New product development in foreign customer relationships: a study of international SMEs

Abstract This study identifies a gap in research concerning how small and medium-sized enterprises (SMEs) can benefit from pursuing locally (rather than globally) oriented internationalization strategies. Becoming overly dependent on one single foreign market could potentially reduce the inflow and diversity of new knowledge that can serve as input for new product development. This study discusses how this risk can be minimized. In this endeavour we create a theoretical model that investigates how the local sales concentration and relationship-specific commitment of SMEs relates to new product development. To do this we draw on the behavioural internationalization process framework. The theoretical model is tested on an effective sample of 188 Swedish SMEs. The results show that relationship-specific commitment mediates the effect of local sales concentration on new product development. The implication is that investments which enable collaboration in important business relationships are crucial requisites for keeping firms innovative and in pace with market fluctuations. The findings thus contribute to international business literature by showing that a local market scope of operations combined with a relationship orientation are beneficial for new product development in international SMEs.


Introduction
In international business research, there is a lack of consensus regarding whether firms should follow local market strategies (Buckley and Casson 1976;Hennart 1982;Rugman 2005) or develop their businesses on a global scope, sometimes already from the start Cavusgil 1996, 2004). Whereas previous studies about the benefits of locally oriented internationalization strategies have predominantly focused on large firms -namely multinational enterprises (MNEs) (e.g. De Martino, McHardy Reid, and Zygliodopoulos 2006;Gellynck, Vermeire, and Viaene 2007;Semlinger 2008) -internationalization orientations with more global scopes have been reported among small and medium-sized enterprises (SMEs) in the international entrepreneurship literature (Knight and Cavusgil 2004;Loane and Bell 2006). Whether -and if so, how -international SMEs can benefit from locally oriented internationalization strategies is understudied. Hence, scholars have called for more studies focusing on the effects of local contexts on the international development of SMEs (Drakopoulou Dodd and Hynes 2012;Kibler 2013). We attempt to address this gap in research by studying how local sales concentration (i.e. the local sales to a specific foreign market/total international sales-ratio) relates to SMEs' abilities to develop new products in that same market. A 'local market' is in this study regarded as a foreign country market in a specific region of the world, in which the investigated sellers and locally situated key customers can interact and exchange products and/or services with each other.
In this study, we apply new product development as the outcome variable because previous research suggests that it is critical but challenging for entrepreneurial firms, such as international SMEs, to devise suitable new products in a timely manner so as to serve needs and wants in markets (Knight and Cavusgil 2004;Yli-Renko and Janakiraman 2008). Our inquiry regarding the relationship between local sales concentration and new product development seems relevant because local sales concentration, potentially, could be an inhibiting factor for the product development and growth of SMEs. SMEs that earn a large part of their revenues in a single foreign market may eventually become entrenched in their business models and not responsive to change. In comparison to firms that are active in a larger set of markets, locally oriented firms will be exposed to less diversified feedback from foreign customers and business partners. In the worst case scenario, a lack of such input will create few incentives to instigate change. This could, ultimately, lead to a reduced level of new product development and overall stagnation. To contribute to a better understanding of the effects of the local sales concentration on the product development of SMEs, we study the mediating effect of local customer relationships. Customer relationships are regarded as central to the core value activities of firms (Gupta, Lehmann, and Stuart 2004;Yli-Renko, Sapienza, and Hay 2001) and studies have indicated that specific customer relationships have a positive impact on new product development (see: Rindfleisch and Moorman 2001;Yli-Renko and Janakiraman 2008). Consequently, the purpose of this study is to create a model that investigates how an SME's local sales concentration and relationship-specific commitment relate to new product development. We base the model on theoretical ideas from the behavioural internationalization process (IP) framework (e.g. Blomstermo et al. 2004;Vahlne 1977, 2009), arguing that this theoretical perspective has the potential to further the understanding of SMEs' new product development in local market contexts owing to its emphasis on the benefits related to foreign business relationships.
As noted by Banalieva and Dhanaraj (2013), the international business field must take a fresh look at the geographical scope of firms and its impact on performance. A feasible way of understanding the effect that the geographical concentration of sales operations has on performance outcomes is to adopt a business relationship approach. Studies have shown that business relationships can nullify perceived barriers of foreign markets and enable SMEs to explore local opportunities (Presutti, Boari, and Fratocchi 2016;Rovira Nordman and Tolstoy 2014). Thus, business relationships may be instrumental in allowing firms to stay in tune with market changes and provide input for product development. Yet little empirical work exist that examines the mediating effect of key customer relationships on new product development in particular market settings (Yli-Renko and Janakiraman 2008;Yli-Renko, Sapienza, and Hay 2001). By empirically demonstrating how SMEs' local sales concentration relates to their relationship-specific commitment as well as to new product development, we are able to contribute new insights into this relatively understudied area within the international business field. The findings presented in this study can specifically extend research focusing on how SMEs best can pursue entrepreneurial opportunities (Presutti, Boari, and Fratocchi 2016;Yli-Renko and Janakiraman 2008) by furthering the understanding of relationship-oriented new product development in a foreign market context.

Theoretical background
Behavioural internationalization process (IP) theory is one theoretical perspective that has had a major influence on the ongoing discussion in international business research on firms' proclivities to operate on a global or local scope. The original IP framework (Johanson and Vahlne 1977) builds on the idea that firms usually go abroad to close and familiar markets to start with and gradually extend foreign operations to more distant and unfamiliar countries. The driving-mechanism of the internationalization process is a firm's development of local market knowledge. In general, local market knowledge includes knowledge about the characteristics of a specific national market, its business climate, cultural patterns, individual customer firms and their personnel (Johanson and Vahlne 1977). This knowledge is critical because it enables a firm to identify concrete business opportunities in a market, such as an opportunity to collaborate with a local business partner. This knowledge also decreases a firm's uncertainty related to operating in a specific market, thus making it more inclined to commit resources to specific business opportunities that are identified (Barkema, Bell, and Pennings 1996;Blomstermo et al. 2004;Johanson and Vahlne 2009). Firms develop local market knowledge based on their activities in a market. By doing business in a specific country, managers learn how customers often act and react in different situations. This subtle understanding of a market, i.e. local market knowledge, cannot be replaced by general market information. It takes time to develop local market knowledge because it is associated with the specific situations and contexts under which it is developed.
Following the reasoning of the traditional IP framework, the internationalization process can be interpreted as an interplay between the development of local market knowledge and the commitment of resources to business opportunities. The original IP framework did not explicitly elaborate on the importance of business relationships in firms' internationalization (see: Johanson and Vahlne 1977). Based on empirical evidence (e.g. Erramilli and Rao 1990;Majkgård and Sharma 1998), later theoretical work suggested that firm internationalization is becoming less a matter of country or market specificity and more a matter of relationship specificity (Johanson and Vahlne 2009). Applying such a relational view on internationalization means that the problems and opportunities that a firm faces in foreign markets are related to specific business relationships instead of country specificities. It also implies that the concepts of local market knowledge and commitment do not primarily concern countries but also the potential and existing relationship partners firms interact with. Within these relationships new business opportunities are identified and acted on which requires resources to be committed. Such relationship-specific commitments are intertwined with the gradual development and growth of business relationships (Anderson and Weitz 1992). Scholars have elaborated on the importance of relationship-specific commitment for enabling firms to share knowledge effectively with each other, which ultimately can enhance the development of new business outcomes in particular market contexts (Ghauri, Hadjikhani, and Johanson 2005). Because relationship commitment reflects a bilateral orientation it is here conceptualized as the mutual desire of two parties to sustain an individual business relationship (Blankenburg Holm, Eriksson, and Johanson 1999). Following the reasoning of the revised IP framework, the important business relationships for an internationalizing firm are signified by close-knit, long-term and stable commitment strategies rather than being signified by arms-length and mostly transactional relationship building. The transactions that collaborating partners engage in are thus rather seen as instances of interactions inwrought in long-term relationships, which in their turn can promote both knowledge development and change (Johanson and Mattsson 1987).
The traditional IP framework has received criticism in the international entrepreneurship literature, where scholars have put forward that the framework cannot explain the phenomenon of small firms that already from inception run sales operations on a global scope (Oviatt and McDougall 1994). Proponents of the IP perspective have, as a response to this criticism, argued that most so-called born global firms actually do not run operations on a worldwide scale, but tend to be more locally oriented than globally oriented (Johanson and Vahlne 2009). Johanson and Vahlne (2009), furthermore, argue that the basic premises for developing international business have not changed. Both globally and locally oriented firms still need to develop local market knowledge by interacting with partners in specific foreign markets, and build trust and commitment in these relationships.
One key argument put forward by international entrepreneurship scholars is that a relationship-based approach is needed to understand how small, resource-constrained firms are able to tap into resources that enable them to expand and develop new business in foreign markets (Loane and Bell 2006). This study is based on the argument that recent developments of the IP framework, focusing on the benefits of relationship-specific commitment, have the potential to add to the predictive power regarding SMEs' business development in local market contexts. Even though some studies have empirically investigated the conceptual ideas presented in the revised IP framework, few empirical studies have investigated how SMEs' relationship-specific commitments relate to the development of new business outcomes in particular market settings. Moreover, the IP framework has not directly elaborated on business outcomes in terms of new product development. The basic mechanism described (i.e. the experiential learning-commitment interplay) has, however, been proven eligible to explain other business outcomes, such as the technological development of multinational firms (Johanson and Vahlne 2003;Zander 1999). In this study, we therefore build on the revised IP framework, but also extend the empirical scope of the framework by arguing that relationships-specific commitments are instrumental for learning about a local market and develop new products in a local market setting.

Hypotheses development
The development of new products in firms has previously been associated with sustained growth in the context of small, new firms (Zahra and Bogner 2000) as well as considered to capture innovation and R&D output (Katila and Ahuja 2002). Highlighting the importance of innovation and R&D output for new product development, we define new product development to be business outcomes in terms of both new products and technologies. This conceptualization builds on other studies (Yli-Renko and Janakiraman 2008; Yli-Renko, Sapienza, and Hay 2001), where new technologies in and of themselves are regarded as possible product outcomes. Because many new products can be based on new technical components (Van de Ven 1986), particularly in the context of small, technology-based firms, development of new products can be difficult to separate from the development of new technologies. We, therefore, argue that new products and new technologies are closely connected which makes it relevant to include both in our concept. Moreover, in line with the purpose of this study, we aim to capture those products and technologies developed as a result of an SME's activities in a specific local market.
Studies focusing on internationally entrepreneurial SMEs emphasize that the knowledge-based resources of these firms enable them to develop and introduce new products in niche markets on a broad international scale (Knight and Cavusgil 2004;Oviatt and McDougall 1994). The empirical studies at the same time describe that short product life cycles, and the lack of first-hand information about foreign market preferences and distribution channels pose challenges to firms in their attempts to meet the product demands of local markets quickly (Autio, Sapienza, and Almeida 2000;Crick and Spence 2005). Combined with a general lack of resources, such organizational challenges may hamper the ability of SMEs to introduce new products and achieve economies of scope globally. Following the reasoning of the IP framework, internationalizing firms often lack local market knowledge. Because this knowledge is experiential, firms which focus on one specific market will have easier access to this knowledge, than firms which are forced to dilute their foreign market presence over several markets. Moreover, a strong exposure to local market knowledge can facilitate the recognition and development of new business ideas that are specifically geared towards that particular market.
Looking at empirical studies on international SMEs and entrepreneurial firms, there is some evidence that a local scope of business is related to distinctive benefits, such as easy access to resources and knowledge from complementary resource bases and supportive industries (Andersson, Evers, and Griot 2013;Gellynck, Vermeire, and Viaene 2007;Johannisson, Ramirez-Pasillas, and Karlsson 2002;Van Geenhuizen 2008). The positive influence of tapping into resources from local industrial clusters has also been highlighted in economic geography studies, showing that firms located in areas where technological activities agglomerate (technology clusters) are more innovative than firms located elsewhere (Deeds, Decarolis, and Coombs 1999;Van Geenhuizen and Reyes-Gonzalez 2007). One reason for this is that the close proximity of organizations with similar interests (for example suppliers and customers) promotes a natural exchange of ideas between business partners which help knowledge spread (Rosenkopf and Almeida 2003;Von Hippel 1988) and allow firms to achieve R&D results fast and/or with few resources (Lecocq et al. 2012). Because of this the location of a firm is an indicator of its propensity to develop new knowledge which can lead to the development of new products. In their study of the biotechnology business in U.S.A., Deeds, Decarolis, and Coombs (1999) also show that a beneficial location near firms with similar interests has a significant positive impact on these firms' new product development.
Hence, these empirical findings as well as the theoretical reasoning of the IP framework suggest that international SMEs with a local scope of business may benefit from drawing on the resources and knowledge of local partners and industries in the specific foreign markets where they are active. Moreover, the access to locally applicable knowledge can spur new product development. We therefore argue that an SME's local sales concentration influences its ability to engage in new product development that is related to these markets.
H1: Local sales concentration is associated with new product development in the local market.
The revised IP framework elaborates on the importance of relationship-specific commitment. Based on this theoretical perspective, relationship-specific commitment can be defined as the closeness between a firm and its partner in an individual business relationship. In specific, such relationship-specific commitments can be manifested in those investments that are made to enhance interaction and mutual orientations in a business relationship (Jonsson and Lindbergh 2010). Relationship-specific commitments can in other words be seen to concern the gradual development of the relationships in which a firm is engaged (Anderson, Håkansson, and Johanson 1994). Marketing literature stipulates that the development of a business relationship is a process that requires time, resources (Dyer and Singh 1998) and responsiveness to partners to increase their mutual commitment. Business relationships develop when parties learn about each other interactively and thereby build trust and increase commitment (Anderson and Weitz 1992;Blankenburg Holm, Eriksson, and Johanson 1999;Morgan and Hunt 1994). Hence, mutual commitment can be equated with relationship-specific investments made by business partners (Blankenburg Holm, Eriksson, and Johanson 1999;Chetty and Eriksson 2002). When business partners learn from their common interaction, they acquire knowledge about for example the counterpart's willingness to adapt products or coordinate activities to strengthen the joint productivity (Johanson and Vahlne 2003). In specific, this coordination can involve the adaptation of production and administrative activities so as to bring about a better match between firms (Hallén, Johanson, and Seyed-Mohamed 1991). In other words, business partners commit to a relationship by adapting to each other (Chetty and Eriksson 2002;Hallén, Johanson, and Seyed-Mohamed 1991). Based on extant literature, we argue that relationship-specific commitments can be reflected by investment made in a specific business relationship, in terms of adaptations and time, which are characterized by mutuality.
Some studies have claimed that SMEs with a global scope of business develop an increased ability for interacting with business partners in the various markets in which they enter (Loane and Bell 2006;Sharma and Blomstermo 2003). At the same time, what these studies demonstrate is that small firms with a global scope of business tend to rely on indirect relationships in the foreign markets. These indirect relationships are typically characterized by limited intensity and duration of interaction, and therefore less cumbersome to maintain (Sharma and Blomstermo 2003). The results of a qualitative study by Chetty and Campbell-Hunt (2003) revealed that whereas globally oriented SMEs tended to rely on indirect representation in foreign markets, locally oriented SMEs favoured direct relationships with end customers in local markets. Moreover, by having a locally concentrated business, the firms had from their own experience learned about the needs of the particular markets, and thereby learned how the firms should adapt their products to serve the needs and problems of the customers (Chetty and Campbell-Hunt 2003). In other words, a local concentration of business helped the investigated firms to adapt products and services to be in line with the needs of the customers. Other studies focusing on locally oriented SMEs have given similar indications as those presented by Chetty and Campbell-Hunt (2003). Kontinen and Ojala (2011), for example, showed that once an SME actually enters a foreign market, the firm focuses more on developing customer relationships characterized by trust and commitment within that market and less on finding new international customers. Laursen, Masciarelli, and Prencipe (2012) argued that a firm that interacts mostly with local actors is more likely to continue to focus on developing the relationships in that particular market. What these studies indicate is that a local sales concentration by SMEs relates to the firms' proclivity to commit to local customer relationships by making investments in and adaptations to these relationships. Hence, we suggest the following hypothesis: H2: Local sales concentration is associated with relationship-specific commitment in the local market.
Applying a relational view on international business highlights the idea that differences between countries are of interest only if these differences have an impact on the interaction between firms. In other words, firms which already are engaged in existing business relationships are more likely to discover the need to adapt to local requirements. Strong commitments in the shape of adaptations to demanding customers are particularly important because such adaptations may results in superior products or production systems (Hallén, Johanson, and Seyed-Mohamed 1991). In a similar vein, a mutual commitment between interacting partners allows the participating firms to share and leverage their respective bodies of knowledge effectively, which can ultimately lead to the development of new business outcomes (Ghauri, Hadjikhani, and Johanson 2005). Hence, relationship-specific commitments allow firms to build on their respective bodies of knowledge, making it possible for them to discover and develop new business ideas in the relationship.
Even though business relationships generally are considered to be very important for the international development of firms, there is a danger involved in focusing too much on customers' needs because this may hinder their development efforts (Fischer and Reuber 2004;Yli-Renko and Janakiraman 2008) and hamper their innovativeness (Christensen and Bower 1996;Macdonald 1995). Research on both entrepreneurial firms and international SMEs have, however, emphasized that strong benefits can be entailed by being involved in business relationships (if firms can avoid getting too closely knit to specific business counterparts). In a study of 180 entrepreneurial, technology based firms, Yli-Renko, Autio, and Sapienza (2001) show that specific key relationships between young firms and their customers can form the basis of alliances or cooperative ventures that can lead to wealth-creating opportunities. The knowledge acquisition that is drawn from these relationships is positively associated with knowledge exploitation for competetive advantage which can be manifested in new product development. In a similar vein, studies focusing on international SMEs emphasize that firms' abilities to interact with foreign customers and partners and make adaptations to these are instrumental for accessing critical market knowledge (e.g. Sharma and Blomstermo 2003;Yli-Renko, Autio, and Tontti 2002). Hence, learning from key partners can fuel the internationalization of young firms (Bruneel, Yli-Renko, and Clarysse 2010).
Other studies of international SMEs and entrepreneurial firms specifically highlight the benefits of investing time and resources to commmit to business relationships, for new product development to occur. For example, firms that are involved in formal interfirm alliances frequently endeavour to use information and know-how for new product development (Rindfleisch and Moorman 2001). By trusting specific business partners and interacting with them, firms can obtain access to an extended knowledge and resource base (Presutti, Boari, and Fratocchi 2016;Yli-Renko and Janakiraman 2008). The closer a firm is to a specific customer, the less time is spent on monitoring and bargaining activities and the better they will understand each other's specialized systems, requirements and capabilities and will be able to tap into external knowledge more quickly (Dyer and Singh 1998). This kind of knowledge access can also provide firms with concrete knowledge and feedback on product improvements or new functional requirements (Presutti, Boari, and Fratocchi 2016;Smith, Collins, and Clark 2005). The benefits that close business relationships confer can thus generate new product outputs (Yli-Renko and Janakiraman 2008), improved effectiveness (Griffin and Hauser 1996) and speed (Rindfleisch and Moorman 2001) of new product development, and a higher number of patents and new products (Wuyts, Dutta, and Stremersch 2004).
Building on the empirical results discussed above and the theoretical reasoning of the IP framework, we argue that close customer relationships in a specific foreign market can enhance an SME's ability to leverage the knowledge accessed from these relationships. Moreover, when a firm invests time in the relationship the probability increases that it can create new products which resonate with the needs and requirements in a local market. Anchoring local sales concentration in relationship-specific commitments thus facilitates efficient knowledge acquisition from customers, spurring activities of new product development within a specific market. Consequently, we suggest that relationship-specific commitment can mediate the effect of local sales concentration on new product development.

H3:
The relationship between local sales concentration and new product development is mediated by relationship-specific commitment.

Research design
To investigate how a local sales concentration relates to relationship-specific commitment and new product development, we conducted a multiple regression analysis. We also conducted a mediation test where we investigated the mediating effect of relationship-specific commitment (between local sales concentration and new product development).

Selection of firms
For the purpose of this study, we focused on Swedish SMEs. Sweden is, for several reasons, an interesting setting to study SMEs. One reason is that 97% of firms in Sweden have 50 or fewer employees. Sweden is, moreover, a relatively small market, which puts pressure on SMEs to break into foreign markets soon after their establishment. Like most countries, Sweden is fragmented in terms of industrial structure. To maintain consistency in our investigation, we focused on firms located within the delimited geographical area of Mälardalen, Sweden (i.e. the area including and surrounding the capital city of Stockholm). This area was chosen because of its high industrial concentration and its geographical accessibility for the investigators.
An initial sample of Swedish SMEs was collected from Statistics Sweden's Business Register in 2003 on the basis of the following criteria: firms that (1) are active in foreign markets, with at least 10% of their turnovers as a result of export sales, and (2) fulfil the definition of an SME. 1 We derived a random sample of 233 international SMEs. These firms were considered to provide a representative group of Swedish internationalizing SMEs within the Mälardalen area. From this sample, 188 case firms participated in our study, a response rate of approximately 81%. The two major reasons given for declining to participate were a lack of time and a reluctance to release information. To control for differences between responding and non-responding firms with regard to industry, size, location and level of internationalization, we used secondary data collected from Statistics Sweden's Business Register. This analysis revealed no significant differences between the groups; therefore, non-response bias is unlikely to be a problematic issue when interpreting the findings of the study.

Data collection
In this study, we combine objective data from the import/export register provided by Statistics Sweden with perceptual data collected from a questionnaire. The questionnaire contained questions revolving around a specific foreign business relationship (with a key customer) and its supporting network in the specific local market. The questions were limited to a specific market so as to capture the local aspects of firms' international activities as well as to provide consistency so as to increase reliability. Each respondent was asked to select a foreign business relationship that met three conditions: it was considered important to the firm, it was ongoing, and it had resulted in realized sales transactions.
The investigators visited the informants personally on-site, thereby increasing the chance that the correct individuals were answering the surveys and likely reducing the number of missing values. The persons answering the questionnaires were the individuals who were considered key informants in the firms -that is, those who made decisions related to foreign market operations (most often, the chief executive officer [CEO] or the marketing manager). When the respondents had answered the questionnaires, the investigators also conducted semi-structured interviews during which the respondents could speak more freely about the selected business relationships.

Control variables
We used four variables as control variables for the multiple regression models: (1) size (number of employees), (2) time in international markets (years elapsed since entry), (3) contractual tie to the foreign customer and (4) cultural distance. Size is measured by the number of permanent employees working in the firm. Time in international markets represents the amount of time (in years) that has elapsed since the firm made its first foreign market entry. The contractual tie to the foreign customer is a dummy variable that denotes whether the customer is an agent or distributor, which would imply a formalized type of business exchange. Cultural distance was operationalized in line with Hofstede's (1980) cultural dimensions, which were transformed into a composite index (see Appendix 1) based on the cultural distance of every country of entry that was represented in our sample in relation to Sweden. This index was created based on the formula developed by Kogut and Singh (1988), which corrects deviations for differences in variances and then averages them arithmetically: where CD j represents the cultural distance from Sweden (the base country, denoted by s) to country j, I ij is the index for cultural dimension I of country j and V i is the variance index of dimension I.

Independent variable
Regional sales concentration and similar constructs related to local international sales have been applied in previous studies (e.g. Rugman 2005). Rugman (2005), who arguably was one of the most influential proponents of the regional perspective, analysed the fraction of international sales linked to triad markets (Europe, North America and Japan) to determine the level of regional orientation of MNEs. Such a measure may work for larger firms, but we argue that it is too blunt of a measure to be applicable to SMEs. To determine the local sales concentration of SMEs in our sample, we considered the frequently used Herfindahl-Hirschman Index, which would provide an estimate of the average share of sales to a foreign market within the firm's portfolio, weighted by the relative sales contribution of each market. We were, however, unable to receive exact sales information of each market in which the firm operated. The next best alternative was to operationalize the local sales concentration by focusing on one key foreign market. This not only would be a reliable method to receive accurate data on the ratio of foreign market sales relative to international sales but also would provide consistency to our study as all variables revolve around one particular foreign market venture. Hence, we operationalized the variable as the fraction of sales to the home market of the foreign key customer in relation to total international sales. This endeavour is akin to that of Cooper and Kleinschmidt (1985), who captured a regional international sales strategy by relating the amount of foreign sales to a neighbouring country to the total amount of international sales. We believe, however, that focusing on a key market is more salient than focusing on a neighbouring market because the internationalizing behaviour of SMEs has been observed to be not as shaped by geographical or psychic distance as used to be the case (Ojala 2015;Rovira Nordman and Tolstoy 2014). Furthermore, we investigate whether a local sales concentration will affect business development in SMEs along the following dimensions (see Appendix 2).

Mediating variable
Relationship-specific commitment can be manifested by investments which enhance interaction and mutual orientation in a business relationship (Jonsson and Lindbergh 2010). These investments can be of different kinds but in this study we use three items to reflect relationship-specific commitment at the foreign market level. The first two items are derived from a study investigating relationship specific commitment by Jonsson and Lindbergh (2010) and concern the extent to which the respondents have invested in specific foreign market relationships in terms of adaptations and time. The third, complementary, item measures whether the respondents perceive the investments made in the relationship to be characterized by mutuality (i.e. that the respondents' perceive their selected customers to participate in making mutual investments in the relationships under study). We consider this item to be important because commitment is not one-sided but involves a mutual orientation (Blankenburg Holm, Eriksson, and Johanson 1999).

Dependent variable
Building on the findings of Yli-Renko and Janakiraman (2008), we focus solely on the new product development portion of the innovation process (thus disregarding the other parts of the innovation process, from idea generation to adoption by customers). In accordance with Yli-Renko and Janakiraman's (2008) research, we can then focus on the specific question of 'how specific customers affect new product development' . Because technology in and of itself can be regarded as a product (to be purchased and sold) in many high-tech firms, the new product development construct is measured by the extent to which the firm has used the focal business relationship to create (1) new products and/or (2) technologies within the same market. These items are derived from a recent study investigating innovation outcomes by Rovira Nordman and Tolstoy (2016). Table 1 presents the bivariate correlations of all the variables in the study. As indicated in Table 1, some variables are significantly related, although not generally highly correlated.

Results from the multiple regression analysis
To investigate the discrete effects of local sales concentration on relationship-specific commitment and new product development we used multiple regression as the technique of analysis (see Table 2). We ran regressions in two models separated on the basis of different dependent variables. Each model was linked to a baseline model to check for alternative effects and thus avoid omitted variable bias. The statistics showed a significant positive effect of size on relationship-specific commitment, indicating that larger firms are relatively more inclined to invest more heavily in foreign business relationships. We can also discern a negative effect of time in international markets on new product development, which suggests that firms that have spent a relatively longer time abroad (or are just older as this variable is highly correlated with age) are likely to stagnate in this respect eventually. The main effects of the study pertaining to hypotheses 1 and 2 are both supported by strong significant beta coefficients, thus showing that local sales concentration has discrete effects on both relationship-specific commitment and new product development. Overall, the models seemed to provide an acceptable, although not ideal, level of accuracy and explanatory power according to fit measures. The high level of unexplained variance of the models evokes suspicion that the models are not optimally specified. Even though hypotheses 1 and 2 are confirmed, we believe that the validity of these results can be enhanced if we check for the interrelatedness between variables as stipulated in hypothesis 3. We claim that the explanatory power of the model can be increased if the relationship-specific commitment variable is specified as a mediator that links the other variables of the study to each other.

Results from the mediation analysis
To test the significance of hypothesis 3, we delineate a causal path in a nested model where relationship-specific commitment is set to serve as a mediating variable between local sales concentration and new product development. To evaluate whether a mediation effect exists, we followed Baron and Kenny's (1986) three-step process by (1) testing whether the independent variable (X) is a significant predictor of the mediator (M), (2) testing whether M is a significant predictor of the variable (Y) and (3) assessing whether the total effect on Y of a model that includes both X and M is larger than the discrete direct effect of a model that contains only X. To strengthen the reliability of our analysis, we also conducted a Sobel test to check the significance of the mediation effect. Our mediation analysis (see Table 3) supports these three steps and thus substantiates hypothesis 3. The direct effect on the variable local sales concentration is not insignificant in the third step of the model, which indicates a partial mediation effect. The Sobel test shows a significant z-value, thus lending further support to the validity of the effect. The mediating effect can be explained by the argument that close relationships characterized by commitment support interaction and provide learning advantages in the local markets in which firms operate. Hence, as partners become more involved in relationships, the probability increases that they can create new products that resonates with the needs and requirements in the local market setting.

Discussion
Whether -and if so, how -international SMEs can benefit from locally oriented internationalization strategies still remains an area that lacks empirical substantiation within the international business literature. In this study, we address this gap in research by studying how Table 3. Baron and Kenny (1986) mediation test. * p < 0.05 (two-tailed); ** p < 0.01.
Step 1 2.544 ** local sales concentration relates to SMEs' abilities to develop new products in that same market. Becoming overly dependent on one single local market could potentially reduce the inflow and diversity of new knowledge that can serve as input for new product development. In the worst case, a lack of such input will create few incentives to instigate change and lead to a reduced level of new product development. Hence, it is relevant to consider the risk that a local sales concentration could inhibit the product development and growth of SMEs. The findings of this study however show that to minimize this risk and get the most out of a strategy that is local in scope, SMEs are dependent on close customer relationships that can stimulate market learning. In this capacity, business relationships make firms more closely connected to specific foreign markets and provide impetuses which enable firms to stay relevant to customers. The findings of this study therefore add new insights into the ongoing discussion within the international business field about the effects of local/global strategies and offer implications to the literature about how SMEs best can pursue entrepreneurial opportunities.

Theoretical implications
A strategy involving a high level of local sales concentration could serve a company well -at least in the short run. An advantage of local sales concentration is that it can spur operational efficiency. However, an overemphasis on efficiency in managing customers could entail that business exchange becomes transactional rather than relational, leading to negative outcomes in the long run. For most firms it is crucial for sustained growth to continually adapt to market changes and develop products that resonate with customers' changing needs. For this purpose they need to stay in touch with the market by acquiring pertinent knowledge. Such local market knowledge will, arguably, be easier to attain in close rather than in arms-length relationships. Hence, the most original and important result of our study is that we can empirically verify that relationship-specific commitment is an important mediating variable that positively affects a firm's ability to leverage the inherent capacity of a local sales strategy/orientation. The findings are thus strongly aligned with the revised IP framework, by demonstrating that relationship commitment is a key mechanism for new product development in a foreign market. Our findings imply that investments which enable collaboration in important business relationships are crucial requisites for keeping firms innovative and in pace with market fluctuations. In relation to the ongoing discussion in international business research regarding the international sales concentration of firms, our findings suggest there is much to gain from using theoretical perspectives that take into account the conducive nature of business relationships. A business relationship can facilitate learning which enable firms to stay up to speed with the changing needs and requirements in specific foreign markets. Such relationships may provide insidership positions in foreign markets (see Johanson and Vahlne 2009), thus promoting the acquisition of local market knowledge. This knowledge can in its turn be used for product development that aligns with the needs of customers. One of the key messages of this study is that locally oriented strategies should not only be built on the principle of transactional efficiency. A transactional approach combined with high local sales concentration may eventually detach firms from their key markets and stifle new product development. Scholars in the international business field have suggested that the field must take a fresh look at the geographical scope of firms and its impact on performance (Banalieva and Dhanaraj 2013). Our findings shed new light on this issue by underscoring the usefulness of a relationship-based approach for assessing how SMEs can succeed with locally-oriented strategies.
The results of our model also contribute to research investigating how SMEs best can pursue entrepreneurial opportunities by adopting locally relationship-oriented strategies to generate international growth (e.g. Presutti, Boari, and Fratocchi 2016;Rovira Nordman and Tolstoy 2014) and new product development (Yli-Renko and Janakiraman 2008). Although various qualitative case-based studies in this field have indicated a direct relationship between locally oriented strategies and close business relationships (e.g. Chetty and Campbell-Hunt 2003;Kontinen and Ojala 2011), our study on international SMEs offers a quantitative validation of how these theoretical constructs are linked to each other. By developing our understanding about relationship-oriented new product development in a local market context, our results also extend the literature on the impact of influential customers (e.g. Yli-Renko, Sapienza, and Hay 2001) and whole customer portfolios (Yli-Renko and Janakiraman 2008). In relation to the ongoing discussion within this literature (Christensen and Bower 1996;Fischer and Reuber 2004;Yli-Renko and Janakiraman 2008), our study emphasizes and elaborates on the positive impact from specific customer relationships on new product development in a specific local market context.

Limitations and further research
Previous research on MNEs has shown that a majority of the world's largest firms actually operate using a local-based strategy (Rugman 2005). Even though locally oriented internationalization strategies have been given less attention in research focusing on SMEs, some studies on British (Beleska-Spasova and Glaister 2010), Japanese (Delios and Beamish 2005) and Costa Rican (Lopez, Kundu, and Ciravegna 2009) firms have revealed that these firms are also predominantly local in their geographic spread. One limitation of this study is that it did not delve into the factors underlying firms' choice of either a locally or a more globally oriented perspective. We argue, however, that more research is required to test our results further and to provide more detailed insights into the mechanisms and circumstances that influence locally based strategies. In the same vein, our models should by no means be viewed as exclusive as they are devised on the principle of parsimony. We humbly encourage researchers to investigate alternative variables that local market orientations of SMEs could affect, such as long-term performance, organizational structures and establishment modes.
A second limitation of this study relates to certain aspects of the sample: the sample was taken from one country, the sample included only internationally active small and medium-sized firms, and the respondents were questioned at only one point in time. These aspects of the sample limit the generalizability of the results to other contexts. Future studies should examine whether our results are valid for larger firms and are applicable to firms with a domestic market orientation (i.e. whether a local market orientation affects the results in a domestic market setting).
A third limitation of this study is the use of new product development as our dependent variable. Even though this is a commonly used measure, it is only one aspect of a firm's innovative output. An equally important aspect is the commercialization phase of a new product in local markets. Some aspects relating to the commercialization phase of new product development, such as the general performance and growth of firms in specific local markets, could be investigated in future research.

Note
1. An SME does not exceed 250 employees. SME firms are firms with fewer than 50 employees, and medium-sized firms are firms with 50-250 employees (OECD 2005