Interorganizational Network Embeddedness and Performance of Companies Active on Foreign Markets

Abstract Purpose: The purpose of the article is to identify the impact of interorganizational network embeddedness and relationships on companies’ performance by exploring the perspective of companies active on foreign markets. Methodology: The article presents results of an empirical quantitative analysis. The data has been collected through an anonymous online and postal survey. The analysis focuses on 345 companies active on foreign markets and involves a Mann-Whitney U test and a regression analysis. Findings: The conducted analysis points to the importance of the degree of interorganizational network embeddedness for the performance of companies active on foreign markets. It shows that the higher the degree of the overall network embeddedness (measured with adaptations, trust and mutuality), the higher the company’s perceived performance relative to its competitors. Limitations: The analysis focuses on the degree of network embeddedness as a moderating factor of companies’ performance, excluding other complex factors that may affect their performance. Value: The analysis highlights the importance of interorganizational network embeddedness for the performance of companies operating on foreign markets. It is not limited to multinational enterprises, but shows the importance of embeddedness for companies of different sizes and structures. The analysis is followed by a set of managerial implications relevant to the subject presented in the paper.


Introduction
Companies are embedded in relationships with their customers, suppliers, and other entities (such as competitors or governmental bodies). This perception of companies' activities is reflected in two theoretical threads, namely the thread of network theory (Ford, Gadde, Håkansson and Snehota, 2011;Håkansson and Snehota, 1989;Håkans son and Snehota, 1995;Johanson and Mattsson, 1987) and the thread of embeddedness theory (Forsgren, Holm, and Johanson, 2005;Granovetter, 1985;Halinen and Törnroos, 1998;Hess, 2004;Polanyi, 1944;Sigfusson and Harris, 2013). This kind of perception of companies and their activities is important not only for cognitive reasons but also for providing managerial solutions for increased performance. Improvement in perfor mance is deemed to be a significant benefit following from business (also called 'inter organizational') relationships (Gadde and Snehota, 2000;Golonka, 2016;Håkansson and Snehota, 1995;Mazurek, 2014;Ratajczak-Mrozek and Małys, 2012). However, the question of what relationships characterized by what qualities lead to a higher performance remains still valid. To answer this question, embeddedness, understood as a rela tional gradable concept, constitutes an important contribution to the analysis of the impact of interorganizational relationships on companies' performance as it concen trates on the quality of their relationships with their surroundings. However, as Bres ciani and Ferraris notice (2016, p. 108): "nowadays relatively little is known about the impact of embeddedness on performance outcomes, such as firms' greater profitability or sustainable advantage in the marketplace".
The assumption that a company is embedded in interorganizational relationships is particularly important when one analyses a given company's international activity. Networks and relationships are becoming the dominant paradigm in the theory of internationalization (Johanson and Mattsson, 1988;Johanson and Vahlne, 2009). If a company has relationships with foreign customers and/or suppliers, it can be viewed as embedded in both domestic and foreign relationships. The question is if this differen tiated spatial embeddedness changes managers' perception of the influence of interorganizational network embeddedness on business performance.
Taking into consideration the abovementioned questions, the purpose of the article is to identify the impact of interorganizational network embeddedness and relationships on companies' performance by taking the perspective of companies active on foreign markets into consideration.

Theoretical background
Embeddedness is a concept rooted in sociology, introduced by Polanyi (1944), popu larized by Granovetter (1985), and later adopted by organization and business studies. In organization and business studies, research on embeddedness is characterized by a lack of definitional consistency as there are many "research perspectives on network embeddedness" (Halinen and Törnroos, 1998, p. 191). Two general ways of interpret ing interorganizational network embeddedness may be distinguished. First of all, network embeddedness is meant to describe "wider industrial settings (...) in which such business networks are embedded" (Fletcher and Barrett, 2001, p. 561) or "com panies' relations with, and dependence on, various types of networks" (Halinen and Törnroos, 1998, p. 187). In this broadest sense, embeddedness is used as a synonym of dependence on a given context (social, organizational or other).
Second of all, interpretation of network embeddedness focuses on specific characteristics of embedded relationships, most of which focus on adaptations. Andersson and Forsgren (1996, p. 497) define embeddedness as the "degree of adaptation of resources to other network actors". Adaptations are seen as a demonstration of commitment and trust in relationships (Andersson, Forsgren and Holm, 2001;Forsgren et al., 2005). Andersson, Blankenburg Holm and Johanson (2005, p. 32) argue that highly embedded exchange is "characterized by extensive interaction, mutual trust, and relationshipspe cific advantage". According to this interpretation, adopted also in the article, network embeddedness is determined by the intensity of adaptations and trust in relationships as well as the resulting interdependence and mutuality. Defined in this quality-related way, embeddedness becomes a gradable concept, with different possible degrees thereof. A similar approach to measurement and assessment of embeddedness, viewed as a con tinuous variable, is proposed also by other authors (Dacin, Ventresca and Beal, 1999;Dellestrand, 2011;Figueiredo, 2011;Forsgren et al., 2005;Nell and Andersson, 2012).
Interorganizational network embeddedness is used to analyse the impact of business relationships on companies' behavior and performance. Research of Ahuja (2000) and Knoben, (2008) shows that a high degree of interorganizational network embeddedness has a positive impact on business performance. A similarly positive impact on companies' market performance is associated with a high degree of technical embeddedness (Andersson et al., 2001). In terms of impact of interorganizational network embedded ness on companies' performance, the most developed research is that conducted within internationalization theory, which, however, focuses mainly on multinational enter prises. Here, the emphasis is placed on the socalled dual embeddedness, which is simultaneous embeddedness of a multinational enterprise (MNE) and especially its

JMBA.CE 147
Interorganizational Network Embeddedness and Performance of Companies Active... subsidiary in both internal (that is within an MNE) and external networks (that is out side an MNE) of interorganizational relationships (Ciabuschi, Holm and Martín Martín, 2014;Figueiredo, 2011). Hallin, Holm, and Sharma (2011) and Bresciani and Ferraris (2016), among others, underline the positive influence of dual embeddedness on business performance. In the context of an analysis of a subsidiary's performance, Oehmichen and Puck (2016, p. 17) have found "a positive moderating effect of changes in owner ship mode on the relationship between external and internal embeddedness and subsidiary performance". Many authors notice that external embeddedness positively affects a subsidiary's performance (Andersson et al., 2001;Nell, Ambos and Schlegelmilch, 2011). For example, Andersson et al. (2001) argue that the higher the degree of a subsidiary's external embeddedness, the higher its expected market performance.
Within the research on the impact of embeddedness on performance, some studies focus on particular effects leading to increased or decreased performance. The most extensively discussed positive effects include gaining access to valuable resources, innovations, and identification of market opportunities. These effects may concern both domestic and international activity of companies. It is argued that embedded interorganizational relationships "provide access to resources and knowledge that would not be accessible (or only at higher costs) through the market, which leads to (competitive) advantages for firms" (Knoben, 2008, p. 15). Embeddedness enables com panies to coproduce resources and utilize these resources more effectively (Chandra, Styles and Wilkinson, 2009). In terms of knowledge, "a high degree of relational embed dedness can be viewed as the capacity of the relationship to hold knowledge" (Anders son et al., 2005, p. 33). Resources, including knowledge, are important for innovation and product and technological development. Andersson and Holm (2002) underline that a high degree of technical embeddedness has a positive influence on a business unit's ability to develop new products or to improve its production processes. As high degree of embeddedness is an investment that influences the company's capacity to identify new information (Andersson and Holm, 2002, p. 364), network embeddedness fosters the creation, identification, and utilization of opportunities (Chandra et al., 2009). The most developed studies in the area of opportunities recognition are focused on the links between embedded interorganizational relationships and international activity of companies.
The impact of interorganizational network embeddedness on companies' performance tends to be perceived in terms of positive influence, but negative aspects of embed dedness should not be overlooked, though. A high degree of embeddedness requires relationship-specific investments and, consequently, is costly to handle. It can also lead to lockin effects (Anderson, Håkansson and Johanson, 1994) (Andersson et al., 2001). Companies may be stuck in interorgani zational relationships that may be characterized by a high degree of embeddedness, but yield little or no profit (Mitręga and Zolkiewski, 2012). Thus embeddedness can also have a negative effect on performance.
Although research on the link between network embeddedness and companies' per formance has been gaining significance, an observation by Bresciani and Ferraris (2016, p. 108) remains still valid: "nowadays relatively little is known about the impact of embeddedness on performance outcomes, such as firms' greater profitability or sustainable advantage in the marketplace". It is especially important for internationa lized companies. Many companies entering foreign markets and developing interna tional operations do so in order to secure the possibility of development and to improve their performance. However, the research on the impact of network embeddedness on performance in the area of internationalization theory focuses on multinational enter prises, that is to say mostly large companies. An important question is if interorganiza tional network embeddedness is equally important for performance of internationally active companies of different sizes (not only MNEs).

Methodology
The data presented in the article was obtained through anonymous postal and online surveys conducted from November 2014 to June 2015. For the sample frame, the Kom pass Poland database of business-to-business companies was used. A random sample of companies from all over Poland and all industries was selected (excluding public education and public administration). Both online and postal questionnaires consisted of the same questions, and each company received only either an online or a postal questionnaire. Additionally, the data obtained from both paper and online question naires was compared and did not reveal any differences related to the manner in which the data had been collected.
The postal questionnaire was sent to 2,000 companies and the link to the online version was emailed to 8,000 companies. The questionnaire was accompanied by a covering letter with a request that the questionnaire be filled in by representatives of the com pany's management. Because some companies in the sample could not be contacted (e.g. outdated addresses, liquidated companies), the paper questionnaire reached 1,910 companies. 226 responses were received, resulting in a response rate of 11.8%. The response rate for the online survey was only 2.4% (which was due to the large number

JMBA.CE 149
Interorganizational Network Embeddedness and Performance of Companies Active... of incorrect email addresses and messages being classified as spam). 419 responses were obtained in total.
The two most important variables for the article are the company's performance and the degree of embeddedness. The performance variable was associated with five items that reflected managers' perception of their company's performance, which included total profit, average annual sales growth rate, market share, return on assets, and the level of innovativeness of products/services/processes. The respondents were asked to assess these items for their own company in relation to direct competitors on a fivepoint Likert scale (1 -much worse, 2 -worse, 3 -almost the same, 4 -better, 5 -much better). As companies are very reluctant to share objective account-based data in interviews or questionnaires, even if these are anonymous, it seemed more reasonable to focus on perceived or subjective assessment of performance (a similar approach is adopted by e.g. Andersson et al., 2001;Fonfara, 2012;Oehmichen and Puck, 2016). The performance variable was defined as a mean of the five abovementioned items, which could range from 1 to 5. The internal consistency of the performance variable was tested using Cronbach's alpha, which amounted to 0.86, which confirmed that the proposed scale was -and is -a reliable measurement tool.
The variable degree of embeddedness is represented by four items reflecting the extent to which important relationships are characterized by adaptation, trust and mutuality. The items described the level of adaptation and trust from the perspective of the respondent's company and the perceived level of adaptation and trust from the per spective of the other actor involved in the relationship (either key customer or key supplier). The joint assessment of the four measures of adaptation and trust provides an indicator of mutuality (if adaptations are assessed as low on both sides, the degree of mutuality in the relationship is assumed to be low; the same relation holds for trust). Respondents were first asked to list relationships with the most important key customers and key suppliers. These relationships were then characterized by standardized questions covering among others adaptations and trust. These were measured on a five-point Likert scale (1 -I definitely disagree, 2 -I disagree, 3 -I partly agree and partly disagree, 4 -I agree, 5 -I definitely agree). The variable degree of embeddedness was defined as a mean of the total score of mutual adaptations and the total score of mutual trust in the relationships with (1) key customers and (2) key suppliers and can range from 2 to 10. The internal consistency of the variable of the degree of embed dedness was tested using Cronbach's alpha which in the case of embeddedness in relationships with customers equals 0.69 and in relationships with suppliers was 0.75, which confirms that the proposed scale is a reliable measurement tool.

Milena Ratajczak-Mrozek
Realizing that present embeddedness influences future performance (Andersson et al., 2001), the questions in the survey were constructed to take this bias into account. All the questions regarding characteristics of key relationships referred to the period of 2010-2014, while the questions about the performance referred to the end of 2014. Additionally an attempt was made to limit potential common method bias (Bresciani and Ferraris, 2016) resulting from the fact that the questionnaire was only completed by managers. Following recommendations among others of Bresciani and Ferraris (2016) and Podsakoff, MacKenzie, Lee and Podsakoff (2003), independent and dependent variables and related questions were placed in different sections of the questionnaire and different scales and response formats were used.

Data analysis
In the article, a total of 419 companies are analysed, focusing on 345 of them that conducted business activity abroad in 2014. Foreign activity was defined as having international customers and/or suppliers, thus it took into account both outward and inward internationalization (Welch and Luostarinen, 1993). 62 companies were only engaged in domestic activities and used as reference point for the analysis of companies active on foreign markets. In the case of 12 companies (2.9%) it was impossible to verify whether they were active or not on foreign markets because of the lack of data. The sample is described in Table 1.
Micro and small companies were the dominant group among the companies that did not conduct foreign activity (74.2%). As regards the companies that conducted foreign activity, only 16.2% of them were micro companies, while the majority consisted of small and medium-sized enterprises (69.8%). In terms of ownership structure, relatively more companies involved in foreign activity were foreignowned enterprises and par tially owned by foreign companies. Considerably more companies that did not conduct foreign activity were involved in services and trade (80.6%) than in the group of the companies that did (52.5%).
282 of internationalized companies answered the question about the forms of foreign activities (67.3% of the sample and 81.7% of companies active on foreign markets). The most commonly reported forms of internationalization were direct exports (58.9%), indirect exports (41.1%), imports (54.3%), and subcontracting (23%). Other forms were rare, accounting for between 5% and 0.7% of responses (it was possible to select more than one form). 14.6 % of 345 companies (50 companies) were exclusively involved in inward forms of internationalization.

JMBA.CE 151
Interorganizational Network Embeddedness and Performance of Companies Active... With the intention to identify differences in the perception of the impact of embedded ness on business performance, the first step involved an analysis of whether there was any difference between the frequency of particular effects related to performance and a given company's involvement in foreign activity. After establishing that the variables of interest were not normally distributed, a Mann-Whitney U test and a U statistic were applied for two groups of companies: those conducting foreign activity and those that did not. Table 2 presents the test results for the positive and negative effects of relation ships with key customers and key suppliers, where the existence of a statistically significant difference in the frequency of particular effects was confirmed. The analysis included both positive and negative effects. The results of the conducted Mann-Whitney U test reveal statistically significant differences in the assessment of particular effects related to performance between companies with and without foreign activity. In assessing the positive effects derived from interorganizational relationships with key customers, companies active on foreign markets have pointed frequently to such benefits as sales growth and development, internationalization of activity and benefits associated with innovation, namely access to knowledge and the creation of new products. As far as interorganizational relation ships with key suppliers are concerned, companies with foreign involvement appear to have placed more importance on internationalization of activity and access to unique resources. Another difference concerns the negative outcomes: companies without foreign involvement have reported the disadvantage associated with their dependence on key customers more frequently. One may conclude that companies engaged in foreign activity perceive positive effects of business relationships related to develop ment, innovation and, by no surprise -internationalization relatively more frequently.
The discovered differences in the effects related to performance do not yet prove the direct impact of interorganizational relationships and interorganizational network embeddedness on performance of companies active on foreign markets. Thus in the second step of the analysis, the link between the network embeddedness in relation ships and the performance of companies involved in foreign activity was analysed by checking the Spearman correlation between the variable degree of overall embedded ness in relationships with key customers and key suppliers and the performance variable. It is proven that there is a weak positive correlation between two variables of interest (r = 0.17; p = 0.01, two tailed). The results demonstrate that as the degree of embeddedness increases, so does the business performance.
In the third step of the analysis, the final test was conducted using a multiple linear regression analysis (concerning the sample of companies conducting foreign activity). In the estimated model the variable degree of the overall network embeddedness explains 2% of the variance of the dependent variable (performance). This is a rather low level, Milena Ratajczak-Mrozek but it is not uncommon in economic studies and can be regarded as acceptable for the purpose of the present analysis, as the main purpose of the analysis is not to achieve a high goodness of fit, but to identify the link between interorganizational network embeddedness and the resulting performance. The dependence is week and positive (with beta = 0.15). The value of the angular coefficient is 0.1 and is statistically significant, t(232) = 2.297, p<0.05. The coefficient suggests that the higher the degree of the overall interorganizational network embeddedness in relationships of a company active on foreign markets, the higher a given company's perceived performance relative to its competitors.

Conclusions
Relationships form an essential part of companies' activity regardless of whether these companies are active on foreign markets or work only with domesticallybased custo mers and suppliers. However, in the case of internationalized companies, the capability to manage such relationships may be a crucial source of improved performance and competitive advantage. It is especially important as many companies enter foreign markets to secure future development, growth, and -eventually -to improve their per formance. It has been shown in the analysis that companies active on foreign markets are more aware of the benefits of relationships, with such benefits taking on the form of innovation and development, than companies active only on domestic markets.
The conducted study showed that the perceived performance of companies active on foreign markets is increased the higher the degree of the overall network embedded ness in interorganizational relationships. This degree of embeddedness, in turn, is increased by particular characteristics of the relationships' quality, which is based on trust, adaptations, and mutuality. This answers the question of what relationships lead to higher performance. Embeddedness, seen as a relational and gradable concept, functions as an important contribution to the analysis of the impact of interorgani zational relationships on business performance. It is important inasmuch as a company may at least partly influence these characteristics (especially adaptations).
The conducted study offers a range of managerial implications. The conclusions are important for improving the competitiveness of companies operating on foreign mar kets. However, when operating on foreign markets, companies should consciously focus on selected relationships and make more adaptations within them. It is important to instantaneously evaluate the partner's (customer's or supplier's) willingness to reciprocate the offered mutual adaptation. Without such a mutuality it is difficult, or

JMBA.CE 155
Interorganizational Network Embeddedness and Performance of Companies Active... rather impossible, to reach a high degree of network embeddedness. At the same time, one has to keep in mind that it is not possible to have all highly embedded interorga nizational relationships and that it is not the proper direction of business activity. It is more important to choose key relationships and to focus on the degree of network embeddedness within these selected relationships, while maintaining the remaining contacts at a lower level of embeddedness.
The conducted analysis is not without limitations, though. It focuses on the degree of interorganizational network embeddedness as a moderating factor of companies' per formance, omitting other performance factors. And as the positive impact of a high degree of network embeddedness can be an important factor in improving business performance, it is important to further deepen the analysis of complex dependencies in this area of science and business practice.