5.1. Grouping of Enterprises Using the Clustering Method
The first stage of the analysis was to distinguish groups of enterprises according to their environmental and social goals. We have chosen environmental and social goals since these two aspects determine whether a given enterprise pursues sustainable entrepreneurship, the essence of which is to conduct business not only bringing a specific economic profit but also enabling the fulfillment of environmental and social goals [
41,
62,
63,
64,
65,
66].
Two selected questions of the survey focused on the implementation of environmental goals (marked as WE.1., WE.15) and social (WS.1., WS.14) consistent with the concept of sustainable entrepreneurship (
Table 2). Answers to these two questions adopted a scale of 1–6 (6—definitely yes, 5—yes, 4—rather yes, 3—probably not, 2—no, 1—definitely no).
The results obtained from the questions on a six-point scale allowed the calculation of indicators showing the company’s involvement in activities consistent with the concept of sustainable development:
The indicators for the areas mentioned above were calculated by summing up the Likert scale points. As a result, we received a table in which we calculated two indicators for each of the 410 enterprises. The higher value of each indicator illustrates the better implementation of sustainable entrepreneurship activities.
The highest possible values of the WE and WS are, respectively, 90 and 84 (if 6 is the answer to each question). Moreover, the lowest 15 and 14 (if 1 were the answer to each question). Only two enterprises achieved the highest values of both indicators. There were no enterprises that would achieve minimum values (the lowest values achieved by one enterprise were WE = 15 and WS = 19).
The next step was to identify Polish enterprise’s characteristic attitudes towards implementing the SE concept and compare the results with the company’s self-assessment. Here, we took the first question of the survey into account, which was to establish how companies perceive their activities in the light of the presented definition. Possible answers to the questions included the options: Yes, no, I do not know.
The enterprises’ characterization based on answers to the two questions was to determine their behavior. The self-assessment question was to illuminate the respondents’ awareness in the area of sustainable entrepreneurship.
The research focused on identifying the similarities between the studied enterprises, determining the specificity of the selected groups, and then identifying their strengths and weaknesses in implementing the concept of sustainable entrepreneurship.
The implementation of this goal and the size of the data set determined using the cluster analysis method, i.e., unsupervised classification, which consists of assigning individual examined objects to separate groups. No predetermined division was assumed, while the study focused on searching for similarities between enterprises and, on this basis isolating specific characteristic patterns of behavior.
The data were explored using the K-means algorithm which is one of the most popular unsupervised machine learning algorithms. This method decomposes a dataset into a set of disjoint clusters. Unsupervised algorithms make inferences from datasets using only input vectors without referring to known or labelled outcomes. The K-means algorithm identifies the k number of centroids, and then allocates every data point to the nearest centroid. The process is performed iteratively until an optimal partition is reached.
Given a set of observations (x
1, x
2, and x
n), where each observation is a d-dimensional real vector, k-means clustering aims to partition the
n observations into
k (≤
n) sets
S = {
S1,
S2, and
Sk} so as to minimize the within—cluster sum of squares (WCSS) (i.e., variance). The objective is to find:
where
μi is the mean of points in
Si. This is equivalent to minimizing the pairwise squared deviations of points in the same cluster.
We used the Orange [
67] software to perform the analysis.
Figure 2 presents the stages of data mining and the tools used.
The data mining tool proposed various possible data divisions and the generated cluster’s quality indicator called Silhouette Score (
Figure 3). A larger Silhouette Score informs about a more precise assignment of elements to the group (more significant similarity of elements in the cluster).
The best match index was found in the division into two, then into five clusters. We chose the division into five clusters due to the more detailed possibility of semantic interpretation of the results.
The input data were the values of the two indicators, as mentioned earlier, calculated based on each respondent’s responses.
To better assess the cluster’s quality, we visualized them using the MDS algorithm. The visualization shows that the groups are relatively homogeneous (clearly distinguished color areas, objects from the same cluster are gathered close to each other and paired).
The MDS algorithm enables the visualization of the features of the tested objects. The figures below show WS (
Figure 4) and WE (
Figure 5) values as the marker size (the better the WE or WS index, the bigger the symbol).
Therefore, enterprises were divided into five clusters due to the specificity of their activities in the social and environmental field. The number of enterprises in individual C1–C5 groups is presented in
Figure 6.
The size of the research sample is representative of the population of small and medium-sized enterprises in Poland. Therefore, it can be assumed that the division into groups reflects the characteristics of the studied population.
Additionally, to facilitate the interpretation of clustering results and increase the division’s clarity, the indicator’s values have been divided into three groups: Low, medium, high. The division was performed using the Discretize algorithm available in the Orange software. The method of discretization according to the equal frequency was chosen, which considers the data set’s nature when partitioning (
Table 3).
Five groups of enterprises have been distinguished, characterized by a distinctly different approach and sustainable development implementation.
Table 4 presents a short characteristic of the selected clusters with their classification according to the level of sustainability.
The analysis of enterprises grouped into clusters revealed their behavior in the ecological and social areas. We analyzed the answers to the question “Does the enterprise operate following the concept of sustainable entrepreneurship.” The answer to this question was to highlight the enterprise’s self-awareness of sustainable development (
Table 5).
As it could be assumed, the enterprises from cluster C5, which best implement the PE concept, also assess their activities correctly. Only in the case of two enterprises from this group, the answer was “no”. In clusters C2 and C3, assessed as implementing the SE concept on average, the responses are similar—about 60% of enterprises have confidence that they implement this concept well. On the other hand, in clusters C1 and C4, whose activities were assessed poorly, about half of the enterprises wrongly consider their sustainable activities. A high percentage of “I do not know” answers in all groups leads to the conclusion that Polish enterprise’s knowledge about sustainable entrepreneurship is insufficient. The next task was the semantic interpretation of the division of enterprises into groups. For this purpose, we compiled data describing cluster’s characteristics in the form of tables and charts.
5.2. Semantic Interpretation of Clustering Results
For facilitating the interpretation of clustering results, a summary of answers to the questions was prepared, broken down into individual clusters (
Figure 7 and
Figure 8). The visualization presents the number of negative (1–3) and positive (4–6) responses that made up WS and WE indicators. We calculated the sum of the negative (1–3) and positive (4–6) answers for each question. We visualized the percentage of negative and positive responses as data bars to make the problems more comfortable to see. The red color of the data bar in the table represents the percentage of negative responses, while green represents positive responses.
There were 75 companies in cluster C1 (18% of the research sample). The list of indicators (
Table 6) shows that these enterprises’ main problem is the low achievement of social goals, while environmental goals are poorly or moderately achieved. Five enterprises achieved a high rate of environmental goals implementation. Nevertheless, it was still at the lower limit (61 or 63 points), while the WS ratio for these enterprises was low (
Table 6).
Therefore, we can assume that the enterprises from cluster C1 are not very good at implementing sustainable entrepreneurship. We analyzed the answers to the questions to diagnose specific problems.
As can be seen in
Table 6, negative responses prevail in nine out of 15 questions. The highest number of negative answers took place in the case of the following questions:
WE.1. Use of renewable energy sources.
WE.13. Use of devices regulating water consumption (aerators, stream pressure regulators, timer stops, photocells).
WE.6. Reduction of radiation, hum, and noise related to the use of electronic equipment.
In the case of questions on social goals, the highest number of negative assessments occurred in the case of the following questions:
WS.13. Conducting activities for the development of local culture.
WS.14. Participation in charity events.
WS.12. Pro-social activity.
WS.4. Conducting activities for the benefit of the local community, e.g., patronage or sponsorship of cultural events.
In cluster C2 (
Table 7), some enterprises cope well with the implementation of environmental goals (lack of low indicators), and on average, with the implementation of social goals.
In the case of questions on implementing ecological goals, the number of negative answers prevailed only for two problems. As in the case of the previously analyzed group, these were the following questions:
WE.1. Use of renewable energy sources.
WE.13. Use of devices regulating water consumption (aerators, stream pressure regulators, timer stops, photocells).
The implementation of social goals in the group of enterprises from cluster C2 is much better. Only one of the questions was slightly dominated by negative answers. It was the question WS13 about activities for the development of local culture. The second question with the most negative answers was WS4—activities for the benefit of the local community, e.g., patronage or sponsorship of cultural events. Based on the answers to the questions, we can conclude that these enterprises are more focused on supporting their employees.
Cluster C3 (
Table 8) represents enterprises that cope poorly or moderately with implementing environmental goals (no positive ratings) but much better with implementing social goals (no negative ratings).
Comparing the relationship between negative and positive assessments (
Figure 7) revealed that enterprises have problems implementing most ecological goals (nine negatives out of 15).
More than 60% of negative answers to the questions:
WE.1. Use of renewable energy sources.
WE.3. Reduction of working time of electronic devices.
WE.13. Use of water intake regulating devices.
WE.4. Use of efficient heating and emission-reducing systems.
WE.6. Reduction of radiation, hum, and noise related to the use of electronic equipment.
WE.11. Producing and selling environmentally friendly products.
On the other hand, the answers to the questions on achieving social goals were very positive. In not even one case, the negative opinions outweighed the positive ones (
Figure 8).
The respondents gave 100% of positive answers to the following questions:
WS.5. Taking care of safety in the workplace.
WS.8. Striving to create a friendly atmosphere at work.
WS.11. Development and application of accepted ethical standards related to the conducted activity.
Thus, we can conclude that the analyzed C3 group of enterprises is pro-social and oriented with particular care for the employee’s welfare.
Cluster C4 represents enterprises that are the worst at implementing sustainable entrepreneurship. The implementation rate of environmental goals was low in all cases, while achieving social goals was low in 80% (
Table 9).
After a closer look at the answers to the questions from the category of ecological goals (
Table 9), we can see that negative opinions prevail in all cases (69% or more). There is an overwhelming number of extremely negative assessments (1 and 2).
Answers to questions from the category of social goals are better than the answers on environmental goals. Similar to other clusters, in this group, the goals related to the development of culture (WS.13), pro-social activity (WS.12), and activities for the local community (WS.4) were also the worst.
Cluster C5 represents leaders of sustainable entrepreneurship in the surveyed group. Unfortunately, it constitutes only 14% of all respondents.
Table 10 presents the ranges of the EC and WS indicators for cluster C5.
The same problem was revealed in the implementation of environmental goals by enterprises from cluster C5, which occurred in other groups—56% of the respondents gave a negative answer to the question on WE.1 (use of renewable energy sources). The implementation of social goals is excellent in the analyzed group. One hundred percent of the answers were positive for eight out of 14 questions and over 72% for the remaining ones. There were also individual negative responses among the group of leaders. In social goals, these were the questions: WS.1—hiring employees based on employment contracts and WS.2—offering work to people from the closest social environment. Enterprises also gave negative assessments (two and three) in question WS.13—carrying out activities to develop local culture. Based on the research results, we can conclude that environmental goals are a bigger problem for Polish enterprises than implementing social goals. The chart (
Figure 9) shows the percentage number of responses on a scale of 1–6 for the categories of social (WS) and environmental (WE) goals.
The environmental problem that was most visible in all the groups identified in the study was the use of renewable energy sources. As the prices of systems for acquiring energy from renewable sources are systematically decreasing, such solution’s profitability grows. Therefore, we can assume that this problem may be caused not by the reluctance or little knowledge of entrepreneurs but by the lack of optimal legislative proposals on the Ministry of Energy. The second ecological problem highlighted was the use of devices regulating water consumption (aerators, stream pressure regulators, timer stops, photocells). It is a significant problem for the entire economy. One of the priorities of the environmental policy actions to rationalize water management should be [
68]. In this area, in the country’s economy, it is necessary to regulate the Polish system of supporting ecological investments and disseminate knowledge on the importance of the problem and obtaining funds by enterprises. Other ecological problems visible in the research results are the decrease of radiation, noise, and noise related to electronic equipment and reduced working time of electronic devices. Solving these problems will contribute to the improvement of working conditions and the reduction of electricity consumption costs.
The most critical problems in achieving social goals are the low activity of enterprises in terms of activities for the development of culture and the local community, e.g., patronage or sponsorship of cultural events and charity auctions. These are important activities that should be adequately stimulated by the state (e.g., through an appropriate tax policy). Polish enterprises are more focused on social aspects related to employee’s care, compliance with the labor code, maintaining a good atmosphere in the team, and ensuring safety.
5.3. Identification of Differences in the Approach to the Business Activity of Enterprises
The next part of the study focused on distinguishing differences in the business activity of enterprises. The characteristics that were not subject to clustering were analyzed. For this analysis, two clusters, C4 and C5, representing extreme levels on the scale of sustainable entrepreneurship implementation, were selected. The analysis aimed at examining the characteristics of enterprises that perform best and worst sustainable entrepreneurship. In particular, the analysis of leader’s attitudes indicates the best practices among Polish enterprises.
RQ1: Is there a difference in the innovativeness of enterprises from clusters C4 and C5?
H1: The enterprises from cluster C5, which are leaders of sustainable entrepreneurship are more innovative than enterprises from cluster C4.
The implementation of sustainable entrepreneurship requires changing the established patterns of action and new technical solutions associated with investments. Thus, innovation has an economic and organizational dimension. To examine innovativeness, the authors interpreted the responses to the selected survey questions (
Table 11):
Selected questions allow examining the features related to innovation. At the same time, we checked whether there was a significant correlation between the answers (
Table 12). A high correlation (0.6–0.8) may indicate that the questions have a very similar scope and describe the same phenomenon. Correlations were calculated for the entire studied population of 410 enterprises.
The highest correlation with the value of 0.6 was observed only in questions 6.13 and 6.14. However, it was assumed that it is still a permissible value. Information technology enables the monitoring and measurement of business processes. Accurate and up-to-date data is the basis for making the right decisions. Information technology also brings significant value in terms of optimizing operations and improving overall efficiency. Lower costs, less time needed to complete the process, and less manual work are examples of IT application’s benefits. With fault tracing, monitoring, and predictive maintenance, industrial and commercial components significantly have longer life cycles, leading to less waste. The implementation of advanced information technologies is an innovative activity. When analyzing the enterprise’s responses, we can notice that sustainable entrepreneurship leaders often use advanced IT solutions (
Figure 10). In cluster C5, 77.97% of the respondents gave a positive answer, while in cluster C4, only 35.42%.
Red color in the table means the percentage of negative answers (1–3), while green color means the percentage of positive answers (4–6).
Striving to improve cooperation between employees is another activity that is directly related to the concept of sustainable entrepreneurship and, at the same time, proves an innovative approach to business (
Figure 11).
There are also significant differences in the enterprise’s answers from clusters C4 and C5 in this area. All companies from the C5 group of leaders try to improve the company’s cooperation with 44% declaring a strong commitment in this area. Better collaboration between the company’s employees also contributes to increasing their involvement in sustainable entrepreneurship activities and their innovativeness. The enterprise’s innovativeness depends on the amount of expenditure to develop new solutions and the quality of cooperation between employees. Good collaboration between the company’s employees also means a high degree of trust, mutual understanding, and a joint effort to transform the organization into a sustainable enterprise. Moreover, the simplification of organizational procedures can be considered innovative, which is related to the restructuring of processes that requires an unconventional approach. The vast majority of cluster C5 enterprises strive to simplify and standardize organizational procedures (
Figure 12). Therefore, we can conclude that this feature is characteristic of sustainable enterprises.
The ability to launch innovation quickly was another phenomenon studied. The answers to the question (
Figure 13) directly show the enterprise’s attitudes to implementing innovations and the perception of their role in the business activity. In this area, we can notice a significant difference between companies from clusters C4 and C5.
The question related to the possibility of using information, which is currently one of the most valuable sources of knowledge in a company.
Figure 14 presents the answers provided by enterprises from clusters C4 and C5.
Both the acquisition and generation of information by an enterprise are associated with sustainable entrepreneurship. Information obtained from the external environment increases the employee’s knowledge, enriching the intellectual capital, enabling them to operate following sustainable entrepreneurship. On the other hand, the enterprise’s information and transfer to the external environment also affect the implementation of activities consistent with sustainable entrepreneurship, most often in the ecological and social dimensions.
Another question was concerning ecological solutions, in line with the idea of sustainable entrepreneurship. Possible answers are: We do not have it, we plan to implement it, we have it (
Figure 15). These solutions applied consistently can significantly improve the natural environment’s condition and indirectly affect the company’s image. Red color in the table means the percentage of negative answers (1–3), while green color means the percentage of positive answers (4–6). Yellow represents neutral responses.
In the case of both clusters, the answers to the questions regarding the implementation of the zero-paper principle, storing documents in electronic form rather than paper, and promoting the use of bicycles or car sharing in commuting are puzzling. As can be seen in both clusters, a small percentage of enterprises apply these principles, and many do not plan to implement them. In the first case, we should bear in mind that implementing the “zero paper” principle is almost impossible since a company operating on the market, cooperating or contacting other business entities or offices, must provide certain documents in a paper version. The implementation of this principle presupposes striving to reduce the paper consumption level rather than eliminating it, which is impossible to meet nowadays. On the other hand, the lack of intention to promote bicycles or carsharing as a means of commuting may result from the fact that enterprises are unwilling to interfere with the way employees commute to work.
Moreover, while bicycles are becoming an increasingly popular means of transport, the idea of carsharing is still relatively new and yet rarely used. In the next step, we analyzed answers to the question concerning innovative activities. The respondent’s task was to indicate how the expenditure on innovations has developed over the last 5 years and the number of implemented innovations. The correlation coefficient between answers was 0.794933. Therefore, we decided to analyze only the question on the number of implemented innovations (
Figure 16). The data bars in the table represent the percentage of respondents who marked a given answer on the questionnaire.
A certain stagnation can be seen in this issue, as almost half of the respondents indicated no change. The answers show the still low level of innovativeness of Polish enterprises. Despite access to many programs supporting innovations, Polish enterprises still show a lower expenditure on innovative activities than others in other EU countries. This outlay, amounting to 0.44% of GDP, places Poland in only the 21st place among the EU countries. Bearing in mind that innovation may be one of the determinants of sustainable entrepreneurship. It can be assumed that increasing the level of expenditure on innovative activities, resulting in a more significant number of implemented innovations, will also contribute to the complete implementation of sustainable entrepreneurship assumptions. Summing up, we can notice that enterprises from cluster C5 are more innovative, which confirms the H1 hypothesis. They more often aim at using advanced IT resources, harnessing the power of information, they want to start innovations faster, and can boast a greater number of implemented innovations.
The second analyzed aspect of activity concerning economic issues was to check how enterprises perceive their economic situation.
RQ2: Is there a difference in the economic situation of enterprises from clusters C4 and C5?
H2: Sustainable enterprises from cluster C5 are in a better economic situation than unsustainable enterprises from cluster C4.
RQ3: Is there a difference in the perception of the future situation by enterprises from clusters C4 and C5?
H3: The sustainable enterprises from cluster C5 are more optimistic about the future.
For this purpose, the answers to the following questions were interpreted:
Enterprise profit for the last 5 years:
Comparing the survey results shows that companies that are leaders in sustainable entrepreneurship do much better economically. For the most part, they recorded a significant or slight increase in profit (
Figure 17 and
Figure 18). Therefore, the H2 hypothesis can be confirmed.
The obtained results indicate that the distribution of responses is relatively proportional. However, the lowest percentage of responses showing a significant improvement in the country’s economic situation stands out. Such various responses may result from the caution of entrepreneurs who prefer to adopt a conservative attitude (
Figure 19).
Enterprises are rather pessimistic about the distant future. Almost half of the companies from both groups believe that the economic situation will deteriorate in the next 5 years. A large number believe that it will remain the same. The rest of the respondents think that the situation will improve. This perception of the future may result in the entrepreneur’s reluctance to act in line with sustainable entrepreneurship. The obtained results can also prove a conservative attitude among entrepreneurs who prefer to assume that the economic situation will deteriorate to be prepared for the worse than to assume its improvement. Therefore, hypothesis H3 on the optimistic attitude of entrepreneurs from cluster C5 should be rejected.
We also analyzed how the surveyed organizations perceive sustainable entrepreneurship in business (
Figure 20).
RQ4: Is there a difference in the perception of the role of sustainable entrepreneurship in both surveyed groups?
H4: Enterprises from cluster C5 treat sustainable entrepreneurship as a key element of competitiveness, while the enterprises from cluster C4 do not appreciate it at all.
The content of the question was: What do you think will be the impact of Sustainable Entrepreneurship on the activities of enterprises in the next 5 years (multiple answers):
More and more companies will carry out their activities in the spirit of sustainable entrepreneurship.
Sustainable entrepreneurship will become the standard in which economic processes will run.
There will be a mixing of concepts with the adjective “balanced.”
Failure to consider the concept of sustainable entrepreneurship will contribute to the enterprise’s failure in the market.
Sustainable entrepreneurship will also affect the attitudes of recipients of services/products offered by the enterprise.
Sustainable entrepreneurship will contribute to increasing the level of innovation in enterprises.
Nothing will change.
Hard to say.
Most companies from both groups believe that more and more enterprises will implement their activities in the spirit of sustainable entrepreneurship. As it results from the literature studies, the actions that, apart from environmental effects, also have a specific economic impact, e.g., replacing light bulbs with energy-saving ones, are part of sustainable entrepreneurship. Therefore, we can assume that sustainable entrepreneurship activities will be carried out over time by every enterprise operating in the market. Many surveyed companies believe that there will be a mixture of concepts containing the adjective “sustainable.” It can be assumed that sustainable development, sustainable production, sustainable enterprises, and sustainable construction will converge, giving rise to new ideas and assumptions. Enterprises from both groups indicated that sustainable entrepreneurship would also affect the recipients of their services and products. It will contribute to disseminating a typical attitude of a sustainable consumer and the phenomenon of sustainable consumption itself. Moreover, companies believe that sustainable entrepreneurship will over time become the standard in which economic processes will run, as in the case of sustainable development, the assumptions in which now take the form of regulations or requirements that companies have to meet. Sustainable entrepreneurship will increase the innovativeness of enterprises since it will force the implementation of innovative solutions. Moreover, the enterprise’s innovativeness may translate into implementing this concept’s assumptions in the broader scope.
On the other hand, the smallest percentage of enterprises from both groups assumes that not considering the sustainable entrepreneurship’s assumptions may lead to the enterprise’s failure in the market. A detailed analysis of the answers to the survey questions allowed us to reject the H4 hypothesis. It can be concluded that enterprises in both groups perceive the role of sustainable entrepreneurship in a similar way.
The modern economic reality is highly profitable and market-oriented. Moreover, many consumers are not aware of the global need for sustainable practices and choices. Therefore, we can safely assume that a company that does not meet the principles of sustainable entrepreneurship but offers an attractive product for a reasonable price will still find customers and will be able to survive in the market and develop. However, quite a large group of respondents indicated answers such as “hard to say” and “nothing will change”, which may still show an individual ignorance of this concept, as well as a lack of willingness to implement its assumptions.