1 Introduction

Digitalisation is the embedding of technology into business processes while digital transformation encompasses the embedding of technologies into the entire business ecosystem (Saariko et al. 2020). Digital transformation has completely changed traditional business models (Verhoef et al. 2021) and the embedding of technologies in companies creates competitive advantages that help improve business performance (Blichfeldt and Faullant 2021). This has led to the creation of digital business ecosystems (Kanda et al. 2021).

Digital business ecosystems are an increasingly desirable context for organisations seeking to embed digital transformation (Wang 2020). These ecosystems rely on digital platforms that bridge technologies to achieve the digitalisation of organisations (Kohtamäki 2022) and drive companies to further introduce new technologies and promote innovation in business models (Yuana et al. 2021). In addition, these tools have developed a change in the relationships with both customers and employees, transferring it to the digital ecosystem (DiPietro et al. 2020; Sun and Zhang 2021), i.e., they contribute to greater stakeholder satisfaction and can be transformed into an improvement in the economic performance of the company (Parise et al. 2016; Castellacci et al. 2019).

Previously, there have been studies that relate digital technologies to job satisfaction, but to the best of our knowledge, there are no studies that relate digitalization to improved customer (outside the company) and employee (within the company) and leads to an improvement in business results. In other words, there is a gap in the literature that measures whether, thanks to digitization, it is possible to improve intra-company and extra-company satisfaction, and this in turn has an impact on improving business performance.

Understanding the connection between digital business ecosystem, satisfaction and performance, a question arises that we seek to answer with this research: Does employees and customer’s satisfaction with their organization’s digital business ecosystem affect their organisation’s performance?

Therefore, the main objective of this research will be to determine the satisfaction of employees with the technologies implemented in organizations and if this satisfaction has an impact on business performance.

From a methodological point of view, in order to achieve the proposed objective, a theoretical model has been developed that connects the digital ecosystem of the companies in the sample with the satisfaction of two of their stakeholders and with their performance or profitability. To this model we have applied the PLS-SEM technique (Partial Least Squares Modelling) and to ensure the reliability of the sample we have applied the FIMIX-POS technique (Prediction Oriented Segmentation), on a sample of 1319 Spanish SMEs. The PLS-SEM technique has proven useful for its ability to model composites and factors, as well as its predictive and structural equation modelling orientation (Henseler et al. 2016a, b in Nitzl et al. 2016; Hair et al. 2019; Sarstedt et al. 2020).

As a result, our model relates the digital business ecosystem of SMEs to customer and employee satisfaction and allows us to identify the impact of customer and employee satisfaction on the performance of the sample organisations.

The main contributions of this research are, firstly, that the satisfaction that employees perceive thanks to technology is slightly higher than that of customers. Secondly, and as a more relevant contribution, this research has allowed us to identify that the performance of an organisation improves mainly when the employee feels satisfied with the technologies implemented in the organisation. As a final contribution, we have been able to divide the sample into two homogeneous groups thanks to the prediction-oriented segmentation carried out by the Smart PLS tool, finding the similarity that employee satisfaction has a positive impact on business performance, thus reinforcing the results obtained in the first part of the analysis. As a final contribution, it provides a new vision of the research topic, as well as new results, because it is a topic that is currently under development and there is no extensive literature on it.

Furthermore, in practical terms, the insights from our study will help companies with characteristics similar to those in our sample to improve their business results through greater customer and employee satisfaction.

Our theoretical model strives to find out if business performance improves thanks to an improvement in the satisfaction of certain stakeholders. To the best of our knowledge, there are no previous studies that measure this in the way that is done here.

Our findings will offer valuable indications and recommendations for policy makers, society, and researchers, promoting a greater appreciation of the satisfaction of clients and employees of organizations.

This paper is structured in 4 main parts, firstly, the introduction where we identify the economic problem that gives rise to the research question, the objectives to be achieved as well as the contributions. Secondly, the theoretical framework thanks to which we can develop our theoretical model with the current literature. This is followed by the empirical framework in which we describe the methodology and show the results of the analysis of the model, as well as the relationship of the results with the literature (discussion). Finally, the conclusions of the study are presented, as well as the limitations found and future lines of research.

2 Theoretical framework

More than a decade ago, the literature showed that digital transformation affects the entire enterprise (Zott et al. 2011). But the first thing to know is how the literature defines this term. In this line, Hinings et al. (2018) define digital transformation as the result of different digital innovations that generate new actors, structures, practices and values, which can change, disrupt or complement established norms in organisations or industries, such as removing barriers to entry and exit from current markets (Li et al. 2018).

Guinan et al. (2019) states that digital transformation is the ability of an organisation to adapt, respond and position itself to the success of rapidly evolving technology, creating competitive advantages in the organisation.

Therefore, digital technologies can help to achieve a competitive advantage by transforming the organisation to take advantage of existing competences or develop new ones (Liu et al. 2011). Along these lines, we find some technologies that have become essential for the development of business activities and for customer relations, some of which are highlighted below (Fig. 1):

Fig. 1
figure 1

Source: Own Elaboration

Technologies within the DEE.

Full size image

These technologies and others have led to digital transformation in organizations, and thus to the formation of digital business ecosystems (Winkelhaus et al. 2022).

A digital business ecosystem (hereafter DBE) is an economic community of interacting organisations and individuals producing valuable goods or services through digital tools (Moore 1993). Later authors such as Nachira et al. (2007) define the DBE as a virtual environment composed of digital entities such as software applications, hardware, and processes.

DBE refers to a virtual environment that connects a technological infrastructure with digital services through shared digital platforms (Nachira et al. 2007). So the ultimate goal of DBE is value creation through digital platforms (Senyo et al. 2019).

Digital platforms are defined as “a technological architecture that enables the development of its IT functionalities and allows the integration of information, computing and connectivity technology into platforms available to an organisation” (Sedera et al. 2016, p. 4).

Previous studies have revealed that digital platforms have helped companies to develop value-creating activities by enhancing their digital transformation processes (Warner and Wägner 2019). Digital platforms enable companies to apply new technologies and build new connections with existing equipment (Cenamor et al. 2019) and promote the innovation of their business model in terms of the learning process, training employees in gaining new knowledge and establishing new interaction capabilities (Li et al. 2020).

Such platforms are typically created and owned by a single company or entity and are part of an ecosystem composed of people, data, processes, and objects connected by the use of technologies that go beyond the scope of a single system (Henfridsson and Bygstad 2013). They thus generate services that are radically changing existing ways of working, altering internal ecosystems and business structures (Montealegre and Iyengar 2021).

These changes have influenced both positively and negatively, and one way to analyse the mode is by measuring satisfaction (Farivar and Richardson 2021).

Satisfaction is the consumer’s response to wish fulfilment. Customer satisfaction “is a judgment about a characteristic of a product or service, or that the product or service itself, provides (or is providing) a pleasurable level of consumption-related satisfaction, including levels of under/over satisfaction” (Oliver 1993, p. 13). This definition implies that satisfaction is a positive state of consumers when purchasing a product or service while it satisfies their desires, including the degree of delight beyond the desired level (Brill et al. 2019). And nowadays, Digital Business Models focused on customer experience are gaining importance, as digital transformation can result in improved firm performance and new ways of creating value (Hanelt et al. 2021).

According to Kotler et al. (2017b), satisfaction is given by several factors, and one of them is improved accessibility to products, increased communication with the company and easy access to company information. This is closely linked to the insertion of technologies in the company and the use of digital platforms (Minh Duc, 2022).

Employee satisfaction can be defined as an individual’s subjective viewpoint that encompasses how they feel about their job and the organisation that employs them. In terms of job satisfaction, the subjective view is related to the business processes and tasks in which employees operate.

DiPietro et al. (2020) distinguish two types of job satisfaction; firstly, the general feeling of satisfaction with the job, and secondly, feelings about aspects of the job such as benefits, salary, and work environment (Abuhashesh et al., 2019).

Several studies claim that satisfied employees tend to have higher productivity and high performance (Ilies et al. 2009). Importantly, retaining satisfied employees can help the company to reduce costs (Grissom et al. 2012) and achieve higher performance in the organisation.

At this point we propose to define the following study hypotheses:

H1 Digital business ecosystem influences customer satisfaction.

H2 Digital business ecosystem influences employee satisfaction.

H3 Customer satisfaction influences organisational performance.

H4 Employee satisfaction influences organisational performance.

2.1 Proposed theoretical model

Figure 2 below shows the proposed model, which shows the proposed relationship of the digital business ecosystem with customer and employee satisfaction, and how these in turn influence organisational performance.

Fig. 2
figure 2

Proposed theoretical model

Full size image

The DBE is made up of twelve elements and the application of this model is intended to achieve the main objective of this research, which is to determine whether employees are satisfied with the technologies implemented in organisations and whether this satisfaction has an impact on business results.

2.2 Empirical framework

As for the empirical study, we will first describe the fieldwork. For this, part of the questionnaire by Durende et al. (2022) has been used, which has been validated by the FAEDPYME network of researchers (Faedpyme 2022), and a company specialising in business surveys has been contracted for data collection, using the CATI system. The empirical study is based on a survey of 1319 SMEs in Spain. The questions in the questionnaire that collect the necessary information can be consulted in Table 1.

Table 1 Questions to form the indicators that make up the constructs.
Full size table

To carry out the data analysis, we have used the Partial Least Squares (PLS) technique, which usually is developed in 2 steps, but we have added one step more, than it is Prediction-oriented segmentation analysis for better the results:

  • Assessment of the global model

  • Assessment of the measurement model

  • Prediction-oriented segmentation analysis.

In addition, in order to know the effect and the relationship between these variables we have based ourselves on the use of Structural Equation Modelling, resorting to the PLS technique, using the SmartPLS 3.2.8 tool.

Different authors highlight PLS’ ability to model compounds and factors (Henseler et al. 2016a, b in Nitzl et al. 2016; Marín-García and Alfalla-Luque 2019; Sarstedt et al 2020) and its prediction orientation of more complex models that are possible to deal with thanks to the emergence of techniques such as Structural Equation Modelling (SEM) (Shmueli et al. 2016; Hair et al. 2019; Sarstedt et al. 2020).

On the other hand, there are authors who prove that unobserved heterogeneity is not only a threat to the validity of any structural model but also to measurement models (Sarstedt et al. 2020; Fordellone and Vichi 2020). For this reason, it is considered necessary to analyse the existence or not of unobserved heterogeneity in this model in order to clarify whether it is ultimately valid and also to be able to identify the population group with the highest predictive capacity. For this purpose, the prediction-oriented segmentation technique FIMIX-PLS and PLS-POS (Robina-Ramírez et al. 2019) will be used. FIMIX-PLS is considered the appropriate and commonly used approach to identify heterogeneity (Fordellone and Vichi 2020; Klesel et al., 2019; Marin-Garcia and Alfalla-Luque 2019). But later, Becker et al. (2013) presented a new method called PLS-POS to detect unobserved heterogeneity with higher fit, which reveals heterogeneity in both the structural model and the measurement model giving the possibility to discover very small segments of the population (Arenas-Gaitán et al. 2019) with disparate characteristics. This methodology has been applied in other current research such as Robina-Ramírez et al. (2019), Silva et al. (2019), Zhiqiang et al. (2020), Fordellone & Vichi (2020) among others.

Finally, a breakdown analysis of the variance explained is performed to identify the variable that contributes most to the explanation of the model developed, specifically to find out which construct is the one that contributes most to entrepreneurial performance.

3 Results

The results of the analysis of the data with the PLS-SEM technique allow us to validate the measurement instrument, the structural model, and the validity of the sample as a whole (Fig. 3). The different estimates of the model provide the distribution of the internal path coefficients (Nitzl et al. 2016; Hair et al. 2019), i.e., thanks to this technique, it will be possible to estimate the impact of business digitalisation on the satisfaction of two of its stakeholders: customers and employees. And consequently, to measure the impact of satisfaction on business results.

Fig. 3
figure 3

Initial result of the model of the impact of business digitalisation on customer and employee satisfaction

Full size image

3.1 Assessment of the global model

Following Hu & Bentler (1998) and later Hair et al. (2020), the values 0.01, 0.05, and 0.08 are used to indicate excellent, good, and mediocre accuracy, respectively, and in our case, the SRMR value for the saturated model is below the threshold of 0.05, which indicates good accuracy (Table 2). Likewise, the model estimate is below 0.08, which is very close to a good level of accuracy.

Table 2 Accuracy rates
Full size table

Henseler et al. (2016b), state that the SRMR values for a two-tailed Bootstrapping, must be less than the Bootstrap result at 97.5% in the relationships between variables and if the SRMR value exceeds this value, it is unlikely that the model is true. In the case of our model, we have used two-tailed bootstrapping in order to know the sign of the hypothesis and, as can be seen in Table 3, this requirement is met, and the model is considered to be true. Furthermore, if we look at the VIF values of the structural model, we can determine that they are all within the parameters established by the literature (see Table 4).

Table 3 Bootstrap accuracy test with sample size at 95% confidence
Full size table
Table 4 VIF estructural model
Full size table

3.2 Assessment of the measurement model

Internal consistency was assessed using Cronbach’s alpha and composite reliability. According to theory, authors such as Nunnally & Bernstein (1994) or Hair et al. (2020), among others, suggest 0.7 as an adequate level for ‘modest’ reliability in early stages of research, although the alpha value should range from 0 to 1. The closer the alpha value is to 1, the higher the internal consistency of the items analysed, i.e. it is assumed that the items are measuring the same dimension.

As the table above shows, our model not only exceeds the acceptance threshold for composite reliability, but also shows a perfect accuracy (Table 5).

Table 5 Reliability and construct validity (internal consistency)
Full size table

The same is true for the Dijkstra-Henseler’s (ρA) indicator (rho_A) all variables analysed exceed the threshold of 0.7 showing a perfect value (Hair et al. 2020).

The mean extracted variance was used to measure divergent validity and, on this occasion, the minimum value recommended by Hair et al. (2014) is 0.50. Interestingly, the analysis shows that these variables explain the entire model, since they all reach a value of 1.

In our model, Cronbach’s Alpha, rhoA, Composite Reliability and Average Extracted Variance are equal to 1, because in type A or reflective indicators, when measured by a single indicator, these values can only be equal to 1.

The square root of the average extracted variance of each latent variable should be greater than the correlations it has with the rest of the variables (Hair et al. 2017) as shown in Table 6, this criterion is met in our model. And once again, if we look again at the VIF values of the measurement model, the parameters established by the literature for all the indicators are met again (see Table 7).

Table 6 Discriminant validity. Fornell-Larcker criterion
Full size table
Table 7 VIF values measurement model
Full size table

On the other hand, following Chin (2010), the evaluation has been checked both at the construct level and at the indicator level. As for the evaluation at the indicator level, it is necessary to indicate that there is no multicollinearity between the indicators, and also according to the assessment of their magnitude of the weights and their significance, it has been decided not to eliminate any of them due to their contribution to the final explained variance and to the fact that these 12 indicators, according to the theory, are those that make up the Digital Business Ecosystem (DBE).

Having determined the validity of the model, as well as the good accuracy of the model, it is necessary to test whether this model would be valid and could be successfully applied to complex models (Hair et al. 2019). For this purpose, the bootstrapping technique was used, bounded by 5000 iterations.

Looking at Fig. 4, it can be seen that the use of Social Media mainly, followed by having one’s own website, are the conditions that have the most significant relationships with the digitalisation of the company and the digital business ecosystem.

Fig. 4
figure 4

Bootstrapping algorithm: model after bootstrap procedure (5000 iterations)

Full size image

In turn, the model shows that, thanks to business digitalisation, both customers and employees are satisfied with this digitalisation, although the satisfaction value of employees is slightly higher, probably because internal processes have improved for them, and they find it easier to carry out their work activities.

Furthermore, as the main contribution, the model shows how employee satisfaction has an impact on the company’s performance and stands out above customer satisfaction, as the value is more than twice as high as that of customer satisfaction.

Following the theory described (Hair et al., 2017), the closer the P-values are to zero, the more accepted they will be. As the Table 8 shows, all the relationships between variables have a zero value, but if we look at the T-statistics and as already anticipated in Fig. 4, the improvement in performance due to employee satisfaction stands out above the other relationships. Therefore, it can be concluded that digitalisation has a positive influence on employee satisfaction, which in turn influences an improvement in organisational performance.

Table 8 Path bootstrapping coefficients (5000 iterations)
Full size table

As for the variance explained (R2) the value required according to theory (Falk and Miller 1992) must be greater than 0.10 and a value of 1.0 indicates a perfect accuracy and, therefore, a very reliable model for future forecasts. The higher the value, the more predictive the model is. In our case, the value exceeds the value for the performance variable exceedsthis acceptance threshold.

Moreover, according to the theory expounded by Chin (1998) and later by Hair et al. (2019), the predictive relevance of the model will be low if Q2 > 0, will be medium if Q2 > 0.25 and will be high if Q2 > 0.50. In this case the predictive relevance is at a low level, for all variables (Table 9).

Table 9 Level of R2 and predictive relevance of the model
Full size table

Subsequently, we searched for unobserved heterogeneity in the model. This analysis of unobserved heterogeneity indicates that there are no significant differences within the sample, which allows us to ensure that we have a homogeneous sample in relation to the dependent variable, and with the structural model defined, therefore, we have a robust sample for the purpose of our study.

In order to deepen our knowledge of the dependent variable, we have Proceed to analyse the decomposition of the variance explained by this variable, based on the previous constructs, as we can see in Table 10.

Table 10 Analysis of the contribution to the explained variance
Full size table

When analysing the decomposition of the Explained Variance of Performance (14.7%), we see that substantial differences can be observed, as employee satisfaction has a contribution to the explained variance of almost 11.9%, much higher than customer satisfaction with only 2.8%.

4 Discussion

Digital tools bring several advantages to the business world, favouring the competitive positioning of organisations (Valdez-Juárez et al. 2020). In turn, these tools have developed a change in both external relationships with customers and internal relationships with employees (DiPietro et al. 2020; Sun & Zhang 2021) i.e., they contribute to greater stakeholder satisfaction and can be transformed into improved economic performance of the company (Parise et al. 2016; Castellacci et al. 2019).

Based on this, we established an objective in line with the beginning of the research, which consisted of analysing the satisfaction of customers and employees of a sample of Spanish SMEs, in terms of the digital ecosystem of their company and, once this satisfaction or dissatisfaction had been identified, identifying how it affects the performance of these organisations.

This research consisted of two steps: firstly, a confirmatory study based on hypotheses, specifically 4. Starting with the stakeholder satisfaction analysed, we found a positive relationship of customer satisfaction with the Digital Business Ecosystem (H1) of 0.136 and another positive but slightly stronger relationship of employee satisfaction with the Digital Business Ecosystem (H2) of 0.155. These results are in line with the literature as it has previously been shown that the application of digital tools as well as the development of a DBE not only improves the internal processes of organisations, but also enhances employee relations (Zhu & Smith 2019; Winkelhaus et al. 2022).

Furthermore, authors such as Hanelt et al. (2021) already identified that digital business models focused on customer satisfaction are gaining importance, as this, as we have confirmed with this analysis, results in their positive influence on business performance.

Following Williams et al. (2009) and Henseler et al. (2016a), the root mean square residual (SRMR) value shows a good model accuracy.

The Bootstrapping analyse shows that the most significant relationship is the one corresponding to the fourth hypothesis (H4) which relates employee satisfaction with an improvement in organisational performance. Again taking employee satisfaction as a reference, the literature shows that a satisfied or motivated employee is one of the most important sources of improvement in business performance (Yunis et al. 2018; Martínez et al. 2020; Grecco et al. 2021) in addition it has also been previously demonstrated that digitalisation favours this satisfaction (Castellacci & Viñas 2019; Torre et al. 2021; Winkelhaus et al. (2022) with this research further reinforcing these theories. In another vein, there is some research contrary to these findings, claiming that the use of these technologies exacerbates work-life conflicts and negatively impacts employee satisfaction, especially in the wake of COVID 19 and the forced digitalisation of organisations (Farivar & Richardson 2021; Nemteanu & Dabija 2021). However, this research provides empirical evidence of the influence of DBE and job satisfaction on the new normal, as the data are for the year 2022.

But what is really important from the analysis is that there is a strong relationship of the employee satisfaction variable with the corporate performance variable. This is again in line with the theory that satisfied employees are one of the causes of good company performance (Grecco et al. 2021). Finally, if we look at the contribution to the variance explained, we again find that the employee satisfaction variable is the one with the highest variance explained and therefore it is the one that has a contribution to business performance thanks to the technologies that are part of the digital business ecosystem.

5 Conclusions, limitations and future research line

Digital business ecosystems are an increasingly desirable context for companies seeking to embrace digital transformation (Wang 2020). The tools that are part of this ecosystem have developed a change in both internal and external relationships, contributing to greater stakeholder satisfaction (DiPietro et al. 2020; Sun and Zhang 2021). This, in turn, can lead to improved business performance (Parise et al. 2016; Castellacci et al., 2019).

The digital business ecosystem and its component technologies have been found to contribute to both customer satisfaction (H1) and employee satisfaction (H2). It was previously shown that the application of digital tools, as well as the development of a DBE, not only improve the internal processes of organizations, but also improve relationships with internal agents, such as employees (Zhu & Smith 2019; Winkelhaus et al. 2022). Furthermore, in the analysis of performance through stakeholder satisfaction, it was confirmed throughout the process that employee satisfaction contributes to a greater extent to improved business performance (H4). Again, taking as a reference the research by Zhu & Smith (2019) and Winkelhaus et al. (2022), where they related the improvement of employee satisfaction with the application of digitization to companies. Furthermore, the analysis of the contribution to the explained variance further reinforces these results.

Therefore, we can conclude that our results show that a good digital business ecosystem and the technological tools that compose it contribute to improved customer and employee satisfaction, and more satisfied employees lead to improved organisational performance. These results are in line with the theories proposed by authors such as Yunis et al. (2018) or more recently by Grecco et al. (2021) and therefore this research provides new evidence on this line, adding that all the hypotheses are contrasted and highlighting hypotheses H2 and H4, related to employee satisfaction.

Moreover, this work not only has a theoretical application but also a practical one. Therefore, we recommend that the business ecosystem of Spanish SMEs, or those with similar characteristics to Spanish SMEs, focus their investments on digital innovations in order to increase stakeholder satisfaction, as it has been shown that greater stakeholder satisfaction will lead to improved business results. The findings of this study have significant implications for organizational managers, as they illustrate how the relationships between the model variables exert different impacts. Consequently, managers must devise strategies to encourage innovation and investment in digitization to improve the satisfaction of their stakeholders both inside and outside organizations, specifically customers and employees. Promoting innovation and digitalization of processes is essential not only because it contributes to improving satisfaction, but also because it allows companies to remain viable in a globalized and fiercely competitive economy, since it contributes to improving their business results.

It should be pointed out that the main limitation of the research is that this study is based on a survey carried out in a specific year (2022) and although the results are very current, they could vary if the survey were carried out at a different point in time. Another limitation is that performance is measured on the basis of perceptions and not numerically, so that carrying out this study on the basis of real economic figures could lead to variations in the results. And from this limitation, a new future line of research is proposed. Finally, it is possible that we have not considered all of the publications related to the research topic. However, the collected studies clearly address the situation of digitalization, satisfaction, and performance.

Furthermore, it is proposed to analyse the relationship between employee satisfaction and customer satisfaction in order to find out how employee satisfaction affects employee satisfaction in the improvement of processes and products to improve customer satisfaction. And to analyse how employee satisfaction affects not only business performance, but also the efficiency and effectiveness of production processes.