiveness, and autonomy. These five dimensions were aggregated together to measure EO. The authors proposed that their approach was consistent with prior studies measuring EO as a unidimensional construct (e.g.,Engelen et al. 2015; Engelen, Neumann, and Schmidt 2016; Gupta and Batra 2016; Rauch et al. 2009). In this study, we do not consider the EO construct to be unidimensional and, in line with several previous studies, do consider the original three-dimensional construct describes EO well; accordingly, competitive aggressiveness and autonomy are not presented here.

Wolff, Pett, and Ring (2015) measured EO using a scale modified from Covin and Slevin (1991) and based on prior works of Miller (1983) and Covin and Slevin (1989). Those studies measured EO by asking respondents 12 questions comparing their firm to similar firms in their particular industry. The study delimits the EO construct with three dimensions – innovativeness, proactiveness and risk-taking and employs a 7-point Likert-type scale anchored with strongly disagree (1) and strongly agree (7).

In this study, the final EO scale to be tested was built based on the three studies described above and their operationalizations. However, not all the statements used in the above studies were included in every dimension of the EO measurement, but efforts were made to combine them with compact statements that best illustrate the original conceptualization of Miller but as multidimensional and adapted for SMEs.

In the final structure of innovativeness, we used three items of Soininen et al. (2013) to measure innovative behavior. To measure the proactiveness of SMEs we used three items: one from Sok et al. (2017), one item from Soininen et al. (2013), and one from Wolff, Pett, and Ring (2015). The risk-taking dimension was measured with three items from Sok et al. (2017). This gave us a versatile but compact structure applicable to measure EO among SMEs. We used 7-point Likert scale for the items anchored with totally disagree (1) and totally agree (7).

We conducted two iteration phases to test the scale. In the first iteration, 10 researchers analyzed the content and wording of the questions. For the second phase, we invited 12 managers and owners from both the manufacturing and service industries to evaluate the questionnaire. Modifications to the content were based on the suggestions of both scholars and managers to increase validity and reliability of the questionnaire.

The validity and reliability of the scale were examined for internal consistency, content validity, and construct validity (Bannigan and Watson 2009). Litwin (1995) states that internal consistency is applied to groups of items that are thought to measure different aspects of the same concept. To test internal consistency for reliability, we used Cronbach’s alpha and assessed the resulting reliability rates of 0.70 or better to be acceptable based on Nunnally’s (1978) recommendation. Content validity is closely related to construct validity and refers to the extent to which the content of a scale includes all the relevant issues (Bannigan and Watson 2009). Sim and Arnell (1993) note that no correct content lists are available; and hence, total content validity is impossible to verify. However, a critical review by an expert panel for clarity and completeness can be implemented, and content can be compared to the literature (Bannigan and Watson 2009).

To increase content validity, we enlisted 10 experts, and the results were compared with the literature to achieve authenticity (Messick 1994). In addition, face validity – defined as the subjects’ acceptance of the text (Payton 1988) – was assessed with 12 managers and owners of companies as described earlier. The measurement tool must be understandable and perceived to be relevant by the subjects. Therefore, the relevance of the scales was assessed and tested with a group of managers representing the respondents. Construct validity can be assessed through factorial validity (Bannigan and Watson 2009). We used explorative factor analysis to test the factorial validity of the scales, before subsequently conducting CFA with structural equation modelling.

2.4 Preliminary Analysis

Preliminary analysis was conducted with SPSS 21. First, we tested the skewness and kurtosis of the scale items and applied an acceptable limit of ±2 indices (Field 2000, 2009; Gravetter and Wallnau 2014; Trochim and Donnelly 2006). Exploratory factor analysis was used to test the construct validity of the EO scale. The Kaiser–Meyer–Olkin Measure of Sampling Adequacy (0.881) and Bartlett’s Test of Sphericity with a significant p-value (0.000) indicated a good suitability for the factor analysis. The factor analysis resulted in two factors with eigenvalues more than 1. Communalities varied between 0.542 and 0.743. The first factor explained 59% of the variance and the second factor 13%.

The first five items (items related to innovativeness and proactiveness) loaded to the first factor. These items reflect the dimension of innovativeness. Item 6 loaded to both factors and was consequently omitted from the final model. The remainder of the items loaded to the second factor, which reflects the dimension of risk-taking. Table 2 presents the factor loadings. The result was a two-factor model that could be examined with structural equation modeling. The factor analysis suggested that there are only two factors in EO: innovativeness and risk-taking. The items related to proactiveness loaded on the innovativeness factor.

2.5 CFA with Structural Equation Modeling for EO Scale

Validating the measurement instrument involved designing a first-order CFA model to test the multidimensionality of EO construct based on the results of the exploratory factor analysis. The model postulates a priori that EO is a two-factor structure composed of innovativeness and risk-taking. We used different measures suggested by Byrne (2010) to evaluate the model fit. Acceptable model fit was operationalized as X2/degrees of freedom (df) ratios (CMIN/DF) less than 3.0, Comparative Fit Index (CFI) values greater than 0.90, Normal Fit Index (NFI) values greater than 0.95 and Root Mean Square Error of Approximation (RMSEA) values less than 0.08. A sample should have around 100 to 200 observations to merit confidence of fit tests (Hoyle 1995). In addition, Mitchell (1993) recommends that a model should contain 10 times as many observations as variables. The sample size of 260 used in the current study can therefore be considered adequate as our model has 18 variables.

Brown (2006) suggests it is important to consider the model’s parameter estimates when determining the model’s goodness of fit. We examined each item’s R2 and factor loadings and accepted values greater than 0.40. Brown (2006) notes that a low R2 for an item indicates high levels of error and is considered an estimate of the item’s reliability. The modification indices (MI) were examined in order to find any misspecifications in the model. The MI suggested correlation of some error terms (indicated by an MI value of above 10). Since the correlation of error terms in the same dimension can be accepted (Byrne 2010), the correlations of error terms were included in the model. We added a correlation between the error terms of items Inno1 and Inno4, between the error terms of items Inno2 and Inno3, and between the error terms of items Inno3 and Inno5. The model fit was good for modified two-factor CFA model; CMIN/DF 1.97; CFI 0.98; NFI 0.98; RMSEA 0.061. Figure 1 presents the final model with standardized estimates.

Figure 1:

Final CFA model with standardized estimates (R2 of items, regression weights and correlations).