The implementation of Industry 4.0 practices in small and medium-sized enterprises (SMEs) has taken place through a variety of channels (Meindl et al., 2021), among which cloud computing remains the most prominent one (Harvey Nash, 2022; Moeuf et al., 2017). Digital transformation aims to trigger improvements in the competitiveness of SMEs through innovations implemented at different firm levels (Teoh et al., 2022). In this regard, adopting digital technologies requires substantial investment by firms, which, if absent, would decrease their capacity to address potential competitive challenges arising in the market (Oliveira et al., 2022; Cambrea et al., 2021).

Digital technologies are strengthened during the internationalization process of firms (Chen & Kamal, 2016), with Müller et al. (2021) emphasizing the importance that the capacity of SMEs to assimilate and apply environmental information has for their innovation strategies. Similarly, Ojha et al. (2023) studied the dynamic strategic benefits of virtual integration as a performance improver for SMEs. The literature has also ventured into the efficiency branch of operations research. Gao et al. (2023) applied a hybrid model consisting of the super-efficiency version of data envelopment analysis within a Malmquist framework to study the effects of digitalization on the servitization level of manufacturing SMEs.

The increase in the productivity of SMEs triggered by digital transformation has been constantly validated in the literature. The analyses performed encompass developed regions such as the US and Europe (Van Ark et al., 2021), as well as developing countries (Aly, 2020), with particular emphasis placed on Africa (Gaglio et al., 2022), China (Du & Jiang, 2022) and Russia (Romanova & Ponomareva, 2022). On the other hand, Radicic & Petković (2023) highlighted the fact that the literature has not generally considered the effect of digital transformation on innovation in SMEs.

There are plenty of potential channels through which digital technologies may affect the functioning of SMEs, ranging from the reduction of transaction costs and new product and process developments to improvements in logistics and communication within and across firms (Grover et al., 2022; Alcácer et al., 2016). In this regard, digital transformation has also had significant effects on the competitive behavior of European SMEs (Garzoni et al., 2020). The activities undertaken by European SMEs are indeed significantly conditioned by digital technologies, including their interactions with customers and suppliers (Matarazzo et al., 2021) and the development of new products and services (Khin & Ho, 2019). Skare et al. (2023) illustrated how the transformation induced by digital technologies conditions the main competences of European firms and their networks. These authors analyzed the effects of digitalization – measured via the Digital Economy and Society Index (DESI) index computed by the European Commission – on different activities of European SMEs, including customer access, competition, access to finance, input costs, skilled labor, exogenous shocks, global crises, and regulatory issues.

Our analysis focuses on the evolution of digital value chains and their effect on the innovation performance of SMEs in terms of patents, product, and process innovations. Value chains have been substantially affected by digital technologies, both via supply chain modifications (Hahn, 2020) and customer interactions (Frederico et al., 2020). The introduction of patents, new products, and processes is determined by the innovation strategy designed by the firms and the acquisition of internal and external knowledge (Guckenbiehl et al., 2021). In this regard, the limited capacity of SMEs to generate and invest in internal knowledge leads to external interactions as alternative sources of knowledge (Obradović et al., 2021). This latter strategy is enhanced by their digitalization processes, with SMEs displaying heterogeneous resource capacities (Eller et al., 2020).

Given these premises, we will define three main hypotheses suggesting positive effects from digital value chains, vertical and institutional cooperation on the innovation capacities and outcomes of SMEs. In a nutshell, we will test the potential positive impact of these digitalization variables on the capacity of firms to develop and introduce patents and technological innovations. The intuition describing the interactions analyzed empirically is presented in Fig. 3, where the main digitalization variables and controls used in the econometric regressions are listed. A detailed description of each variable and the corresponding references justifying their choice and describing their expected effect on the innovation outcomes of firms is presented in the next section. We conclude by noting that these hypotheses will be tested for large firms and SMEs as a whole and separately so that general and specific size effects can be identified and analyzed.

Fig. 3.

Conceptual framework within the innovation ecosystem.

(0.16MB).

Patent analysis constitutes a well-known standard measure of technological competitiveness (Zhu et al., 2023; Ahn, 2020). Radicic & Petković (2023) analyzed the impact of digitalization on product and process innovations, namely, technological innovations. These authors highlighted the fact that this potentially positive relationship has not been properly analyzed in the literature, which contrasts with the significant number of analyses performed regarding the effects of digital transformation on the productivity and efficiency of firms (Kádárová et al., 2023; Schuh et al., 2014).

The interdependencies between both types of innovation have been consistently highlighted in the literature. Product innovation aims at increasing the profitability of firms, emphasizing the interactions between firms and consumers (Pesch et al., 2021; López-Cabarcos et al., 2020). Innovations are introduced in production processes to reduce costs and improve quality, enhancing interactions across the value chain (Domnich, 2022; Gogokhia & Berulava, 2021; Mohnen & Hall, 2013).

Three main variables are used to describe the digitalization capacity and implementation of SMEs.

Control variables are introduced to capture the main characteristics of markets and firms. We list the variables below together with the expected effects from their respective inclusion. Our approach differs from that of the recent paper by Radicic & Petković (2023), who used the internal R&D performed by SMEs as a moderating variable when considering the relation between the digitalization of firms and their technological innovations.

The data on manufacturing firms used to perform the empirical analysis has been retrieved from a database named Survey on Business Strategies (Encuesta sobre Estrategias Empresariales – ESEE).1 This survey has been conducted periodically since 1990 under the auspices of the Spanish Ministry of Finance and Public Function and encompasses manufacturing firms in Spain with more than 10 employees. ESEE collects annual data on firms from 20 manufacturing industries, covering the total number of firms with more than 200 employees and a representative sample of firms employing between 10 and 200 workers evaluated through a random stratified sampling scheme.

This data source provides a large amount of information at the micro level on the productive activities of firms, their business strategies and performance, while offering specialized information regarding their technological activities. As a result, ESEE has been used in a variety of contributions analyzing the determinants and sources of innovative capabilities across firms (Radicic & Pinto, 2019; Barge-Gil et al., 2011; Cassiman et al., 2010; Salomon & Jin, 2008; Huergo, 2006; Álvarez & Molero, 2005).

Since 2000, ESEE includes a set of variables related to the firm's use of internet-based digital communication technologies, such as e-commerce transactions within the supply value chain. These variables will allow us to analyze how much the digitalization of the value chain contributes to the innovative performance of firms. The sample used in the empirical analysis covers the period 2000-2017 and comprises 1,369 firms, out of which 862 are SMEs (between 10 and 250 employees) and 508 are large firms (more than 250 employees). Note that despite the availability of information regarding the digitalization of firms, its consequences for their technological activities remain mainly unexplored.

To test whether digitalization contributes to the innovative performance of firms, the econometric model applied takes the form of a knowledge production function that allows us to explore to what extent different knowledge sources, beyond formal R&D activities, affect innovation outputs (Griliches, 1979). More precisely, the empirical model is defined as follows:

We define three empirical models, one for each innovation output, as follows:

Each dependent variable is regressed against the explanatory factors discussed in the previous section. The main independent variable capturing the effect of digitalization on the innovative performance of firms is the digitalization of the value chain (Digital value chain). This variable takes the value 1 if the firm sells to customers or purchases from suppliers through Internet-based channels that require sophisticated electronic data interchange processes.

The second set of independent variables accounts for the role of innovation networks in the success of innovation processes. The insertion of firms into innovation networks is operationalized through their technological cooperation with customers and/or suppliers (Vertical cooperation) as well as universities and R&D centers (Institutional cooperation). These two regressors are dummy variables that adopt the value 1 if the firm cooperates with other agents or 0 if not, representing the mean effect of knowledge networks on its innovation capabilities.

As control variables, we include the R&D intensity of firms, measured as the percentage of R&D investments over sales, since R&D is considered one of the main factors determining technological innovation (Acs and Audretsch, 1988; Hall et al., 1986). We also include the following control variables related to the characteristics of firms that the literature on economics of innovation highlights as positively associated with technological performance: Productivity –measured in terms of value added by employee –, size – measured by the number of employees – and firm age.

A dummy variable capturing the Foreign ownership2 of firms is also included to control for the innovative activities that subsidiaries carry out in host countries, which are conditioned by the competence creating or competence exploiting mandates of the multinational enterprise subunit (Cantwell & Piscitello, 2015; Cantwell & Mudambi, 2005). Finally, we also include the Export intensity of the firm – measured by the export volume as a share of sales – to capture the learning process that takes place by exporting, characterized by the exchange of knowledge in foreign markets (Salomon & Shaver, 2005).

Table 1 summarizes the variables used in the analysis and presents some descriptive statistics. It can be observed that large firms are more innovative than SMEs. In the case of patents granted, the frequency of innovation within large firms is 13.52 %, almost four times higher than for SMEs. These percentages increase for both groups of firms when considering product and process innovations. Even though substantial differences remain between both groups, the technological gap that separates them decreases. For instance, 36 % of large firms introduced product innovations relative to 16.21 % of SMEs. Note that SMEs and large firms are more engaged in process innovations than in the rest of technological activities, validating the trend observed in Fig. 2.

When considering digitalization, we observe that more than 30 % of firms are involved in digital value chains, with this share increasing up to 41 % in the case of large companies. Regarding technological cooperation, the percentage of firms cooperating with customers and suppliers is similar to that of firms cooperating with universities and R&D centers. However, large firms are more involved in innovation networks.

As expected, control variables illustrate that large firms are more R&D-intensive, productive, and export-oriented than SMEs while also being characterized by substantial heterogeneity. In addition, 47 % of large firms are foreign-owned, a percentage that decreases to 11 % when considering SMEs.

Finally, Table 2 presents the correlation matrix, highlighting the absence of any significant relationship among the variables considered.

We have shown how digital transformation has a positive effect on technological innovations across Spanish firms, a quality that is particularly evident among SMEs. The literature has consistently emphasized the fact that the innovation capacity of larger firms and their implementation of digital transformation processes are both higher than those of smaller firms. In this regard, we have observed how digitalization helps the latter improve their innovation performance, particularly when considering patents and product innovations. At the same time, cooperation with customers, suppliers, and research institutions across the value chain constitutes a source of knowledge consistently exploited by all firms to develop technological innovations.

Another important result is the consistently positive and significant contribution of R&D activities to technological innovation for all firms. This result contrasts with Usai et al. (2021), who highlighted the standardized quality of the knowledge driving digital transformation. As a result, these authors suggested that SMEs engaging in R&D activities might be unable to use it to introduce new products and processes, while those SMEs not performing R&D activities could rely on this knowledge when developing innovations.

The remaining effects, particularly those corresponding to the size, age, and productivity variables, coincide with the intuition provided throughout the literature review section.

A couple of important remarks are due, requiring a more detailed analysis of the differences in innovation incentives and outcomes between domestic and foreign firms. To do so, we select a subset of domestic and foreign firms from the database according to their prevalence throughout the whole period of analysis. That is, given the length of the period considered, many firms enter and exit the database due to a variety of reasons. Thus, we have selected domestic and foreign firms with observations available for each sample year, allowing us to define a consistent profile of their evolution. The subset is composed of 397 firms, out of which 320 are domestic and 77 foreign. Two of them shifted from domestic to foreign ownership through the period analyzed.

Fig. 4 illustrates the average value – across time periods and firms – of the binary variables defining patents and product and process innovations for domestic and foreign firms. The evidence presented by Almodóvar & Nguyen (2022) is validated across all categories. Foreign subsidiaries established in Spain throughout the whole period of analysis introduce a higher number of patents and technological innovations on average than domestic firms. In this regard, the negative innovation coefficients displayed by the whole set of foreign firms align with the application of strategies favoring competence exploitation over exploration (Griffith et al., 2021; Ramachandran et al., 2019). We must, however, note that different results could be obtained when categorizing firms by the technological content of sectors, a feature defining a potential line of future research.

Fig. 4.

Average patents and technological innovations by firm ownership.

(0.07MB).

Fig. 5 illustrates the similar R&D intensities exhibited by the domestic and foreign firms composing the consistent subset. Note that the subsidiaries displaying a less intense R&D behavior than domestic firms tend to be more productive.

Fig. 5.

R&D intensity and productivity among domestic and foreign firms.

Note: Variables are defined in natural logarithms.

(0.22MB).

Fig. 6 describes similar export intensities among domestic and foreign firms. Note how, as domestic and foreign firms grow older, they tend to increase their export intensity. Thus, the negative signs of the export coefficient could be due to the foreign orientation of companies constituting an attempt to increase their market share, absent any innovation incentives from international competition (Acikdilli et al., 2022; Aboushady & Zaki, 2021).

Fig. 6.

Export intensity and age among domestic and foreign firms.

Note: Variables are defined in natural logarithms.

(0.35MB).

Despite the potential advantages of digitalization, the adoption of digital technologies among SMEs is generally limited due to the financial and capital constraints derived from their relative size (Estensoro et al., 2022; Santos-Arteaga et al., 2020). In this regard, Saratchandra et al. (2022) highlighted the importance that external knowledge and the capacity of SMEs to source it have for innovation. This knowledge can be used to complement the one generated internally by R&D-based SMEs. On the other hand, if the SMEs are non-R&D-based, this knowledge would help foster their capacity to innovate.

The Spanish digitalization strategy is summarized in Ministerio de Asuntos Económicos y Transformación Digital (2023). This report emphasizes the limited progress achieved in the digitalization and R&D activities of SMEs over the last two decades and the subsequent priority assigned to correcting this tendency. Furthermore, the website dedicated to the digital transformation of Spain across its economic and social domains, http://espanadigital.gob.es/home, highlights the strategic importance of digitalization for SMEs and the general population.

The results obtained from the current analysis imply that the digital strategy implemented by Spain is paying off, a quality validated through the behavior of its DESI index. The consistently superior performance of the latter relative to the European Union (EU) average is described in Fig. 7. However, despite these features and as illustrated in Table 1, the digital competitiveness of Spanish firms has not been reflected in their innovation score.

Fig. 7.

Evolution of DESI scores across Greece, Germany, Spain, Romania, and the EU.

(0.12MB).

A couple of remarks must be made regarding this difference in scores. The main variables composing the dimensions and subdimensions that define the DESI index are presented in Table 6. Note that there is a prevalence of variables describing the social and institutional qualities of ICTs and their market penetration process. The index does not focus on the industrial characteristics of ICTs such as, for instance, their effect on labor productivity. This quality determines the evolution of the DESI index observed across the four reference countries described in Fig. 7.

These countries have been selected according to their relative performances in terms of the European Innovation Scores illustrated in Fig. 1. Greece and Romania define the upper and lower score limits, respectively. Germany, a standard European reference, displays the opposite trend relative to Spain, namely, a higher innovation score than digital competitiveness. This latter variable is indeed quite similar for both countries while Spain exhibits a substantially lower innovation score than Germany.

As can be observed when comparing Figs. 2 and 7, Romania performs poorly in both scenarios in terms of all variables. The high innovation score displayed by Greece is not reflected in the evolution of its DESI index, which seems more related to its low digital competitiveness. Germany, which displays one of the highest innovation scores among European countries, performs below Spain in terms of DESI throughout the whole period the index is available. Thus, as can be inferred from comparing both figures, the DESI index and the innovation score lack a clear relationship, while digital competitiveness seems closer to the actual behavior of the DESI scores.

All in all, given the fact that digital transformation and cooperation have a positive effect on technological innovations across Spanish firms, particularly SMEs, any policy designed to foster the digitalization incentives of firms and the generation of stable links among the actors composing the value chain should be continued and reinforced.

The database used to perform the empirical analyses contains many variables whose effect on the innovation incentives of firms could be studied. Quite a few additional variables and scenarios have been formalized to study the relationship between digitalization and innovation performance in SMEs. For instance, Haug et al. (2023) highlighted the conditioning quality of technological orientation, while Bouwman et al. (2019) focused on business model experimentation and strategy implementation. Gruber (2019) listed four features of SMEs that limit their adoption of digital technologies: lower exposure to the consequences of digitalization, lack of managerial vision, gradual transformation processes, and limited financial resources.

Thus, while the variables used in the current paper to analyze the relationship between digital transformation and technological innovations are standard in the literature, alternative scenarios could be defined and validated. Among these, relevant complementary analyses should aim at categorizing the impact of digitalization on SMEs across sectors of different technological intensity. These extensions should help provide additional intuition regarding the differences observed between the digital competitiveness of Spanish firms and the value of their innovation score.