Service innovation and internationalization in SMEs: antecedents and profitability outcomes

The development of small- and medium-sized enterprises (SMEs) is essentially influenced by the extent of their success in innovation and internationalization (Louart and Martin 2012; Hagen et al. 2014), and innovative SMEs are more likely to export than non-innovative SMEs (Saridakis et al. 2019). Innovative organizational culture is linked to rapid entry into foreign markets (Knight and Cavusgil 2004; Jones and Coviello 2005; Chetty and Stangl 2010; Onetti et al. 2012). McDougall and Oviatt (2000, p.903) already incorporated innovation and internationalization into the very definition of international entrepreneurship (IE), i.e., “a combination of innovative, proactive and risk-seeking behavior that crosses national borders and is intended to create value in organizations”. This definition brings an international dimension to the concept of entrepreneurial orientation (Covin and Slevin 1989, 1991; Lumpkin and Dess 1996; Miller 2011), and later studies have indeed identified positive linkages between higher levels of entrepreneurial orientation (EO) and internationalization decisions, activities, and performance in SMEs (Jantunen et al. 2005; Kuivalainen et al. 2007; Ripollès-Melià et al. 2007). However, extant studies have also left several aspects of entrepreneurial internationalization and innovation relatively unexamined.

One such area relates to clarifying the relationship between entrepreneurial internationalization and innovation further. The two have been found to be intertwined (Onetti et al. 2012; Denicolai et al. 2015), but whether innovation is a driver of internationalization (Golovko and Valentini 2011; Veglio and Zucchella 2015; Calvo and Villarreal 2019; Weerawardena et al. 2019) or rather, if the opposite is the case (Kafouros et al. 2008; Zahra et al. 2009; Genc et al. 2019; Vrontis and Christofi 2019) and under which conditions, could do with further study as the extant literature has found both types of relationships, and the linking literature overall is of recent origin (Hagen et al. 2014). As Veglio and Zucchella (2015, p. 139) note, there are studies on the dynamics of innovation and internationalization, but mostly from the context of large companies (see Cano-Kollmann et al. 2016). Recently, Saridakis et al. (2019) have found that the link between innovation and internationalization is also dependent on the type of innovations, and Weerawardena et al. (2019) posit that learning capabilities can underlie the development of innovation in born globals. Bello et al. (2016) also recently examined how innovativeness of service offerings relates to financial performance, but that study did not account for internationalization, which we in this study aim to do.

Specifically, the antecedents and outcomes of internationalization- and innovation-related activities in SMEs should be clarified further in order to establish what types of trade-offs do SMEs have to consider in their internationalization and innovation efforts: Although both internationalization and innovation activities may reside in the organizational culture (Dimitratos and Plakoyiannaki 2003), in practice, a SME may not have the resources necessary to engage seriously in both simultaneously. Indeed, innovativeness may, in some cases, hinder internationalization efforts (Dai et al. 2013). Such a situation presents SMEs with the necessity of trade-offs.

Second, extant studies examining innovation in the context of SME internationalization tend to examine different types of innovativeness, namely product innovation (e.g., Zahra et al. 2009), design innovation (Veglio and Zucchella 2015), and business model innovation (Rask 2014). However, the role of service innovation in the context has received very little attention so far (Weerawardena et al. 2019). This is despite the fact that service innovation as concept that has performance implications (Bello et al. 2016), is distinct scientifically, and offers promise for cross-disciplinary research in particular (Karniouchina et al. 2006). Instead, research on entrepreneurship and innovations in general has tended to be mostly limited on examining product innovations instead (Oly Ndubisi 2014). As internationalization can be related to the type of innovation in question (Denicolai et al. 2015; Saridakis et al. 2019), examining the role of service innovation in international entrepreneurship can bring forth added knowledge to complement and extend the extant studies concentrating on different types of innovations (Rask 2014; Veglio and Zucchella 2015).

Indeed, a large portion of entrepreneurial internationalization occurs in high-technology fields where service innovation is a core element in corporate decision-making (Van Riel et al. 2004). This is particularly the case with born globals (Rennie 1993; Madsen and Servais 1997), which tend to be knowledge-intensive in nature (Kuivalainen et al. 2012) and to operate in service-oriented industrial sectors such as software (Bell 1995). There is thus a clear need to examine the role of service innovation in international entrepreneurship.

This study aims to respond to these omissions in literature, by illustrating the relationships between service innovation and internationalization on successful entrepreneurial internationalization. As most existing studies have not assessed both the antecedents and the related outcomes of internationalization and innovation in general, we also explore the antecedents of both the service innovation (SI) and international expansion in SMEs, and link the two to performance outcomes, namely to profitability. This is due to earlier literature examining the impact of international expansion on business performance having reported contradictory and ambiguous results (see, e.g., Capar and Kotabe 2003; Lu and Beamish 2004; Thomas and Eden 2004), and most of the studies in question having been conducted in manufacturing contexts (Capar and Kotabe 2003). Only a few studies in IE assess the profitability aspect of internationalizing SMEs, and those that do give conflicting results (Lu and Beamish 2001, 2006; Qian and Li 2003; Zhou et al. 2007). Thus, there is also a need for further clarification on the relevant question of how internationalization affects the profitability of SMEs. Given that engaging in both internationalization and innovation may stretch the financial and other resources of SMEs, which by nature are resource-constrained to begin with (Madsen and Servais 1997), maximizing profitability may be of the essence for an innovating SME aiming for both foreign-market entry and long-term survival.

The extant literature identifies two concepts in particular that may, in part, underlie and determine the extent of both service innovation and internationalization: human capital (Becker 1964; Youndt and Snell 2004) and entrepreneurial orientation (Miller 1983; Covin and Slevin 1989). Genc et al. (2019) have recently posited that entrepreneurial orientation can mediate a relationship between internationalization and innovation performance, while the study by Bello et al. (2016) finds that entrepreneurial orientation and human capital predict the innovativeness of service offering in Indian professional service firms. While the latter study did not account for internationalization in their empirical model, it does provide the basis for postulating on these antecedents of service innovation also in international entrepreneurship. In general, both human resources of SMEs and their entrepreneurial orientation have been found to drive business performance (Grimmer et al. 2013). However, there is only limited amount of research examining causal relationships between entrepreneurial orientation and the multitude of outcomes, e.g., firm profitability, international scope, and product development performance, to which it has been connected (Wales et al. 2013).

Therefore, this study examines the impact of human capital (HC) and entrepreneurial orientation (EO) on internationalization and service innovation of SMEs, and the subsequent impact of the latter two on the profitability of internationally operating SMEs. We continue as follows: Sect. 2 gives a more in-depth literature review of the relevant research, examining the theoretical foundations underlying the proposed relationships between the antecedents of DOI and SI and their profitability outcomes; Sect. 3 introduces the research methodology and describes the data collection procedure and the measures we developed; Sect. 4 gives the empirical results, and we conclude in Sect. 5 with a discussion of their implications and limitations, also identifying potential future research avenues.

We tested the hypotheses by means of structural equation modeling (SEM). The most common estimation procedure within SEM is maximum likelihood, which is used for small sample sizes of 50 and has generated valid results, although the appropriate sample size is usually between 100 and 200. The significance level should be dropped to 0.025 or 0.01 with smaller sample sizes due to certain statistical characteristics of maximum likelihood estimation (Hair et al. 1998, p. 605; p. 613). The path model of profitability was based on the regression analysis results. We used the lavaan.survey R package for the SEM analyses. Three goodness-of-fit measures are used to evaluate the model fit in SEM: absolute fit, incremental fit, and parsimonious fit (Hair et al. 1998, p. 611; Hooper et al. 2008). The maximum likelihood method with robust standard errors was used for the model estimation. All the variables were standardized during the estimation due to the impact of normality, skewness, and kurtosis on the results, and the covariance matrix was generated via the software. Although the one-step model in which both confirmatory factor analysis and regression analysis is performed simultaneously is preferred, the two-step model has been widely used (Hair et al. 1998, p. 600).

Chi-square test statistics, and the ratio between chi-square (χ2) and degrees of freedom (df) were used to evaluate the overall model fit. The value of the latter in the model with the profitability outcome was 0.835 (15.022/18), suggesting no significant correlations between the matrices of the hypothesized model and the sampled data in the models of international performance and growth. The RMSEA value was 0.00 and thus below the suggested 0.08 threshold, indicating that errors in the fit of the covariance matrix were reasonable, although a good error of approximation was indicated in the profitability model. The SRMR value (0.04) was also below the recommended level of 0.08, indicating that the model was not misspecified. All in all, the fit measures (CFI 1.000, TLI 1.027, RMSEA 0.000, and SRMR 0.040) in the profitability model indicated an adequate fit to the data (see Table 4). Both the CFI and TLI values were above 0.95, suggesting a sufficient fit between the model and the sample data. This indicates that there was little possibility of measurement error because the difference between the null model and the sampled model was small. In addition, because all the measures in the model were at an adequate level, we checked that the SRMR values were between 0.06 and 0.03, which lowers the likelihood of error.

Figure 1 depicts the overall model. The results give partial support to H1. Namely, the risk-taking dimension of EO (H1c) did have a positive and significant (0.29, sig. < 0.01) impact on DOI. We did not consider this a surprising result given that entering foreign markets tends to require SMEs to focus their resources on internationalization, thus assuming a risk-taking strategic posture. The impact of innovativeness on DOI, although statistically significant, was nevertheless negative (− 0.19, sig. < 0.05). Thus, more innovation-oriented SMEs tended to exhibit lower levels of DOI, and H1a was not supported. The proactiveness dimension of EO was not associated with higher levels of DOI in the SEM model; thus, H1b was not supported either. In sum, the impact of EO on DOI was complex in that increased risk-taking led to higher levels of DOI, whereas a more innovative posture had the opposite effect.

Our second main hypothesis (H2) postulated a positive impact of EO on SI. As Fig. 1 indicates, the results were similarly dependent on the EO dimension in question: the innovativeness dimension had a significantly positive impact (0.35, sig. < 0.01). Although this was not a surprising result in itself, we note that the impact was the opposite to that of DOI, suggesting that in assuming an innovative posture the SMEs were, in fact, prioritizing service innovation at the cost of expanding their international activities. Additionally, the risk-taking dimension of EO was not associated with SI, and the relationship of the proactiveness dimension, although positive, was non-significant (0.22, sig. < 0.05). Thus, H2 was partially supported through the support for H2a.

The third hypothesis postulated a positive relationship between HC and DOI. As Fig. 1 indicates, higher levels of HC led to higher levels of SI rather than more extensive foreign operations. Thus, H3 was not supported. Conversely however, H4 was partially supported in that the skills and know-how dimension of HC was positively and significantly (0.25, sig. < 0.05) associated with higher levels of SI. The creativity dimension was positive (0.17) but not statistically significant in our model.

Finally, hypotheses H5 and H6 postulated a positive impact of SI and a negative impact of DOI (respectively) on the profitability of the international SMEs. As Fig. 1 shows, H6 was supported (− 0.33, sig. < 0.01). In other words, higher levels of DOI among the companies led to decreased profitability, whereas SI was not associated to profitability in our model.

In sum, we conclude that the profitability model suggested a sufficient fit to the sample data, and thus indicated the following: the risk-taking dimension of entrepreneurial orientation positively and significantly affected DOI, and the innovativeness and proactiveness dimensions significantly and positively affected SI. Furthermore, according to the model, DOI was significantly negatively associated with profitability. Overall, the SEM results show that EO and HC may be antecedents of internationalization and service innovation among SMEs, but their respective impacts are largely contrary. The implication is that the aptitude of SMEs to engage in either service innovation or internationalization may preclude them from achieving the other. The findings also indicate that SMEs may have to make a choice between innovating and internationalizing, and if they were to choose the latter, they may have to be prepared for a negative impact on their profitability. One reason for this may be that increasing DOI requires a risk-taking entrepreneurial attitude, which may result in an increased likelihood of market failure.

While entrepreneurial internationalization and innovation have been linked together in international entrepreneurship for a long time (e.g., Knight and Cavusgil 2004), studying the two jointly has recently started to gain steam in the IE field (e.g., Veglio and Zucchella 2015; Denicolai et al. 2015; Weerawardena et al. 2019). However, the role of service innovation in the international entrepreneurship domain overall has received little attention, and thus, the aim of this study was to seek empirical clarification concerning the antecedents and outcomes related to service innovation in the context of SME internationalization. The results of this study indicate, first, that entrepreneurial orientation underlies both internationalization and service innovation in SMEs, but that the impact depends partly on the EO dimension, as the innovativeness dimension of EO has significant but opposite effects on DOI and SI, suggesting a trade-off between innovation and internationalization when it comes to the entrepreneurial culture in SMEs. This is an interesting result when contrasted to those of Genc et al. (2019), who found EO to mediate between internationalization and innovation outcomes. Contrary to that study, we did not find EO to have a mutually beneficial and enriching role in linking innovation and internationalization outcomes, rather that the individual components of EO predict (service) innovation and internationalization separately. This difference may be due to the fact that Genc et al. measured innovation performance through the number of patents owned, thus their study context was not service innovation specifically. In addition, while Saridakis et al. (2019) found more innovative firms to also export more, the results of the present study contradict that result, by finding that the innovativeness dimension of EO has positive effect only on the extent of service innovation, but a negative one on the degree of internationalization. We suggest that these results may be due to the fact that the relationship between innovation and internationalization can depend on the type of innovation in question (Saridakis et al. 2019). The present study suggests that with service innovation, there may be a trade-off between engaging in innovation and international expansion.

Human capital would be an antecedent to SI, but not to the degree of internationalization among the SMEs and, although increasing the scale and scope of internationalization resulted in lower profitability among the SMEs, engaging in service innovation did not incur profitability costs for them. Consequently, our findings also imply that accumulated human capital may have neither a direct nor an indirect association with the profitability of an internationalizing SME. Our analyses did not yield evidence of a relationship between SI and profitability, but we did find a negative relationship between DOI and profitability. In this, our results are in conflict with Qian and Li’s (2003) finding of a positive impact of DOI on SME profitability, while confirming those reported by Lu and Beamish (2006) indicating that exporting has a positive impact on growth, but a negative one on profitability. This may imply that, as Zahra et al. (2000) suggest, the relationship depends on the characteristics of the firm.

The main contribution of this study is multi-faceted. First, in examining both service innovation and internationalization jointly in the SME context, we have highlighted their potential antecedents of entrepreneurial orientation and human capital, while at the same time as investigating an under-researched internationalization outcome in profitability. Thus, the study adds to the knowledge and arising discussion on the role of innovation in internationalizing SMEs (Genc et al. 2019; Saridakis et al. 2019; Vrontis and Christofi 2019; Weerawardena et al. 2019), while also extending the study by Bello et al. (2016) to account also for internationalization outcomes. Second, the findings contribute to the literature on the determinants of firm performance, and on the successful internationalization of high-technology SMEs (Bell 1995; Coviello and Munro 1997; Moen et al. 2004; Saarenketo et al. 2004; Vadana et al. 2019)—a context relevant in IE. In particular, our results enhance knowledge of service innovation in the context of internationally operating SMEs (cf. Weerawardena et al. 2019): SMEs concentrating on service innovation are able to gain certain benefits by operating internationally but accruing these benefits may be complex and depend on the strategic goals of the firm. A further contribution is to add to the body of knowledge on the determinants and outcomes of DOI, the findings thus far having been mainly inconclusive and conflicting. Given that DOI had a negative relationship with the profitability of internationally operating SMEs while SI did not, the main implication of our analysis is that, at least in the short run, SMEs may remain more profitable through innovation-rather than internationalization-related activities.

Our finding that different dimensions of EO act as antecedents of both DOI and SI provides further empirical support for the suggestion in the literature that EO should be conceptualized as a multidimensional construct (Naman and Slevin 1993; Frishammar and Andersson 2009; Miller 2011; Cadogan 2012; Lomberg et al. 2017). On the other hand and in conflict with some of the extant literature on the topic (e.g., Fletcher 2004; Knight and Cavusgil 2004; Genc et al. 2019; Saridakis et al. 2019), we found that different dimensions of EO had varying impacts on internationalization and service innovation, namely that innovativeness had a positive impact on SI, and a negative relationship with DOI. Our results suggest that innovativeness among top management may in fact decrease DOI, which could imply that innovative managers tend to focus on developing and commercializing innovations and therefore allocate more resources to the process of innovating. Consequently, a trade-off may present itself due to the limited resources SMEs possess, forcing them to spend less on engaging in internationalization.

Moreover, we distinguished a positive relationship between the proactiveness dimension of EO and SI, and between the risk-taking dimension and DOI. Surprisingly and contrary to what de Jong et al. (2003) and den Hertog (2010) suggest, in our analyses, risk-taking had no influence on SI. We believe this may be attributable to the fact that service innovations are carefully planned and researched, and that management risk-taking does not play a role. On the other hand, the lack of a significant relationship between proactiveness and DOI implies that international expansion in these companies may be more an act of risk-taking than a carefully considered strategic choice, which may further imply that in some cases, SMEs follow effectuation logic (Sarasvathy 2001) rather than strategic causality.

We also found a relationship between human capital (HC) and SI, but not between HC and DOI. Instead, international experience was positively related and industry experience negatively related with DOI. This suggests that top management’s international experience may play a more crucial role in internationalization in these types of companies than employees’ skills and capabilities. The role of international experience is acknowledged in the literature (e.g., Reuber and Fischer 1997; Manolova et al. 2002; Ruzzier et al. 2007), but in contrast to the results reported by Shane (2000), Westhead et al. (2001) and D’Angelo and Presutti (2019), our findings imply that industry experience may have a negative impact on DOI. In addition, overall educational level did not have a significant effect on SI or DOI, which contradicts Wang and Chang’s (2005) suggestion that educational background could be a dimension of human capital. Although there may be several explanations for the contrasting result in the present study, it could have had something to do with the nature of the industry: many software SMEs are startups founded by students, university graduates, or serial entrepreneurs who may vary in educational background.

In combination, the results regarding the effects of the EO dimensions and human capital on SI indicate that innovativeness among top management has a bigger impact than employee creativity in that the innovativeness dimension had a highly significant and positive relationship with SI, whereas in the case of creativity, the impact was only partially significant and positive. This could also imply that service innovations in SMEs tend to be created and developed from the top down, meaning that members of top management come up with new ideas that are then developed and commercialized together with employees. Furthermore, our findings also imply that SMEs may not be taking advantage of the potential of employees when assessing the role of skills and know-how in SI, or they may not see their creativity as beneficial. Notably, this could also imply that employees do not share their ideas sufficiently within the organization.

We also believe that our findings have relevant managerial implications. First, one could argue that the role of human capital in the development and commercialization of innovations is important. The human capital of both top managers and employees matters, although the creativity of employees matters less than their skills and know-how. The implication here is that companies should perhaps focus more on applicants’ skills and know-how rather than their creativity and innovativeness in the hiring process. Furthermore, the importance of international experience as well as skills and know-how in DOI and SI highlights the role of training and the opportunity to undertake international assignments. Second, as opposed to what is suggested in the literature, our findings do not support the adoption of a dual strategy combining innovation and internationalization, not at least if the internationalizing SME aims to maintain profitability. SMEs should therefore focus first either on innovation or on internationalization.

The main limitations of this study include its single-country and single-industry focus, which limits the generalizability of the results (cf. Bryman and Bell 2003 p. 105). The proposed model should subsequently be applied across industries, and a cross-country study could reveal more about the potential generalizability of the results obtained. For example, Lomberg et al. (2017) have shown that in the case of low- and high-tech firms, the degree to which each dimension of EO explains the variance in performance differs. Our findings may be valid in the context of high-technology SMEs in the Scandinavian countries and in other small economies that are similar to Finland. They may also be relevant to knowledge-intensive SMEs aiming at international expansion as well as to start-ups, which tend to be innovative and eager to internationalize by nature. The use of panel data would also allow more detailed examination of the DOI-profitability relationship.

Second, the sample mainly comprised small service companies in the ICT industry. Another limitation and, consequently, a potential avenue for future research is that we did not account for the radicalness of service innovations. This could also require applying den Hertog’s (2010) six-dimensional model of SI, given that we used the four-dimensional model on account of its fit to our specific empirical context (i.e., technology, which plays a central role in ICT companies, is at the center of the four-dimensional model). This study was conducted in a specific cultural environment, and EO in particular has been found to be influenced by culture (Lee et al. 2011). Expanding the empirical view to cultures other than Finland might provide us with more information on the dynamics of EO with SME innovation and internationalization. There is some evidence that in emerging markets, the role of institutions and institutional development may impact the dynamics of internationalization and innovation (e.g., Wu et al. 2016). In order to make further comparison, future research could thus assess the role of the institutional environment from the developed economies point of view. We also note that the 10% response rate in the empirical sample was not particularly high. Thus, we considered the possibility of biases arising from the response rate and, as Rutherford et al. (2017) recently found in their review, low response rates are not expected to bias results in the entrepreneurship domain.

For the future, the significance of international experience and the non-significance of human capital in the human-capital-DOI relationship imply that, if structured differently, human capital could yield more evidence about the association. The structure suggested by Huang (2003) and Neal (1995) could give further insight into these dynamics. Furthermore, including tacit and explicit technological knowledge in the measures instead of educational level could provide different evidence about the effect of human capital on DOI and SI.

Another potential future research avenue could be taking the unit of analysis to the individual level, as human capital has been found to lead to accumulating entrepreneurial and managerial competencies (Kungwansupaphan and Siengthai 2012), as well as to overall venture success (Hormiga et al. 2011; De Cleyn et al. 2014). Examining the mindset of SME managers on the innovation-internationalization behavior activities among SMEs, as both global mindset and global orientation may affect the behavior related to internationalization (Felício et al. 2013). Moreover, gender may also have an influence in the entrepreneurial intentions of individuals (Díaz-García and Jiménez-Moreno 2010) and thus by extension, on their individual entrepreneurial orientation (Goktan and Gupta 2013). While our study did not approach the examined phenomena from the point of view of the individuality or the gender of the respondents, extant studies do thus indicate that accounting for the gender of the entrepreneur might yield additional insight into constructs such as EO in this context. In addition, social capital could also be considered an antecedent of DOI and SI, given the recognition of the important role of relationships in contemporary literature on internationalization and innovation (e.g., Chesbrough 2011, p. 35; Johanson and Vahlne 2011). Moreover, service innovations tend to require collaboration between the service provider and the customer or other external parties (Alam and Perry 2002; Magnusson 2003; Kristensson et al. 2008; Kuusisto and Riepula 2011).

The dynamics of internationalization and innovation have recently gained attention in research on international entrepreneurship (e.g., Hagen et al. 2014; Onetti et al. 2012; Veglio and Zucchella 2015; Saridakis et al. 2019; Weerawardena et al. 2019). However, most of the extant research incorporating innovation and internationalization is still from the context of large companies (Cano-Kollmann et al. 2016). It is important to distinguish between different types of innovations in the context of internationalization (Denicolai et al. 2015; Saridakis et al. 2019). Thus, this study has sought to add to the knowledge on a type of innovation that extant research in IE has mostly neglected, service innovation, and to highlight the common antecedents and performance implications underlying service innovation and successful internationalization of SMEs in a joint setting.