Summary highlights

Contributions: The study contributes to the growing body of knowledge on international entrepreneurship by matching three theoretical perspectives with empirical data to explain the long-term trajectory of high-tech INVs (covering at least 20 years from inception) from an emerging economy—Brazil.

Research questions/purpose: The study departs from the following research question: To what extent does the internationalization process of high-tech firms fit the predictions of the Uppsala internationalization process model, the network theory, and the international entrepreneurship perspective?

Methodology: The research method adopted was the multiple case study using a longitudinal approach. A total of 28 interviews were conducted with the managerial team of seven firms during a substantial period of time. In addition, several secondary sources were used allowing for triangulation between different data sources and between interviews conducted in different moments of time.

Results/findings: The results show that no single theoretical perspective examined in the study (the Uppsala model, network theory, or the international entrepreneurship approach) seems to be able to explain the international trajectory of the INVs over time, but all of them may contribute to explain part of a firm’s long-term international trajectory.

Limitations: Due to the limitations arising from the use of the case method, it is not possible to state the extent to which each of the theoretical perspectives adopted applies to the population from which the cases were drawn. In addition, since all firms were from Brazil, the country of origin may have, at least to some extent, influenced the results.

Practical implications: Entrepreneurs should be aware of the different options available for the firm’s internationalization process and not be limited by any path deemed superior to others.

Public policy recommendations: As for policy makers, this study shows that entrepreneurial firms may follow different paths in their internationalization process, and there is no single recipe for success. Therefore, instruments designed to promote exports or other forms of internationalization should take into consideration the diversity of possible trajectories that these firms may follow.

Future research directions: Further research is needed, both qualitative and quantitative, to explore the long-term development of INVs, taking into consideration their different contexts of origin.

Introduction

The phenomenon of international new ventures (INVs) has received substantial attention in the International Business (IB) literature since the pioneering work of Oviatt and McDougall (1994), Knight and Cavusgil (1996), and Madsen and Servais (1997). However, in spite of the already large number of studies on the international trajectory of new ventures during the first years after inception, only recently scholars have paid attention at how these firms develop over time (Romanello and Chiarvesio 2017). In fact, researchers have argued that this line of inquiry is still “embryonic” (Knight and Liesch 2016) and insufficiently understood (Coviello 2015; Jones et al. 2011; Zettinig and Benson-Rea 2008), in spite of its importance to the field. The present study addresses this research gap and examines the long-term international trajectory of INVs using a longitudinal research design, in response to a call by several scholars (e.g., Coombs et al. 2009; Hagen and Zucchella 2014; Keupp and Gassmann 2009; Vahlne and Johanson 2014; Welch and Paavilainen-Mäntymäki 2014).

Following the recommendations in the extant literature for the use of a multi-theoretical approach (e.g., Cavusgil and Knight 2015; Coviello 2015; Peiris et al. 2012), the study uses three different theoretical perspectives to examine the trajectory of seven Brazilian software firms over time: the Uppsala internationalization process model, network theory, and the international entrepreneurship perspective. These three theoretical perspectives were identified as the most cited ones in empirical studies by Welch and Paavilainen-Mäntymäki (2014), in their comprehensive review of the literature on the internationalization process of the firm. As to the so-called economic theories of the multinational firm, and management theories applied to internationalization (such as the resource-based view and contingency theory), they do not present “an explicit time orientation” (Welch and Paavilainen-Mäntymäki 2014, p. 8).

It has been argued that INVs differ from traditional firms in their internationalization trajectory, and that gradual internationalization theories could not explain the international expansion of this new breed of firms (e.g., Oviatt and McDougall 1994; Gabrielsson and Gabrielsson 2013). However, there is insufficient evidence as to whether the new patterns, characteristic of INVs’ entry phase, remain in later phases of their lifecycle. For example, Hashai and Almor (2004) noted that Israeli BGs followed a gradual internationalization pattern later in their trajectory.

The study also responds to a call by scholars to study INVs from emerging markets, a subject that has received limited attention from scholars (Glaister et al. 2014; Kiss et al. 2012; Peiris et al. 2012; Zander et al. 2015), as well as from a large domestic market (Cavusgil and Knight 2015). Kazlauskaitė et al. (2015) indicate that “it is necessary to study to what extent other logics, such as the process theory of internationalisation, international entrepreneurship, network theory, institutional theory, etc.” contribute to explain the phenomenon of INVs in emerging countries.

The study thus explores the following research question: To what extent does the internationalization process of high-tech firms fit the predictions of the three theoretical perspectives selected? Propositions emanating from the three theoretical perspectives are compared with the empirical data regarding pre-internationalization behavior, motives to internationalize, attitude towards risk, speed of internationalization, choice of markets and entry modes, relationship between growth and internationalization, role of networks, and the engine of internationalization along the firm’s internationalization process.

We contribute to the extant literature in three ways: first, it provides a fresh view of the application of three IB theoretical perspectives to examine the long-term trajectory of high-tech INVs; second, it adopts a longitudinal view of this trajectory; third, it studies the creation and development of INVs originating from an emerging country.

Theoretical perspectives and propositions

Traditional theories aiming at explaining the internationalization process of firms, such as the original Uppsala internationalization process model (U-Model) (Johanson and Vahlne 1977; Johanson and Wiedersheim-Paul 1975) and other stage models (e.g., Bilkey and Tesar 1977; Cavusgil 1980, 1982), are based on empirical observations during the 1960s and 1970s, when such processes were slower and more typical of larger firms. In the late 1980s, another contribution to the network theory emphasized the role of networks in the internationalization process (Johanson and Mattsson 1988; Coviello and Munro 1995, 1997). Network theory departed from the observation that smaller firms are not sole actors in the international arena, but that they often internationalize with others (Bonaccorsi 1992), or within their networks (Chetty and Holm 2000). Subsequently, the revised Uppsala model (Johanson and Vahlne 2009) brought the concept of networking to the core of the internationalization process. In the 1990s, with the increased entry of new ventures in international markets as a result of globalization, two different groups of researchers studied and theorized upon the new phenomenon, called INV by one group (e.g., Oviatt and McDougall 1994) and born global (BG) by the other (e.g., Knight and Cavusgil 1996, 2004). These two research lines later merged in the field of international entrepreneurship (IE), and the born global firm was recognized as one case of a more complex and diverse phenomenon (Coviello 2015; Jones et al. 2011; Madsen 2013).

These three theoretical perspectives compete to explain the internationalization process of smaller firms, thus suggesting the need to confront them with empirical evidence. We now provide theoretical support to the propositions emanating from the three theoretical perspectives: the U-Model (noted as UM), network theory (NT), and the international entrepreneurship perspective (IE). Appendix 1 lists the propositions extracted from the literature review and Table 1 summarizes these propositions.

Table 1 Summary of propositions
Full size table

The U-Model

The U-Model has been developed by a group of scholars and is based on case studies of Swedish multinationals (Johanson and Vahlne 1977; Johanson and Wiedersheim-Paul 1975). Firms start their activities in the domestic market (UM-1) and only later in their trajectory enter international markets (UM-8) (Johanson and Wiedersheim-Paul 1975). Initiation can be the result of previous international experience of members of the management team (Johanson and Wiedersheim-Paul 1975) or of opportunities coming from outside the firm (UM-1) (Johanson and Vahlne 1977). Petersen and Pedersen (1997, p. 125) argue that the U-Model implicitly assumes that “the dominant, or sole, motive of internationalisation is market seeking” (UM-2). Another assumption of the model is that decision-makers are permanently “striving to keep risks low” (UM-3) (Johanson and Vahlne 1977, p. 29).

The internationalization process is seen as a series of slow, cautious, incremental moves (UM-4) comprising an “establishment chain” characterized by two decisions: choice of markets and choice of entry modes (Johanson and Wiedersheim-Paul 1975; Björkman and Forsgren 1997). The choice of markets follows the logic of psychic distance; firms first move to markets with lower psychic distance and only later in their international development enter psychically-distant markets (UM-5). Firms initially adopt low-commitment entry modes (UM-6) and later move to high-commitment ones (UM-7). The original model does not make predictions about the use of networks. In fact, Johanson and Vahlne (1990, p. 19) state that the incorporation of the network view in the U-Model required “to make the concepts […] multilateral rather than unilateral as in the original model” (UM-9).

An important characteristic of the U-Model is the mechanism by which the internationalization process feeds itself: the more the knowledge, the higher the commitment. Knowledge, however, “is acquired primarily through experience” (Björkman and Forsgren 1997, p. 18). Learning is the basic mechanism of the process: learning reduces uncertainty and risk, thus increasing the firm’s ability to explore foreign opportunities. The process feeds into itself (UM-10); it is self-sustaining inasmuch as the firm continues to gain knowledge and the experience is satisfactory (Johanson and Vahlne 2009).

Network theory

The network approach applied to international business proposes that “the internationalization of the firm means that the firm establishes and develops positions in relation to counterparts in foreign networks” (Johanson and Mattsson 1988, p. 309). Early studies (Coviello and Munro 1995, 1997) provided evidence that the internationalization of smaller firms is driven by networks, and subsequent research has given “substantial empirical support for this approach” (Loane and Bell 2009, p. 92).

Networks are seen both as initiating forces and as facilitators of internationalization (Chetty and Holm 2000; Ellis 2000; Lindell and Karazoglu 1997). Networks are often present from the outset of a firm’s internationalization (Johanson and Mattsson 1988; Johanson and Vahlne 1990, 2003, 2009), as in the following situations (NT-1): when the leader of the network or another focal firm internationalizes (Johanson and Vahlne 1990); when internationalization results from previous social, business, or personal ties (Chetty and Holm 2000; Eriksson and Chetty 2003); when firms imitate or follow other firms (Forsgren 2002; Johanson and Vahlne 1990); when firms organize together to export (Ghauri et al. 2003; Welch et al. 1998), or even when the firm actively searches for foreign partners (Johanson and Vahlne 2006). Firms may utilize a foreign network to acquire market knowledge (Chetty and Holm 2000; Forsgren 2002; Welch et al. 1998); to reach new opportunities (Johanson and Vahlne 2006); to gain legitimacy (Tang 2009); to overcome obstacles (Ghauri et al. 2003), or to obtain access to foreign capital (Chetty and Holm 2000), among other motives (NT-2). Relationships with partners from a foreign network may reduce risk and uncertainty (Johanson and Mattsson 1987) (NT-3), thus offering “a sort of safety net” (Johanson and Vahlne 2011, p. 489).

Networks are important throughout the internationalization process, influencing its pace (NT-4). The speed of internationalization depends on the amount of time required to develop the network relationships. Networks impact, and even shape, certain decisions. Choices of markets and entry modes are contingent on available alternatives provided by the network (NT-6) (e.g., Coviello and Munro 1995, 1997; Torkkeli et al. 2012). Firms may choose “a foreign market in which the partner has a strong position” (Johanson and Vahlne 2009, p. 1425) or even a new location based on an assessment of the easiness of establishing new relationships (NT-5). On the downside, networks often impose long-term restrictions to the firm’s international trajectory (Mattson 1989) (NT-7).

Johanson and Vahlne (2009) argue that internationalization should not be seen as independent actions of the individual firm in a foreign market any more, but that “the internationalization process is pursued within a network” (p. 1424) (NT-9), “as a position-building process” (p. 1415). Entry into a foreign market using a network is a long-term gradual investment in relationships and “a time-demanding process,” involving spending “considerable managerial resources” (Johanson and Vahlne 2011, p. 486). As relationships evolve, trust, commitment, and interdependence increase (NT-10).

The IE approach

The extant literature on international entrepreneurship has been described as “fragmented” and lacking “common theoretical integration” (Keupp and Gassmann 2009, p. 601) or as “lacking an unified direction” (Coombs et al. 2009, p. 24). In addition, because contributions emanate from two originally distinct research lines that depart from somewhat different assumptions and theoretical frameworks, the research results are often difficult to integrate. In spite of this situation, there is also substantial convergence and complementariness between the BG and INV studies; the differences are reconcilable. In fact, Jones et al. (2011) and Peiris et al. (2012) literature reviews provide an integrative view of IE research, combining INV and BG studies.

Internationalization “is preceded by opportunity recognition” (IE-1) (Peiris et al. 2012, p. 294). The main motives for internationalization are the search for value creation and growth (IE-2, IE-8) (Autio et al. 2005; Fernhaber and McDougall 2005; Oviatt and McDougall 1997). The entrepreneur has a crucial role in the decision to internationalize the new venture (Andersson 2000; McDougall et al. 1994). Entrepreneurs who create INVs/BGs accept the risks of an early international expansion or “tolerate initial failure” (IE-3) (Jones et al. 2011, p. 639), more than their counterparts in traditionally internationalized firms (Harveston et al. 2000; Knight and Cavusgil 1996). Most studies remark the importance of a global mind-set or an international orientation for BGs/INVs’ internationalization (IE-10) (e.g., Bloodgood et al. 1996; Moen 2002; Oviatt and McDougall 1995), which is considered “the most distinguishing feature of an INV” (Aspelund et al. 2007, p. 1437).

Internationalization evolves rapidly and is not limited by firm resources, or “asset parsimony” (Cavusgil and Knight 2015, p. 5), since knowledge (INVs’ most important resource) is mobile, intangible, and combined with partners’ assets in foreign markets (IE-4) (Peiris et al. 2012). Because of their limited resources, INVs tend to adopt alternative structures of governance using low-cost, low-investment entry modes (IE-6) (Oviatt and McDougall 1994), and, quite often, “favor exporting as their primary entry mode” (Cavusgil and Knight 2015, p. 4), but often use multiple entry modes in different markets (Freeman et al. 2006; Oviatt and McDougall 1994). As to market selection, there is some debate on how BGs/INVs behave. INVs/BGs tend to use focus strategies to serve international market niches (Knight and Cavusgil 1996; Madsen and Servais 1997; Rialp et al. 2005). Aspelund, Madsen and Moen (2007, p. 1433) claim that there is reasonable agreement that these firms “will normally follow a niche focus strategy” (IE-5). There is also some evidence that psychic distance is not a deterrent to entering foreign markets (Freeman et al. 2015; Gripsrud et al. 2015), a finding that may be explained by the desire to serve global niches, as well as by the lack of experiential knowledge due to rapid internationalization (Pellegrino and McNaughton 2015). Trudgen and Freeman (2014) showed that psychically distant markets were often accessed using a pre-existing network.

Networks play a key role for these new ventures to enter and operate in foreign markets (Freeman et al. 2006; Rasmussen et al. 2001). There is some discussion as to whether INV/BG managers and entrepreneurs have a proactive (Coviello 2006; Schwens and Kabst 2009) or reactive (Crick and Spence 2005; Vasilchenko and Morrish 2011) approach to the use of networks in the internationalization process. It is possible that the first situation occurs mainly in the initial steps of internationalization (Sepulveda and Gabrielsson 2013). In any case, the literature shows that entrepreneurs use, develop, and strengthen their networks to support the internationalization process (IE-9).

Previous studies on the long-term trajectory of INVs

Only recently scholars turned their attention to the later development of INVs or of BGs. One of the earlier studies (Hashai and Almor 2004) to examine the subsequent international trajectory found that Israeli BGs in high-tech industries showed gradual internationalization patterns after their initial market entry. Studies that focused on the survival rates of INVs and BGs found that only a limited number of these firms achieve success (survival, increased sales, high profits, or increased market value) over time. Almor, Tarba and Margalit (2014, p.437) concluded that the performance of mature BGs was “disappointing,” and Hagen and Zucchella (2014) found evidence that most BGs tend to stagnate and decline. However, their study, as well as Mudambi and Zahra’s (2007), detected no differences in their probability of failure compared to established firms, while Sleuwaegen and Onkelinx (2013) found that global start-ups were more prone to failure than other start-ups or gradual exporters. Romanello and Chiarvesio (2017) suggest that there is a “turning point,” from entry to post-entry stage, considered by the authors crucial for the survival of the firm.

Other studies identified several factors (or a combination of them) that appear to be associated with long-term survival and success (Efrat and Shoham 2012; Hagen and Zucchella 2014; Mudambi and Zahra 2007). Ughetto (2016) finds, in a sample of BGs, a positive impact on growth of the entrepreneurial experience and the receipt of venture capital. Gabrielsson et al. (2014), examining four cases of Finnish INVs, find that two major factors seem to positively affect INVs throughout their long-term trajectory: international learning and networking. However, motivation to internationalize, propensity to innovate, attitude towards risk, market orientation, and proactiveness seem to have a positive effect only in the early phases.

A few studies identify, or recommend, strategies and practices associated with later stages of an INV/BG trajectory. The importance of building dynamic capabilities has been emphasized as critical to the successful long-term trajectory of INVs (Al-Aali and Teece 2014; Gabrielsson and Gabrielsson 2013). Oxtorp (2014) argues that mature INVs need to develop new capabilities, and change structure, processes, and rules. Rialp-Criado et al. (2010) suggest that they tend to adopt a more structured planning process in more advanced stages of their trajectory. Changes in scope—customer scope, country scope, or product scope— may also open new paths to long-term growth (Almor 2013).

The present study contributes to this growing body of knowledge by matching the three theoretical perspectives with empirical data to explain the long-term trajectory of high-tech INVs (covering at least 20 years from inception) from an emerging economy, Brazil.

Methodology

The research method adopted is the case study, considered “a key research strategy in the field” of IB (Piekkari and Welch 2011, p. 3). Aharoni (2011, p. 41) advocates that the method “enables the researchers to gain a holistic view of a certain phenomenon” and is particularly “well suited to produce … context-dependent knowledge” (p. 49). The study follows a longitudinal approach to examine how the internationalization process of INVs unfolds over time. The study uses a “multiple shot design,” that is “when research time and case time coincide but only at intervals” (Blazejewski 2011, p. 251). Accordingly, data was collected in two or more different moments in time for each unit of analysis, with the purpose of identifying changes that occurred as their international trajectory evolved (Hassett and Paavilainen-Mäntymäki 2013). Because the study looked at changes over time in the firms’ internationalization processes, the main criteria for company selection were (i) the availability of transcripts of interviews conducted before 2015 and (ii) company foundation in the 1990s, leading to the choice of seven software developers.

The study relied on personal interviews and secondary data. A total of 28 interviews were conducted with the managerial team of the seven firms during a substantial period of time. Although some of the initial interviews were not conducted with the purpose of developing a longitudinal study, this decision was taken in 2008, when six of the seven companies were interviewed. The earlier interviews were conducted by one of the authors of the present study, or by research assistants under the guidance of one or more authors. Additional interviews were carried out with the seven companies in 2015 and 2016. At the end, the number of interviews per firm varied from two to six, covering a research period (between the first and the last interview) of a minimum of 8 to a maximum of 15 years. Nevertheless, the total time interval examined in the study—the case time—covered the period from the company’s inception to the present. All interviewees, with the exception of one, were part of the founding team (Table 2).

Table 2 Interviews
Full size table

Data were both real-time and retrospective (Hassett and Paavilainen-Mäntymäki 2013), a practice often used in longitudinal case studies (Blazejewski 2011). In each interview, entrepreneurs were asked questions that required narrating events that happened between the specific interview and the previous one, as well as to further explain events previously reported. By revisiting events that were already reported in previous interviews, it was possible to recover missing elements, as well as to check for consistency, avoiding some of the common traps associated with retrospective accounts (Soulsby and Clark 2011). In addition, several secondary sources were used, including company websites, articles published in business magazines and newspapers, information from the internet, presentations by company managers, and company documents. This process allowed for triangulation between different data sources and between interviews carried out in different moments of time and, in three of the firms, with more than one member of the managerial team.

Data analysis was conducted following a variant of the standard protocol (Ghauri 2004). First, we prepared descriptive reports for each case, using all the data available from all sources. The intention was to develop a comprehensive narrative of each company’s internationalization history, and to prepare detailed timelines, “to reveal the sequential order of events” (Soulsby and Clark 2011, p. 277). Tables and graphs were used to help organize the data. Second, the data were coded using the ten conceptual categories and related sub-categories previously identified. Using pattern-matching logic, two members of the research team analyzed each case. The pattern-matching strategy, as suggested by Yin (1989, p. 109), “compares an empirically based pattern with a predicted one (or with several alternative predictions). If the patterns coincide, the results can help a case study to strengthen its internal validity.” In this study, we compared the patterns discernible in each case study with the rival theoretical perspectives (or predictions). The purpose here was to determine which theoretical perspective seemed to better explain how the internationalization of each company unfolded over time. The interpretations given by each researcher were then compared and debated, until reaching consensus. Third, a comparative analysis between cases (cross-case analysis) permitted drawing conclusions concerning the set of cases analyzed vis-à-vis the theoretical debate and previous empirical evidence on the long-term trajectory of INVs.

Results

Table 3 presents general company information and Appendix 2 a summary of company histories. The firms took from 1 to 6 years between their inception and their first international initiative. Between 2007 and 2015, only two companies showed substantial growth in terms of number of employees (InBox and Lumina), one company reduced its size (F-Expert), and the others remained about the same size. All companies survived until the first months of 2016. As to the intensity of internationalization measured by the percentage of turnover originating from international markets, three companies de-internationalized, three substantially increased their international turnover, and one slightly reduced this percentage. It should be noted, however, that until 2015 these firms faced a continuous trend of appreciation of the Brazilian currency, making it hard to export from Brazil.

Table 3 Company information
Full size table

Table 4 presents the triggers for the initiation of international activities, the initial attitudes of top management towards international activities, and the overall motivation during the internationalization process.

Table 4 Pre-internationalization behavior and motivations
Full size table

Pre-internationalization behavior

Unsolicited orders from foreign clients were a trigger for three firms to start their international activities. In these cases, management had a reactive behavior (UM) at the beginning, but changed to more proactive as internationalization evolved. Three other firms have shown proactive behavior (IE) since their early international moves. For example, one of the partners at InBox, a firm located in a less developed region of Brazil, explained that, as the local market showed signs of saturation, management considered to enter more developed regional markets in Brazil, but concluded it would be easier to expand to foreign markets. Only one firm (MetaPar) showed a pattern consistent with network theory (NT).

Main motivations

With the exception of MetaPar, all other firms presented a market-seeking motive (UM) as a general pattern along the internationalization process. MetaPar followed the multinational corporation, serving first its subsidiaries and later also some of its foreign customers (NT). The other firms aimed at acquiring new customers, not necessarily in search of rapid growth (as predicted by the IE perspective), but more often because local demand was insufficient (for example, a narrow market niche for the firm’s products), or simply because new opportunities in the international markets materialized.

Table 5 presents the initial levels of market uncertainty faced by each firm, the attitude towards risk as internationalization evolved, the speed of internationalization, and the outcomes of the internationalization process by 2015.

Table 5 Uncertainty, risk, speed and outcomes of internationalization
Full size table

Attitude towards market uncertainty and risk

As we examined the cases, we decided to separate uncertainty and risk, because these two constructs present different manifestations during the international trajectory of the firms examined. Uncertainty is characterized by “a high level of ignorance” (Barney 2011, p. 217) about a situation, in such a way that the probability of future outcomes is unknown. Risk, however, implies that a decision must be taken based on “a [known] probability distribution of possible outcomes.” Market uncertainty is the relevant kind of uncertainty faced by a firm that enters a new market (Johanson and Vahlne 1977, 1990). If the firm lacks market knowledge, it often seems very difficult, or even impossible, to anticipate how an exchange will evolve, making the firm vulnerable to opportunism (Barney 2011). However, when probabilities are known, firms may deliberately accept different levels of risk in their international activities based on their preferences and expectations.

With the exception of InBox and MetaPar, the other entrepreneurs faced substantial market uncertainty in their early experiences in international markets. InBox started its activities in Portugal (a country with low psychic distance to Brazil because of colonial ties) with a partnership with a local firm, in which a top executive was a Brazilian, a relationship established when he and InBox founders worked together for IBM. MetaPar started and continued its internationalization within a network; as a consequence, their market was the network itself (Johanson and Vahlne 2011). Four other firms made rather rapid steps into the unknown (F-Expert, L-Move, Lumina, Xperience). The high level of uncertainty faced by these firms seems to be related to their lack of international experience. Only four entrepreneurs had studied abroad (three from the same firm, Lumina) and only five (from three firms) had previous professional experience in multinational companies (all but one in Brazil).

F-Expert provides a good example of how some entrepreneurs show low aversion to uncertainty. In fact, F-Expert’s entrepreneurs entered markets with very limited knowledge, if any, of how to do business in these markets (“we did not know what kind of game we were playing”). In the USA, they suffered from the liabilities of smallness and learned that their size was an impediment to sell their product to the government (“it is almost impossible for a small firm”). They also faced a major loss in a project in the USA due to the opportunistic behavior of a partner and learned that not all foreign partners should be trusted. In Colombia, they faced liabilities of outsidership and corruption, and ended up exiting the market (“we bet and lost”). The entrepreneurs incurred in severe financial losses that threatened the firm’s very survival, and decided to de-internationalize.

The attitude towards risk, however, differed. Most firms, if not risk-averse, aimed to minimize risks. InBox adopted from the very beginning a gradual approach to internationalization, striving to keep risks low (UM). In fact, with the exception of L-Move, entrepreneurs seem to become more risk averse as they progress in their internationalization process, and the main reason for the change seems to be the experience gained from a trial-and-error approach to international activities. L-Move shifted from a low-risk approach to internationalization (based on foreign representatives) to a more risk-taking strategy (opening an office in the USA), which proved unsuccessful (“but one has to try”), also leading to de-internationalization. Xperience founders quickly learned from an unsuccessful experience and changed their approach in order to use networks as a “safety net.” Networks have been used by four firms as a means of reducing uncertainty and risk (NT).

Speed of internationalization

The speed of internationalization varied: one firm (InBox) showed a slow and gradual pace (UM) in their internationalization process; four others presented an initial accelerated process (IE), but in two of them, the process slowed down later, culminating with de-internationalization (L-Move and F-Expert). One firm (MetaPar) had the speed of internationalization determined by the network (NT). Another (Tandem) did not fit any predicted pattern; after two initial reactive export experiences, it took the firm a few additional years to establish a joint office in a foreign country.

Sequence of foreign markets

To determine the level of psychic distance from a foreign market to Brazil, we used the results of Leite et al. (1988) study with Brazilian executives. Table 6 shows the sequence of foreign markets for each company according to their psychic distance to Brazil. Only one firm started internationalization in a market with low psychic distance to Brazil.

Table 6 Sequence of foreign markets in terms of psychic distance
Full size table

The temporal sequence of foreign markets entered follows different patterns: logic of psychic distance (UM), logic of relationships (NT), logic of market niche (IE), and, in some cases, no pattern. Two firms followed the logic of relationships predicted by network theory: firms chose markets that they could access using their relationships (Lumina and MetaPar). MetaPar presents a symbiotic relationship with a multinational corporation, serving different subsidiaries all over the world, while Lumina accesses multiple network relationships. Two firms follow the logic of market niche (F-Expert, Xperience) and another follows the logic of psychic distance (InBox). However, it should be noted that even in these last cases, relationships have also been mobilized to enter certain foreign markets, although this was not, in our judgment, the main criterion. The other two firms (L-Move and Tandem) did not fit any predicted pattern.

Sequence of entry modes

We have looked at (i) the sequence of modes adopted over time independently of markets (Table 7) and (ii) the sequence adopted in specific markets (Table 8).

Table 7 Sequence of modes independently of markets
Full size table
Table 8 Sequence of modes adopted in key markets
Full size table

All seven firms examined used direct exporting at some point of their international trajectory. Typically, direct exporting tends to be one-shot, arms-length transactions, either in the beginning of the internationalization process, or at any time, responding to unsolicited orders, or first approaching a new market. Local representatives tend to be established later. Interestingly, however, five firms have, or have had, offices in a foreign country, either temporary offices (in foreign incubators), joint offices (with a local partner), or their own marketing subsidiary (InBox and L-Move). Four firms started their internationalization with exporting, a low-commitment entry mode (UM). Two firms made their initial move to a foreign market by establishing themselves in a foreign incubator financed by a Brazilian government agency, a low cost, low investment mode (IE).

When one looks at the sequential pattern of modes in specific markets (Table 8), however, a different picture emerges. One firm (F-Expert) has only used exporting as an entry mode, a decision that can be interpreted as a desire to serve foreign markets with low-commitment entry modes. However, InBox and L-Move show a pattern of increased commitment, moving from lesser to higher-commitment entry modes (UM), and MetaPar serves the MNE network (NT). The other three firms show an unusual sequence of low cost, low investment entry modes (IE). They use local facilities (temporary office or joint office) to develop familiarity with the market, establish relationships with local firms and individuals, and initiate local sales, and then they choose a local partner to perform commercial and post-sales services locally, and put an end to their physical presence in the specific market. An entrepreneur from Lumina explained the logic behind this sequence: “To sell in the U.S., or in any other country, you need to have a presence for a certain period of time. And until you have a local partner that can deliver good services to your clients there, you have to have a physical structure in that market… But once things start to get going, we don’t need to be physically there, especially after the advent of the internet…”. He stressed that one of the main advantages was learning about cultural differences and how to deal with them.

Relationship between growth and internationalization

Growth was neither a major concern of the firms nor an imposition of stakeholders. One reason might be the fact that these firms did not receive angel investments, or even later, venture capital from outside investors. Therefore, they did not have to achieve high growth rates to fulfill investors’ expectations, as many of their counterparts in developed countries. In addition, they operate in a much more hostile domestic environment, with one of the highest interest rates in the world, and substantial (and often abrupt) changes in exchange rates. The one firm that showed very high growth rates was InBox, but in this case, growth was not induced by internationalization.

Role of networks in the internationalization process

With the purpose of identifying the importance of networks in the internationalization process, we looked at the following indicators: (i) whether the first international activity was prompted by pre-existing relationships; (ii) considering all new market entries, what percentage came from network relationships; and (iii) the overall pattern identified (Table 9).

Table 9 Network relationships
Full size table

Network relationships appear in several occasions of each firm’s internationalization process. For three firms, the first market experience was the only one that was not prompted or supported by a network relationship. However, in spite of the frequency of network events in foreign market entries, the role played by networks was quite different. For two firms, relationships were neither central nor strategic, but could be mobilized to facilitate the internationalization process. On the contrary, for MetaPar, networks were the essence of the process. Finally, network relationships were seen as a valuable asset and a critical element for the success of internationalization for the other four firms (as noted by a key entrepreneur at Lumina: “nothing happens without networking”).

Engine of internationalization

Finally, we looked at the overall mechanism of internationalization suggested by the three theoretical perspectives examined. InBox shows the typical cycle of learning and commitment proposed by the Uppsala model (UM), and MetaPar follows the predictions of network theory (NT). Four other firms nicely fit the central concept of international entrepreneurship (IE), with the entrepreneur as the engine of the firm’s internationalization (F-Expert, L-Move, Lumina, and Xperience). One firm (Tandem) does not fit any predicted pattern.

Appendix Table 10 summarizes the findings of this study. Examining how each case adheres or not to the theoretical propositions, it becomes clear that InBox follows the Uppsala model almost strictly, except for the more proactive behavior in the first international move, and the lack of connection between growth and internationalization. Neither of these aspects, however, are key assumptions of the model. It should be noted that this firm presented a very fast growth trajectory in the domestic market; by 2008, it was already much larger than the others. Therefore, it is possible that entrepreneurial cautiousness is associated with the fact that the entrepreneurs had much to lose very early in the firm’s international development. In addition, L-Move’s and Tandem’s internationalization also seem to be better (but only partially) explained by the U-Model, although their trajectory also fits some predictions of other theoretical perspectives. As to MetaPar, this firm’s international trajectory fully adheres to the propositions of network theory. Finally, Xperience’s trajectory is mostly explained by the predictions of the international entrepreneurship approach, and F-Expert and Lumina are truly “mixed” cases.

Final considerations

The results show that no single theoretical perspective examined in this study (the Uppsala model, network theory, or the international entrepreneurship approach) seems to be able to explain the international trajectory of the INVs over time. We identified “pure” cases (cases that can be explained by one single theoretical perspective) and “mixed” cases (cases that need different theoretical perspectives to explain different periods and aspects of their evolution). This means that none of these perspectives, by themselves, is capable of fully explaining the phenomenon. The present study suggests, therefore, that there is no single international trajectory for INVs. In spite of the differences, some aspects of the theoretical perspectives examined can still be reconciled when analyzing their internationalization process over time. Also, different theoretical perspectives may be more useful in explaining certain phases of internationalization than others.

The study contributes to the understanding of how INVs evolve in their long-term trajectory. Interestingly, despite their early internationalization, which allows to classify them as INVs, the firms studied did not follow the INV’s expected pattern in their subsequent lifecycle phases. In fact, although several scholars claim that the Uppsala model is no longer valid to explain SMEs internationalization (e.g., Oviatt and McDougall 1994; Gabrielsson and Gabrielsson 2013), our study shows that this model can still be used in some cases, at least in the context of emerging economies. As to network theory, our findings do not fully support the contention in the extant literature that networking is very important for INVs’ success (e.g., Gabrielsson and Gabrielsson 2013; Romanello and Chiarvesio 2017). In our study, although all the firms studied practiced some sort of networking, de-internationalization occurred when there was a strategic use of networking, even if the data gathered did not show any logical connection between these two variables. Finally, IE literature is useful in understanding the role of the international orientation of the entrepreneur as the engine of INV internationalization even in later phases of their lifecycle.

Due to the limitations arising from the research method used, it is not possible to state, however, the extent to which each of these theoretical perspectives applies to the population from which the cases were drawn. In addition, it is reasonable to suppose that the trajectory of INVs is strongly influenced by context. In this study, the fact that all firms were from Brazil must have to some extent shaped their trajectory. Therefore, it is possible that our results are biased by the specific context of Brazil: a resource-rich land with a tradition of exporting commodities, lacking a country image in terms of high technology. In addition, Brazilian firms operate in a hostile environment, at least to some extent: volatile exchange rates and very high interest rates.

It is clear that there is “a paucity of information” concerning INVs’ “evolutionary trajectory from birth to maturity” (Etemad 2017, p. 118), especially when considering firms from emerging countries, and, even more from Latin America (Kiss et al. 2012). In addition, Kazlauskaitė et al. (2015) point out that INVs from emerging economies may behave differently than those form advanced economies. Further research is needed to explore the long-term development of INVs, taking into consideration their different contexts of origin.