Introduction

The challenge of adopting and implementing information and telecommunication technologies (ICT) will pose a critical issue for many companies in the coming decades. Digital technologies are impacting almost every economic sector in multiple ways, by creating new markets for new products, and transforming production and distribution processes. Digitalization has become an engine for entrepreneurship, productivity growth and, therefore, is crucial for business survival and growth (Gartner et al., 2022; OECD, 2020; Reuschke et al., 2021; Zahra et al., 2023; Zhiyong et al., 2023). Moreover, the COVID-19 pandemic has accelerated the digital transformation process of all types of companies, especially of small and medium-sized enterprises (SMEs) (Mandviwalla & Flanagan, 2021).

Following Vial (2019:118), digital transformation can be viewed as “a process where digital technologies create disruptions triggering strategic responses from organizations that seek to alter their value creation paths while managing the structural changes and organizational barriers that affect the positive and negative outcomes of this process”. Thus, for companies, digitalization involves a transformation process that can affect their products, their production processes, the marketing of their products, and their business organization. As a result, totally new or substantially modified products and processes can emerge (Berman, 2012; Kohli & Melville, 2019; Nambisan et al., 2017). Therefore, digital transformation conceptually implies an innovation process for a company.

Given the significance of this process for the future of many companies, it is necessary to better understand the factors that enable and hamper digital innovation. Although the previous literature has explored this issue (Hanelt et al., 2021; Rêgo et al., 2022; Vial, 2019), it has focused on specific conditioning factors or particular technologies, such as big data (Ciampi et al., 2021), artificial intelligence (Anica-Popa et al., 2021), blockchain (Morkunas et al., 2019), and the Internet of Things (IoT) (Ben-Daya et al., 2017) and fails to provide a holistic vision of the nature and determinants of the digital transformation of companies. Furthermore, empirical evidence remains limited, is mainly qualitative and concentrates on large companies (Ghosh et al., 2022; Rêgo et al., 2022; Reuschke et al., 2021). However, in general terms, SMEs experience a certain delay in their advances toward digitalization (OECD, 2020; Reuschke et al., 2021) due to the obstacles they face to digitally transform their organizations. Owing to their specific characteristics, such as their limited size and the low level of professionalization of their management, their access to financial, technological, and human resources can be constrained (Li et al., 2018; Meier et al., 2022). Therefore, it is convenient to dive into the specific determinants of digitalization in this type of companies.

In this respect, the current article seeks to fill this gap in the literature by investigating the drivers and obstacles to digital transformation in SMEs. To this end, this paper studies digital transformation as an innovation process in the company, which leads to the implementation of new or significantly modified products and processes through the application of digital technologies. Therefore, the article approaches the phenomenon of digital transformation through the concept of “new to firm” innovation (OECD/Eurostat, 2018).

From a theoretical perspective, this paper contributes to this research area by proposing a holistic approach to investigating the determinants of digital transformation in SMEs that integrates factors explored in the previous literature and adds new factors to the debate. The analytical framework presented differentiates three dimensions (technological, human capital, and organizational/relational) and three levels of analysis (the individual entrepreneur/manager, the firm, and the business environment) to investigate the factors affecting the digital transformation process of SMEs. The literature shows a lack of integration of these dimensions/levels of analysis, which can affect the robustness of the results obtained in studies with a specific focus. In this respect, the approach adopted in this research enables us to address the issues associated with omitted variables that arise in other studies focusing on specific determinants of digitalization. In addition, the analysis presented in this paper contributes to the literature by showing that the digitalization of products and processes is driven by different factors to some extent, and that the determinants of digital transformation partially vary between microenterprises and other SMEs.

The empirical study included in this article is based on a dataset of 802 SMEs obtained from a representative survey of the SME sector in Spain. A stratified sampling approach was employed, distinguishing between two subgroups: microenterprises (fewer than 10 employees) and other SMEs (10 employees or more). This rich database allows us to consistently evaluate the hypotheses proposed in this study. Regarding the econometric methodology, the paper employs logistic regression modeling.

The results obtained suggest that human capital and organizational/relational factors are major barriers for digital transformation in SMEs beyond the mere technological aspects. Thus, both internal ICT skills and external knowledge through ICT consultants, suppliers, universities, and technological centers are key resources for digital transformation in this type of company. Therefore, the results of the current research emphasize the significance of interaction and partnership leading to the formation of digital ecosystems. The analysis also reveals the importance of an explicit digital transformation strategy in the company and the distribution of responsibilities regarding digitalization beyond the leadership of the entrepreneur/manager. The profile of entrepreneurs/managers as Internet users and their ambition for growth are also observed to influence the digital product and process innovations, respectively. In addition, the paper highlights differences in the importance of the previously mentioned determinants of digital innovation between microenterprises and other SMEs.

Theoretical Foundations for the Analysis of the Digital Transformation of SMEs

In this section, the previous literature on the digital transformation of SMEs is reviewed and the analytical framework for this research is presented. Following that, the hypotheses to be tested in the empirical part of the paper are introduced.

Literature Review and Analytical Framework

Although the literature on the topic of digital transformation is growing (Hanelt et al., 2021; Rêgo et al., 2022; Vial, 2019), it remains scarce and has yet to clarify the most relevant mechanisms that carry out this process (Kohli & Melville, 2019; Ritala et al., 2021). Moreover, the existing literature mainly focuses on large companies (Ghosh et al., 2022; Giustiziero et al., 2021) and on particular technologies (Morkunas et al., 2019; Ben-Daya et al., 2017; Anica-Popa et al., 2021; Ciampi et al., 2021) and, therefore, fails to provide a holistic framework that supports the analysis of the circumstances that SMEs face to transform themselves digitally.

The changes in companies associated with the adoption and assimilation of digital technologies involve an innovation process. The previous literature has explored the innovation involved in the digital transformation process using various concepts (Bogers et al., 2022; Kohli & Melville, 2019). Thus, the terms “information technology innovation” (Jeyaraj et al., 2006) and “information system innovation” (Swanson, 1994) have been employed regarding the introduction of new processes, products, and services associated with the adoption of an already existing ICT that is new to an organization. Furthermore, the term “digital innovation” (Lee & Berente, 2012) has been utilized to adopt a more “product-centric perspective” to refer to new combinations of physical and digital products to form new products (Kohli & Melville, 2019).

Overall, the adoption, application, and assimilation of ICT in companies can lead to innovations in products, production processes, marketing practices, and organizational systems and routines. In this paper, the determinants of ICT-driven product and process innovations will be considered.

The digital transformation of SMEs is a complex and multifaceted phenomenon. In order to approach it holistically, a theoretical framework that differentiates between a set of functional dimensions and levels of analysis serves as a useful tool. This paper identifies three dimensions in the digital transformation process: the technological, the human capital, and the organizational/relational dimensions. Each of these dimensions has manifestations at three different levels of analysis: the individual entrepreneurs/managers, the firm, and its external environment (see Table 1).

Table 1 Analytical framework. A typology of determinants of digital transformation in SMEs
Full size table

Regarding the levels of analysis, the role of entrepreneurs/managers is particularly relevant in the creation, organization, and performance of SMEs compared to their influence on larger organizations (Guzmán & Santos, 2001; Hadjielias et al., 2022; Romero & Martínez-Román, 2012). Therefore, at the individual level (i), the personal characteristics and mindset of the entrepreneur/manager can condition the digital transformation process in SMEs (Ghosh et al., 2022; Li et al., 2018). Furthermore, the digitalization process of SMEs is naturally shaped by the characteristics of the firm, such as its resources, organization, and management. These aspects configure the firm level of analysis (ii) in the study of digital transformation of SMEs. Finally, companies operate in a particular context and interact with other agents, such as their clients, suppliers, competitors, public administration, universities and research centers, and business associations (Kohli & Melville, 2019). Therefore, a third level of analysis of the digital transformation process gathers the characteristics of this external environment (iii).

Furthermore, the three different functional dimensions of the business digital transformation are presented in greater detail below.

  1. A.

    Technological dimension. This digital transformation process is associated to advances in ICT that enable information that is translated into a binary code to be generated, collected, stored, processed, analyzed, and shared. This includes electronic devices, such as computers, mobile phones, video cameras, personal digital agendas, robots, and drones, and applications, such as blockchain, artificial intelligence, cloud computing, the Internet of Things (IoT), voice interfaces or chatbots, and video streaming (Ciampi et al., 2021; Ben-Daya et al., 2017; Morkunas et al., 2019; Anica-Popa et al., 2021). These digital applications are possible through the use of the Internet, which represents a key enabler for the development of the digital economy. From this perspective, a primary technological conditioning factor for business digital transformation involves the degree of connectivity and use of the Internet in the company.

  2. B.

    Human capital dimension. Firms are knowledge-generating actors that undertake learning processes from inside and outside their own organization and can apply the new knowledge to innovate (Alavi & Leidner, 2001). Knowledge and skills, together with the learning process that leads to their acquisition, therefore constitute a second key ingredient in the digital transformation process in companies. Thus, as Kohli and Melville (2019: 205) state, “knowledge and learning are intimately tied to the notion of digital innovation”. In this respect, from an internal perspective, the educational level and previous training of entrepreneurs/managers and the rest of the staff could provide the learning capabilities that facilitate access to the knowledge and skills necessary to tackle the digital transformation of the company. However, the knowledge and skills that are critical for the digital transformation process are logically those related to digital technologies and their applications, that is, digital skills and capabilities. Furthermore, from an external perspective, companies can benefit from human capital and expertise of other companies and institutions by contracting consultant services, participating in training activities, and through other forms of collaboration.

  3. C.

    Organizational/relational dimension. Finally, the digital transformation process in SMEs is conditioned by the manner in which the company integrates the digital technology, the digital capabilities, and other resources to innovate as a result of the management of the internal organizational processes within the company and the relationships with external agents. In this regard, Ghosh et al., (2022: 2) have proposed the concept of digital transformation capability defined as “the ability of a firm to systematically identify and develop core capabilities for digital transformation”. Likewise, Li et al. (2018) defend that inadequate managerial capabilities are a key barrier for digital transformation in SMEs. In this respect, Ghosh et al. (2022) point out that the organizational structure of the company can represent either an enabler or an obstacle to digital transformation, and hence companies need to configure internal operational structures with a suitable assignment of roles and responsibilities for the successful development of the process. Likewise, the innovations associated with the introduction of ICT in the firm pose changes that require the support of the entrepreneur, the managerial team, and the rest of the staff, and therefore in order to succeed, the digital transformation initiatives need to adjust to the prevalent organizational culture (Ghosh et al., 2022; Jeyaraj et al., 2006). Consequently, the success in the implementation of the process of digital transformation in SMEs can be conditioned by the way in which the process is promoted, perceived, and managed within the organization. Likewise, from an external perspective, relationships with other agents reveal the potential to shape the digital transformation process of SMEs. Knowledge sharing and collaboration within communities can foster access to the intangible and tangible resources necessary to drive digital transformation (Kohli & Melville, 2019; Wang & Ramiller, 2009). In this regard, Ghosh et al. (2022) highlight the importance of strategic partnerships as a channel towards the acquisition of digital transformation capabilities, but they also recognize that these partnerships are difficult to manage, since the partners may have contradictory interests. This relational view of how digital transformation can be conditioned by collaboration with a firm’s stakeholders (clients, suppliers, other companies, public administrations, universities and research centers, business associations, etc.) have led to the notion of “digital ecosystems” (Dyba et al., 2022; Li et al., 2012; Senyo et al., 2019). Digital ecosystems drive value co-creation processes associated with digital transformation and the development of the digital economy. Furthermore, Sussan and Acs (2017) have proposed the concept of a “digital entrepreneurial ecosystem” that highlights the role of Schumpeterian entrepreneurs creating digital companies and innovative products and services. Digital ecosystems often emerge in the context of the digital platform economy, generating interactions between different stakeholders, such as governments, platform-dependent firms, and end users (Aminullah et al., 2022; Lafuente et al., 2022).

Research Hypotheses

In this section, the research hypotheses to be tested in the empirical part of this article are presented, and are classified in accordance with the analytical framework proposed in Table 1.

  1. A.

    Technological dimension

    Although there are numerous digital technologies with multiple applications, in this paper the access and usage of the Internet is considered a proxy factor that captures the technological dimension of the digital transformation process. This decision is justified by its fundamental role as an enabler of other digital technologies and digital business applications.

    At the individual level, the entrepreneur’s/manager’s awareness constitutes a key factor in the digital transformation of SMEs (Hanelt et al., 2021; Li et al., 2018). In this respect, the use of the Internet by entrepreneurs and managers allows them to identify the possibilities associated with the digital transformation of their companies (Orrensalo et al., 2022). Hence, it determines the role that entrepreneurs/managers can assume in leading and pushing the innovation efforts oriented towards the digital transformation process of their companies (Martín-Martín et al., 2022).

    At the firm level, the possibilities to innovate to digitally transform the firm’s products and processes are naturally conditioned by the ICT infrastructure of the company and the extent to which the company’s staff have access to electronic devices connected to the Internet (Armstrong & Sambamurthy, 1999). This factor represents a key enabler for any innovation process based on digital technologies within the firm.

    Finally, with respect to the characteristics of the business environment, several investigations have highlighted the advantages of good Internet connectivity (Belloc et al., 2012; Falch & Henten, 2018). The use of broadband can increase the productivity of SME and favor its digitalization (Haller & Lyons, 2015). Therefore, the problems of accessibility to the Internet have been considered a critical factor that can lead, for instance, to a spatial digital divide between cities and rural areas (Labrianidis & Kalogeressis, 2006; Prieger, 2013).

    In this respect, the following hypotheses are postulated regarding the technological dimension of the digital transformation process:

    H1: Entrepreneurs/managers who make a more intense use of the Internet are more likely to introduce innovations driving the digital transformation of their SMEs.

    H2: SMEs with a higher proportion of staff using electronic equipment (such as computers, tablets, and mobile phones) with Internet connectivity are more likely to introduce digital innovations to drive their digital transformation.

    H3: SMEs that suffer difficulties in attaining a proper Internet connection (stable, with sufficient capacity and reasonable cost) are less likely to introduce innovations to drive their digital transformation.

  2. B.

    Human capital dimension

    The educational level of the entrepreneur/manager has been observed to be an important determinant of innovation in SMEs (Romero & Martínez-Román, 2012, 2015) and top management support has been identified as a primary factor favoring digital transformation (Ghosh et al., 2022; Gono et al., 2016; Jeyaraj et al., 2006). In this respect, the education and skills of the entrepreneur/manager can present relevant factors in the successful adoption of ICT (Ramayah et al., 2016).

    Furthermore, at the firm level, research has linked an enterprise’s creative mindset to the quality and training of the staff (Cooper & Zmud, 1990). This represents a major source of competitive advantage, and hence enterprises with skilled staff can differentiate themselves from competitors, thanks to employee-driven innovation. The management of these knowledge assets is thus critical for the innovation capabilities of the companies (Kesting & Ulhøi, 2010) and the ICT skills gain particular importance in this respect (Ghobakhloo & Ching, 2019). Motivation of staff to increase their own abilities and innovation by using technical knowledge may lead to innovation, while strengthening digital competencies in companies can promote ICT adoption and assimilation.

    Nevertheless, companies can drive their digital transformation process by either intra-organizational (internal capabilities and in-house R&D), inter-organizational (open innovation) channels, or a combination of the two (Scuotto et al., 2017). They can therefore benefit from the external human capital available in their external environment. One straightforward way to acquire external knowledge and skills is to contract ICT consulting services. This could favor the ability of SMEs to boast their digital transformation. Similarly, though public support for SMEs varies across countries (Meramveliotakis & Manioudis, 2021), SMEs can benefit from it in different areas (Chan et al., 2022; Wang et al., 2023). Thus, public institutions can increase awareness of the relevance of digitalization and the opportunities and risks it poses to SMEs. Additionally, public programs can also support the acquisition of the digital capabilities that SMEs often lack.

    In this respect, the following hypotheses are considered in this paper:

    H4: Entrepreneurs/managers with a university degree or higher professional training are more likely to introduce innovations to drive their digital transformation.

    H5: SMEs that employ staff with ICT expertise or digital skills are more likely to introduce innovations to drive their digital transformation.

    H6: SMEs that contract external ICT consulting services are more likely to introduce innovations to drive their digital transformation.

    H7: SMEs participating in public activities to raise digital awareness and training are more likely to introduce innovations to drive their digital transformation.

  3. C.

    Organizational/relational dimension

    At the individual level, the motivation and entrepreneurial orientation of managers/entrepreneurs have been identified as significant components in the development of SMEs (Forcadell & Úbeda, 2022; Hermans et al., 2015). In particular, ambitious entrepreneurs/managers enjoy a greater probability of establishing successful businesses (Guzmán & Santos, 2001; Shane, 2009) and this growth ambition can lead them to promote the digital transformation of their companies. At the firm level, digitalization transforms the strategic management and internal organization of companies in different ways (Berman, 2012; Rêgo et al., 2022) and a specific vision, serving as a guide for business digitalization, can be essential for the successful development of the process (Matt et al., 2015; Matt et al., 2016). Some companies count on a digital strategy, understood as an “organizational strategy formulated and executed by leveraging digital resources to create differential value” (Bharadwaj et al., 2013:472). However, the digitalization process in SMEs is not always the result of a deliberate and explicit plan, but it is sometimes based on “ad hoc approaches” and individual initiatives of the entrepreneur, members of the management team, and of specific employees (Kohli & Melville, 2019). Furthermore, the responsibilities and tasks derived from the digital transformation in companies can be distributed in different ways within the organization: they can be led directly by the own entrepreneur/manager, they can be in charge of a specific unit headed by a Chief Digital Officer (CDO), or they can be distributed among different units/departments/individuals. In other cases, these tasks/responsibilities are not formally assigned, and the process is driven based on disperse initiatives from various individuals in a spontaneous manner. The literature does not offer a response to the question of which organizational design is more efficient (Matt et al., 2015; Ghosh et al., 2022). Nevertheless, the fact that the entrepreneur/manager themselves is in charge of the digitalization tasks and responsibilities may result in a constraint. The entrepreneur/manager in SMEs may have limited digital knowledge and skills and a narrow view of the possibilities of digitalization and may lack the time to reflect on the digitalization process and to implement it efficiently. In any case, digital transformation is a process that potentially changes the whole organization, and it is important to be able to count on the acceptance of the employees who must necessarily participate in the process. When staff view the digitalization process as a threat, by thinking that it may put their jobs at risk or cause some disturbance and drawbacks, then resistance to change may arise and consequently hamper the digital transformation process. Furthermore, SMEs can boost their digital transformation by collaborating with other agents within their business environment (Sussan & Acs, 2017). In this respect, suppliers can provide a channel for innovation (Fernández-Serrano et al., 2019) and encourage the digitalization of SMEs as a result of their needs, requests, and knowledge. In particular, supply chain management is influenced by digital technologies such as blockchain and IoT (Ko et al., 2005; Ben-Daya et al., 2017). Similarly, the expanding pace of technological progress has created a situation in which coordinated R&D activities are required (Chesbrough, 2003). In this respect, SMEs can collaborate with research centers and academic institutions to cover R&D expenses and access other resources to accelerate their digitalization process (Chan et al., 2022).Footnote 1

Based on the aforementioned arguments, the following hypotheses are postulated:

  • H8: Entrepreneurs/managers with growth ambition are more likely to introduce innovations to drive the digital transformation of SMEs.

  • H9: SMEs with an explicit and ongoing digital strategy are more likely to introduce innovations to drive their digital transformation.

  • H10: SMEs in which the entrepreneur/manager directly assumes the responsibilities regarding digitalization are less likely to introduce innovations to drive their digital transformation.

  • H11: SMEs whose employees have a positive attitude toward digitalization are more likely to introduce innovations to drive their digital transformation.

  • H12: SMEs that cooperate with other actors in their ecosystems are more likely to introduce innovations to drive their digital transformation.

Data and Methodology

In this section, the data used in this research and the methodology adopted are presented.

Data and Variables

This research is based on a survey conducted in the second quarter of 2022 on Spanish SMEs with at least one employee and up to 200 employees. A stratified sampling method was used, differentiating two subgroups: microenterprises (with fewer than 10 employees) and other SMEs (with 10 employees or more). Small and medium-sized enterprises participating in the survey were randomly selected from the System of Iberian Balance Analyses (SABI) database. This stratified sample is representative of the SME population in Spain with a margin of error of ± 5.0% at a 95% confidence level. This was achieved by applying simple random sampling for both subgroups and assuming a binomial population distribution (p = q = 0.5, the most unfavorable situation). The survey technique used was that of computer-assisted telephone interviewing (CATI). In the fieldwork, a response rate of 21.5% was obtained. No bias was detected between respondents and non-respondents. The final dataset for this analysis consists of 802 valid observations, comprising 379 microenterprises and 423 SMEs with 10 employees or more.

  1. A.

    Dependent variables

    The dependent variables in this study capture two types of innovation associated with the digital transformation of SMEs:

    - Product innovation (prod_inn). This binary variable is based on the answers to the following question: “During the 2019–2021 period, did your company introduce innovations in new or significantly improved products/services that involve a digitalization process?”.

    - Process innovation (proc_inn). This binary variable comes from the following question: “During the 2019–2021 period, did your company introduce innovations in its production, logistics, or distribution processes, either for new processes or significant improvements in existing ones, that imply a digitalization process?”.

    Both variables take a value of 1 if the firm has implemented any innovation (product or process) with a digitalization dimension, or 0 in the negative case.

  2. B.

    Explanatory variables

    The explanatory variables considered in this study are the following:

    i) Technological dimension

    - Intensity in the use of the Internet by the entrepreneur/manager (Entrep_Internet): The entrepreneurs/managers interviewed were asked “How often and for what purpose do you use the Internet?” The answers were coded as an ordinal variable which takes values from 1 to 5, whereby 1 implicates a less frequent and more simple use of the Internet, and 5 a more frequent and complex use.

    - Percentage of the staff that use computers, tablets, and/or mobile phones with Internet connection for work tasks (Emp_Internet).

    - Broadband quality (Internet_connec): Managers interviewed were asked whether they “experience difficulty in attaining a proper Internet connection”. The answers were coded as an ordinal variable which takes values from 0 to 3, whereby 3 means high and 0 low.

    ii) Human capital dimension

    - Entrepreneurs/managers with university degrees or higher professional studies (Entrep_edu): This binary variable takes the value of 1 for those SMEs with entrepreneurs/managers who had a university degree or higher professional training, and 0 for the rest.

    - Percentage of employees with higher education (Emplo_edu).

    - Medium-level staff with ICT expertise (Techn_TIC): This binary variable takes the value 1 for those companies that have ICT experts in medium-level positions (0 otherwise).

    - ICT training initiatives in the company (ICT_train): This binary variable takes the value of 1 if the company developed training activities to improve the staff’s ICT skills in the three years previous to the survey (0 otherwise).

    - External ICT consultancy (ICT_consu): This binary variable takes the value of 1 if the company contracted functions related to ICT with external suppliers and consultants in the period 2019–2021 (0 otherwise).

    - Public training activities (Publ_train): This binary variable takes value 1 if, during the 2019–2021 period, the company participated in any activity carried out by any public administration (local, regional, national, EU) to raise awareness, for dissemination, or for training activities related to business digital transformation (0 otherwise).

    iii) Organizational/relational dimension

    - Ambition for firm growth (Growth_amb): this variable captures the relevance of the entrepreneur’s/manager’s attitude towards the growth of the company in the future. The answers were coded as an ordinal variable that takes values from 1 to 5, whereby 5 means that they want the company becomes “as big as possible” and 1 just “manageable by myself or with few employees”.

    - Digital strategies (Dig_strategy): This binary variable takes the value of 1 if the company was implementing an explicitly digital transformation strategy (0 otherwise).

    - Responsibilities of digital transformation assumed by the entrepreneur/manager (Entr_dig_head): This binary variable takes the value of 1 if the decisions regarding the digital transformation of the company were made directly by the entrepreneur/manager.

    - Workers’ attitudes towards digitalization (Staff_dig_mot): this variable is based on the following question: “Are your employees generally motivated about the company’s digital transformation?”. The answers were coded as an ordinal variable using a Likert scale that ranges from 1 to 4, whereby 4 means highly motivated and 1 complete resistance.

    - Cooperation with suppliers (Coop_suppl). This variable captures the intensity of cooperation with suppliers located in the region, in other regions of Spain, and abroad, in aspects related to digitalization. The responses to each of the three spatial areas were coded as ordinal variables that take values from 0 to 3, whereby 3 means high and 0 low. The final variable is constructed as a sum of these ordinal variables.

    - Cooperation with Universities and R&D centers (Coop_univ). This variable captures the intensity of cooperation with universities and research centers located in their region, in other regions of Spain, and abroad in aspects related to digitalization. The variable was constructed in an analogous way to that in the case of cooperation with suppliers.

  3. C.

    Control variables

    A set of control variables is also included in the estimated models in order to isolate the effect of the main explanatory variables.

    - Firm size (Employees). The size of the company is included in the analysis measured in terms of the number of employees. Jeyaraj et al. (2006) observed that firm size is one of the three most predictive factors for ICT innovation adoption based on a meta-analysis of the previous literature.

    - Sectoral dummies: the SMEs in the sample were classified into four sectors: industry, construction, trade, and services. Dummy variables were included in the estimated models for the first three (the service sector was taken as the base category).

Table 2 illustrates the main descriptive statistics of the dependent, explanatory, and control variables. The average company in the sample was a small business with 19 employees operating in the service sector. Most of the companies did not introduce any innovation associated with digitalization in the period 2019–2021, and hence it could be considered that the intensity of the digital transformation process of the average SME surveyed was low-medium.

Table 2 Descriptive statistics
Full size table

From a technological perspective, the average company in the dataset did not report significant problems in accessing a proper Internet connection, approximately 3 out of 4 of their employees worked with electronic equipment connected to the Internet, and the entrepreneur/manager had a frequent and varied use of the Internet.

From a human capital perspective, the average entrepreneur/manager had a university degree or higher professional training. Less than half of the employees in the average company had a university degree or higher professional training and the company had no ICT expert as part of their medium-level staff.

From an organizational perspective, the average company declared itself to be implementing a digital transformation strategy led by the entrepreneur/manager. The average entrepreneur/manager reported that they had a medium level of ambition for growth and observed a reasonable level of motivation in their employees regarding the digital transformation of the company.

Table 2 also presents the descriptive indicators separately for microenterprises and other SMEs. As observed, microenterprises have a higher proportion of their staff using computers, tablets, and/or mobile phones with Internet connections. However, they exhibit lower availability of ICT experts within the company, are less active in developing ICT training initiatives and using external ICT consultancy, have lower growth ambition, less frequently implement a formal and ongoing digital strategy, and more often concentrate digitalization responsibilities on the entrepreneur/manager. Furthermore, microenterprises are less active in cooperating with suppliers, universities, and research centers on digitalization issues and less frequently introduce digital innovations compared to other SMEs.

Econometric Methodology

The logistic regression method is employed in this research to evaluate the influence of independent variables on dichotomous innovation variables. This econometric model can be presented as follows:

$$\text{ln}\left(\frac{p}{1-p}\right)=z={\beta }_{0}+{\beta }_{1}{x}_{1}+{\beta }_{2}{x}_{2}+\dots +{\beta }_{k}{x}_{k}$$
(1)

In Eq. (1), p stands for the probability that \(y=1\), where y alternatively represents the innovation dummies (prod_inn and proc_inn), \({x}_{j}\) are the independent variables (explanatory and control variables), and \({\beta }_{j}\) denote the regression coefficients \(\left(j=1\dots k\right)\).

The probability that a company innovates in its products and processes as a result of the application of digital technologies, for a given value of \({x}_{j}\), is given by the following expression:

$$p=\frac{exp\left({\beta }_{0}+\sum_{j}{\beta }_{j}{x}_{j}\right)}{1+exp\left({\beta }_{0}+\sum_{j}{\beta }_{j}{x}_{j}\right)}$$
(2)

This logistic regression model is estimated using the maximum likelihood method.

Results

Models for product and process innovation were estimated separately. Collinearity was not observed to be a problem in these models.

The results for digital product innovations are presented in Table 3. Regarding control variables, the number of employees shows a negative and marginally significant impact on innovation. This could indicate that smaller companies present greater flexibility in introducing innovations in digital products.

Table 3 Logistic regression. Digital product innovation
Full size table

Regarding the technological dimension, a complex Internet use by the entrepreneur/manager has a strong positive effect on the introduction of product innovations. In this respect, it seems that the entrepreneur’s/manager’s awareness and familiarity with digital technology substantially conditions the effective digital transformation of SMEs, as Li et al. (2018) and Hanelt et al. (2021) have pointed out. In contrast, no significant effect of the use of electronic equipment by employees or of access to a proper Internet connection in the company is observed.

With respect to the human capital dimension, the presence of technical staff with ICT expertise in the company, the development of training activities oriented towards digitalization in the company, and the use of ICT external consulting services, are observed to have a strong positive effect on digital product innovation. This indicates that both internal and external digital skills can accelerate the digital transformation process of SMEs However, the formal educational background of the entrepreneur/manager and the staff does not exert a significant effect on digital product innovation.

Finally, regarding the organizational/relational dimension, the existence of an explicit digitalization strategy is observed to have a strong positive effect on digital product innovation. This is in line with Matt et al. (2015) and Matt et al. (2016) among others. The motivation of staff towards digital transformation also seems to have a significant positive effect on digital innovation in SMEs. Likewise, the cooperation activities with suppliers and (more importantly) universities and technological centers have significant positive impacts on product digital innovation. This reveals that, in accordance with Sussan and Acs (2017), the interaction of SMEs with other agents within entrepreneurial digital ecosystems can accelerate their digital transformation.

The results regarding process innovation are presented in Table 4. In this case, there are no variables within the technological dimension that have a statistically significant effect on digital process innovation.

Table 4 Logistic regression. Digital process innovation
Full size table

Regarding the human capital dimension, only the presence of technical staff with ICT expertise and the use of external ICT consultancy show positive statistically significant effects on process innovation.

However, organizational/relational factors appear to play a major role in digital process innovation. In this case, the growth ambition of the company is observed to have a significant positive effect (that was not present in the results for digital product innovation). Furthermore, the existence of a digital strategy exerts a strong positive effect on digital process innovations, as was also observed in the case of digital product innovations. In this respect, it is worth mentioning that those SMEs in which the digital transformation is a responsibility assumed primarily by the entrepreneur/manager show a lower probability to introduce digital process innovations. This could be explained by the need, in the case of process innovations, of a more systematic and planned implementation that would be associated with an approach towards digitalization that is more organic and less dependent on the personal impulse of the entrepreneur/manager. Finally, cooperation with suppliers has a highly significant and positive effect on process innovation, as occurs in the case of product innovation, while in this case, cooperation with universities and R&D centers yields no significant effect.

The joint consideration of the results for digital product and process innovations consistently supports hypotheses H5, H6, H9, and H12. The results support hypotheses H1 and H11 only regarding digital product innovations, and hypotheses H8 and H10 only regarding process digital innovation. In contrast, hypotheses H2, H3, H4, and H7 can be rejected based on the results presented.

Overall, the analysis indicates that the technological dimension does not constitute the main barrier to attempts at digital transformation of SMEs in Spain. This is in line with Tabrizi et al. (2019) that “digital transformation is not about technology”. The majority of SMEs have access to basic digital technology, but, in contrast, they could be affected by human capital and organizational/relational limitations in order to successfully integrate digital technologies into their businesses. Regarding human capital, the results show that the formal educational background of the entrepreneur/manager and the staff is not as relevant as their digital competences. External sources of knowledge from suppliers, universities, and research centers are also observed to be critical factors since they can compensate, complement, and multiply the internal digital skills. Finally, SMEs with a more formalized approach towards digital transformation, based on an explicit digital strategy and distribution of responsibilities on digitalization, are observed to advance more firmly towards digital transformation. According to this finding, individual and spontaneous initiatives towards digital transformation represent a limited and inefficient approach leading to unsuccessful results in many SMEs. It is also worth noting that public activities to raise digital awareness and support digital training appear to exert no effect on digital innovation in SMEs. This may indicate, in line with Meier et al. (2022), that those initiatives must be tailored to the company’s particular circumstances in order to be effective.

In addition, it is interesting to study the determinants of digital innovation separately for microenterprises and other SMEs with 10 employees or more. Table 5 allows for a comparison of the results for digital product innovation between these two size groups. As can be observed, training activities regarding digitalization organized within the companies and the motivation of the staff regarding digitalization have no significant effects on digital product innovation for microenterprises, in contrast to the statistically significant impact seen in other SMEs. The availability of ICT experts and access to ICT consultancy present a lower impact on digital product innovation in microenterprises compared to the rest of SMEs. Furthermore, in SMEs with at least ten employees, growth ambition shows a significant effect on digital product innovation, whereas the impact of digital strategy vanishes. Moreover, cooperation with universities and research centers exerts a higher impact on product digital innovation in microenterprises compared to SMEs.

Table 5 Logistic regression. Digital product innovation by size groups within SMEs
Full size table

Table 6 shows the results for process digital innovation separately for microenterprises and other SMEs. The personal profile of the entrepreneurs/managers as Internet users and their educational background appear as particularly significant aspects for digital process innovation in microenterprises, in contrast with other SMEs. Microenterprises whose entrepreneurs or managers make more frequent and complex use of the Internet and do not have higher education degrees introduced more digital process innovations. This could be explained by the fact that entrepreneurs without formal education but with practical experience prioritize solutions that focus on immediate business needs and practical outcomes rather than theoretical or long-term considerations. In this context, digital tools can provide rapid solutions to pressing business challenges. Furthermore, the increasing accessibility and user-friendliness of digital tools mean that formal education is not a prerequisite for their adoption. Entrepreneurs without formal education can engage in alternative forms of learning, such as online courses, hands-on training, and user communities. These forms of education often emphasize practical digital skills and innovations directly applicable to their businesses.

Table 6 Logistic regression. Digital process innovation by size groups within SMEs
Full size table

Furthermore, microenterprises with a higher percentage of staff holding higher education levels and participation in public administration activities to support digitalization introduce more digital process innovations. Cooperation with suppliers has a stronger impact on process digital innovation in microenterprises than in larger SMEs. However, the availability of ICT experts and access to ICT consultancy are not significant determinants of process digital innovation in microenterprises, unlike in larger SMEs.

Additionally, in SMEs with ten or more employees, a negative significant effect of the leadership of the entrepreneur/manager on digital process innovation can be observed. This indicates that, in these companies, the organizational complexity associated with digitalization necessitates assigning responsibilities in this area based on a more organic and less personal approach. Furthermore, in SMEs with at least ten employees, growth ambition and a digital strategy show significant effects on digital process innovation. In these companies, staff motivation towards digitalization has a marginally significant and negative effect on process digital innovation, which is not observed in microenterprises. This suggests that digital process innovations in larger SMEs often cause significant resistance or discontent among staff, which should be managed to ensure the success of digitalization projects.

Considering the results for both product and process digital innovation, it is suggested that microenterprises adopt simplistic approaches to digital innovation that are less demanding in terms of specific knowledge resources. When they do require these resources, the significant cost of acquiring such knowledge and skills makes collaboration with external partners an efficient channel for access. Thus, collaboration with suppliers, universities, and research centers plays a critical role in leveraging external knowledge to drive digital innovation in microenterprises.

In contrast, larger SMEs have more complex approaches to digital innovation that require more specific knowledge inputs. These companies can obtain these inputs through hiring ICT experts, contracting external consultants, and implementing training activities within the company. Additionally, digital innovation projects in SMEs with 10 or more employees require higher staff involvement and ambitious, professional management of the digitalization process.

Discussion and Conclusion

This paper investigates digital transformation in SMEs and approaches it as an innovation process. The study aims to address a research gap by analyzing the drivers and barriers to digitalization in SMEs from a holistic perspective, and by integrating factors that have received attention in the literature with others that have been overlooked. In order to do this, a theoretical framework is proposed that differentiates between three dimensions of enablers — technological, human capital, and organizational/relational — and three levels of analysis — the individual entrepreneur/manager, the company, and the business environment. Moreover, this theoretical framework is particularly oriented to the characteristics of SMEs and the specific obstacles and conditionings that affect them in their processes of digital transformation. Therefore, the analysis carried out allows delving in the nature and determinants of digitalization in SMEs complementing previous research on this topic that is conceived predominantly from the perspective of large companies. Additionally, this research contributes to the literature by investigating digital innovations in products and processes separately, and showing that their determinants differ to some extent for microenterprises and the rest of SMEs.

The analysis provides quantitative empirical evidence on this insufficiently investigated topic using a representative sample of the SME sector in Spain. The holistic approached proposed differs from other studies that focus on specific determinants of digitalization, which can lead to misleading results due to the bias associated with omitted variables. The results indicate that, regarding the technological dimension, the complexity of the use of the Internet by entrepreneurs/managers exerts a significant positive influence on the introduction of product digital innovations. Regarding the human capital dimension, on the one hand, the existence of technical personnel with ICT experience, the technical assessment by external ICT consultants, and the development of digitalization-oriented training programs by the company are observed to have a substantial impact on digital innovation. Therefore, both internal and external skills seem to be critical for the digital transformation process in SMEs. With respect to the organizational/relational dimension, an explicit digitalization strategy represents a key factor that favors success in digital innovation. Moreover, it is also important that the implementation of this strategy is not the direct responsibility of the entrepreneur/manager but is instead assumed by a specific department/person or shared by several departments/people. Likewise, cooperation actions with suppliers, on the one hand, and universities and technological centers, on the other, represent key enablers of digital innovations for SMEs. Furthermore, an entrepreneurial orientation, resulting in a high growth ambition, is found to have a positive impact on the development of innovation in the company’s production processes through digitalization.

The analysis also shows that the impact of the mentioned determinants of digital innovation in SMEs varies to some extent between microenterprises and other SMEs. Microenterprises exhibit a more simplistic approach to digital innovation, relying more on the personal characteristics of entrepreneurs/managers and less on complex knowledge resources. When specialized knowledge is needed, collaboration with suppliers, universities, and research centers plays a critical role in leveraging external knowledge. In contrast, larger SMEs face higher complexity in digital innovation activities, requiring more specific knowledge inputs, greater staff involvement, and ambitious, professional management of the digitalization process.

These results suggest that the technological dimension is not the principal obstacle for SMEs in their digitalization projects since the basic technological resources to digitalize this type of company are already available and accessible. Therefore, human capital and organizational/relational dimensions are observed to be more determinant in the approach of SMEs towards their digital transformation. On the one hand, the capacity to incorporate digital technologies into the firm’s product and process in an innovative way is directly related to the digital skills available in the company, which condition the firm’s capacity to absorb new knowledge in the digital area, assimilate such knowledge, and translate it into new digital products and processes. On the other hand, organizational/relational aspects are found to be crucial for digital innovation in SMEs. From an internal perspective, the results presented in this paper reveal that a spontaneous and unplanned approach towards digitalization based on entrepreneur’s/manager’s personal initiative is not the most efficient option for SMEs. In contrast, SMEs should develop a formalized, strategically designed, and planned approach towards their digital transformation to successfully convert their efforts into product and process innovations. Furthermore, it is critical for SMEs to develop an open innovation approach for their digital transformation by means of intelligently managing their relationships with other actors (such as ICT consultants, suppliers, universities, and research centers) to overcome the limitations that their size implies. The results presented in this paper also suggest that SMEs need to adapt their organizational capabilities along their growth trajectory in order to accommodate the increasing complexity of the digital innovation process.

The findings of this study involve a variety of implications from the management and public policy perspectives. Our results highlight that investing in human capital and training is a necessary component of any action plan to address the challenge of the digital transformation of SMEs. Public administration could contribute towards this objective, although the most efficient approach does not directly involve carrying out training programs, but instead stimulating, with incentives, the training initiatives demanded by SMEs and facilitating their access to ICT consulting services. The results presented in this paper also emphasize the importance of interaction and partnership within digital ecosystems. Small and medium-sized enterprises should strive to contribute towards the formation of these networks and insert themselves into such existing structures. Likewise, public administration can play a role as a catalyst for the building and development of digital ecosystems in which SMEs can successfully participate.