The Relationship Between Timing, Speed, and Performance in Foreign Market Network Entry

The purpose of this article is to examine the performance consequences of an early and rapid foreign network entry process. We develop a theoretical model that uses theories on first mover advantage and capability development as mechanisms to explain the relationship between the timing and speed of entering the network and SME performance in the network. The theoretical model is tested with a dataset collected on-site at 198 SMEs. The results show that earlier network entry positively affects performance and that the speed of relationship development in the network entered has a curvilinear (inverted U-shaped) effect on SMEs’ performance. Although the extant literature recognizes that firm performance is affected by the degree of insidership reached in foreign business networks, little is known about the performance effects of the timing and the speed of entering the network. Our findings suggest that managers of exporting SMEs should move first into new networks to give themselves time to grow slowly in the foreign market network to mitigate diseconomies of time compression.


Introduction
Two of the most dominating recent research streams on small-and medium-sized enterprises' (SMEs) internationalization -that is (1) research that highlights the importance of networks for SMEs' internationalization and (2) studies on the timing, earliness and/or speed of SMEs internationalization -have rarely connected to engage in a serious conversation . While the former stream claims that SMEs' internationalization is best understood as a network entry process (Sandberg, 2013) and to become an insider (Johanson & Vahlne, 2009;Schweizer, 2013), little is known about the timing and speed of such a network entry . Similarly, even though recent research on SME internationalization processes has made progress in its understanding of the impact of timing and speed on the process (Hilmersson et al., 2017), little of these insights have been employed to understand foreign network entry processes . Since internationalization traditionally has been considered as the spread of activities over country borders (Batsakis & Theoharakis, 2021) where the main liability to be overcome is that of foreignness , research defines earliness of internationalization as the time elapsed between inception and the first sales in a foreign country (Acedo & Jones, 2007;Gassmann & Keupp, 2007;Musteen et al., 2010;Ramos et al., 2011;Rialp et al., 2005). In addition -highlighting the liability of foreignness (Johanson & Vahlne, 1977) -speed of international expansion is measured by dividing the number of countries entered by the time elapsed since the first foreign sales Chetty et al., 2014;Hilmersson et al., 2017;Hsieh, Child, Narooz, Elbanna, Karmowska, Marinova, Puthusserry, Tsai, & Zhang, 2019;Sadeghi et al., 2018).
Drawing on the above, this paper takes the initiative to bridge the above-sketched lack of conversation between the stream of literature pertaining to the internationalization of SMEs as a network entry process and the stream of literature focusing on earliness and speed of entries into new countries as suggested by Forsgren, Holm and Johanson (2015). We use a unique data set from Swedish SMEs' entries into foreign business networks that were opened to enter at a specific point in time. In the early 1990s, the former Soviet Union and China went through major liberalizations and the business networks in Russia, the Baltic States and Poland along with the Chinese business network offered previously unseen business opportunities for Western firms (Jansson & Sandberg, 2008). In this study, we track the entry of Swedish SMEs into these major growth markets of that period, and we seek to explain the performance consequences of the timing and speed of entry into the newly opened business networks. Inspired by Jiang et al., (2014), we capture the timing of network entry by subtracting the year an SME had its first sales in the network from the year the foreign network was opened for international trade. Furthermore, we capture the speed of network entry by dividing the network position gained by the time taken to reach this position as suggested by .
Since most of the SMEs in our study use exporting as their predominant mode of doing business in foreign markets, our study is specifically about exporting firms. Assuming that a customer is connected to a network, in our paper the first export sale to a customer in a foreign market represents the entry node into a new network (Hilmersson & Jansson, 2012). Accordingly, we use the point of time when the first sales in the foreign network took place as our measure of timing of network entry. Our approach is consistent with the network perspective (Blankenburg Holm et al., 1996;Johanson & Vahlne, 2009) arguing that international business takes place in a network setting because each business is part of a network of other relationships. Thus, in our study we consider that the customer does not work in isolation, but has relationships with other actors in the network, such as suppliers, distributors, customers' customers, and institutional and governmental actors. We build on the rationale that the customer relationship allows the SME to enter this wider network in the foreign market where the customer resides. As the firm interacts with this customer, the firm learns about the customer's strengths and weaknesses and about the foreign market (Johanson & Vahlne, 2009;Fraccastoro et al., 2021). Furthermore, drawing on , we understand network entry speed as the relation between a well-developed (in terms of business transactions) position gained in the business network and the time taken to reach this position. We thereby contribute to the internationalization literature by empirically testing an alternative view on internationalization speed. Understanding the internationalization process from a network perspective, we consider internationalization as a process of network entry and positioning by the firm. This is in line with Johanson & Vahlne (2009), suggesting that the main challenge of the internationalizing firm is to move from a network outsidership to an insidership position in relevant networks. We contribute to this literature by shedding light on the timing and speed involved in the process of moving from outsider-to insidership in business networks.
Second, while existing research has suggested that an SME's network position is positively related to the firm's performance (Chetty & Patterson, 2002;Gerschewski et al., 2020;Hilmersson & Jansson, 2012;Johanson & Vahlne, 2009), we still lack knowledge on how timing and speed of network entry influences performance within the network. This leads to our second contribution, which is to reveal the performance consequences of both the timing and the speed of a firm's network entry.
In sum, we seek to answer the following research question: How does the timing and speed of foreign network entry influence SME performance? Along the lines suggested by Johanson & Vahlne (2009;, we view internationalization as a capability development process and hence, we develop our theoretical model by drawing on mechanisms in capability development theory and insights from the first mover advantage literature. We test our theoretical model on a sample of 198 foreign network entries by Swedish exporting SMEs. The data, which was collected on site at all firms' facilities, shows that: (1) the earlier in time the network was entered, the better the performance, and (2) there is a curvilinear (inverted U-shaped) relationship between network entry speed and firm performance. Consequently, we suggest that firms should move first to allow themselves time to grow slowly in the market entered. If the entry is postponed for a later period, then the network position will need to be developed with a high speed to catch up with competitors. Thus, there is a greater risk that the firm's resources and capabilities will be overstrained because of time compression diseconomies (TCD) in their development process.
We structure the remainder of the paper as follows. First, we present a review of the literature. Second, we provide the theoretical background of the paper, wherein we introduce the underlying mechanism of the hypothesis. Third, the hypotheses of the paper are developed. Fourth, we present the method and results of the study. Fifth, the results are discussed, followed by the conclusions, managerial implications, and suggestions for further research.

Internationalization as a Network Entry Process
The network perspective views markets as systems of long-term interdependent relationships between customers and suppliers (Blankenburg- Holm, Eriksson & Johanson, 1996;Håkansson & Johanson 1993;Yamin & Kurt, 2018). In these relationships, firms adapt and modify their operations, which means that mutuality and interdependence emerge. The network perspective suggests that internationalization is the process of firms' actions to establish relationships and strengthen network positions (Johanson & Vahlne, 2009). In this process, firms establish new positions, develop old positions, or increase the coordination between positions in a network (Johanson & Mattson, 1988). Their decisions are contingent on interactions with other firms (Chetty & Blankenburg-Holm, 2000;Ellis & Pecotich 2001;Sedziniauskiene et al., 2019;Wilkinson & Young, 2005). The firm's position is better developed if the firm has more long-term and robust relationships, and the strength of relationships reflects the magnitude and symmetry of interdependence in the relationships.
A firm's network position is a source of information and strengthens the resource base (Gerschewski et al., 2018;Lee et al., 2020;Puthusserry et al., 2020). However, the network also hinders opportunity development in path-dependency rigidness (Sandberg, 2013). When a firm enters a specific country market, it establishes a position in the foreign market's network and progresses from being an outsider to holding an insidership position (Johanson & Vahlne, 2009;Kurt & Yamin, 2018).
An insidership position implies many well-developed and interdependent relationships (Hilmersson & Jansson, 2012), which promotes cooperation and coordination (Coleman, 1990;Gerschewski et al., 2018). This is because insidership entails a firm network, where firms have a lot of knowledge about the other firms in the network and where the same knowledge reaches the firm from different sources, such as suppliers, customers, and competitors. The firm has a central position and few or no relationships with outside firms that can introduce new knowledge (Burt, 1992). An insidership position thus implies that much of the knowledge gained is a repetition of the knowledge already circulating in the network, as the knowledge received from one counterpart already contains the knowledge from another counterpart, since they are likely to be connected (Burt, 1992;Cuypers et al., 2020).
While earlier literature focuses on network position, our paper draws on the insights of  about entering the network. They purport that network entry speed relates to what happens after internationalization has begun. Hence, the starting point is the first activity in the foreign market network, which is in line with Sandberg (2013) directing her interest to entry node into a foreign busi-ness network.  argued that since business networks are dynamic and changing, entering a foreign market's network is about synchronizing the firm's organization to the dynamics of a new and changing network. Synchronization activities, i.e., temporal adaptation and coordination of the activities in the network, influence the speed of entering the network. In other words,  suggested that synchronization is an important factor in explaining how fast the firm can establish a position in the foreign market network-not at least in a changing network (cf. Helfert et al., 2002). Lack of timely synchronization between the firm and the network's activities drags out the entry process. In contrast, if there is a good fit between the entering firm and the foreign market network's activities, it is more likely that entry could be expedited. It is important to note that adaptation or coordination can be mutual, but firms can also adapt without their counterparts doing the same .

Capability Development and Internationalization
Internationalization is a capability development process dependent on the experience generated by the firm (Hilmersson & Johanson, 2021;Mélen Hånell, Rovira Nordman and Mattsson, 2021). The dynamics of capability development is a basis for the incrementalism in Johanson and Vahlne's internationalization process model (cf. 2009). Thus, TCD provides a suitable lens to examine the dynamics of quickly entering a network and its impact on performance. Furthermore, previous studies have shown that by taking a short period of time to develop their capabilities, internationalizing SMEs can face hurdles that have a negative impact on their performance (Hilmersson & Johanson, 2021;Sadeghi et al., 2018).
Research on TCD holds that individuals and organizations will be subject to diminishing rates of return when subjected to time compression. When all input to a process is held constant except the time, the output will be weaker (Dierickx & Cool, 1989). TCD are a consequence of human and organizational limitations in information processing (Vermeulen & Barkema, 2002;Chen et al., 2012). Within a certain period of time, humans and organizations can only process limited amounts of information efficiently. Bounded rational managers may take irrational or inappropriate decisions when overloaded with information within a short time and inefficiencies may appear.
At the organizational level, the absorptive capacity of the firm refers to the organization's ability to handle information. Absorptive capacity has been defined as "the ability of a firm to recognize the value of new external information, assimilate it and apply it to commercial ends" (Cohen & Levinthal, 1990, p. 128). As the internationalizing firm expands, there is a risk that it becomes overloaded with information, which diminishes its capacity to absorb new information, and thus it experiences TCD (Hilmersson & Johanson, 2016). This research stream on TCD and speed of internationalization suggests that the faster the essential resources or capabilities are developed, the higher the cost of this process (Freixanet & Renart, 2020;Jiang et al., 2014;Pacheco-de-Almeida & Zemsky, 2007;Sapienza et al., 2006). To date, the extant literature on TCD and internationalization has focused on speed of foreign expansion (Vermeulen & Barkema, 2002;Jiang et al., 2014), early internationaliza-tion and its impact on firm survival and growth (Sapienza et al., 2006), geographic scope, managers' competencies on SME survival (Freixanet & Renart, 2020), and speed of SME internationalization and performance (García-García et al., 2017). We build on this stream of literature by taking a novel approach to focus on timing and speed of entry into a specific foreign network, and we use theories on capability development as our mechanisms to understand the impact that the timing and speed of entering the network has on performance.

Timing of Network Entry and Firm Performance
Researchers are becoming increasingly interested in the temporal nature of what the internationalizing firm does (Grace Masango & Lasalle, 2020;Hult et al., 2020). We expand on this because we are interested in the timing of the network entry. Following the arguments of Johanson & Vahlne (2009), we established that entering a foreign market network is about creating an insider position in the foreign business network.  described insidership and outsidership as the set of direct relationships with customers and suppliers, the relationships with institutional organizations, and the indirect relationships with the customers and suppliers' counterparts that the firm has developed in the foreign market. The network entry generally takes place as the firm replaces other network actors, connects previously unconnected network actors, and/or brings value to the existing network.
When a firm enters a foreign market, it establishes a position in this market's network and changes from being an outsider to being an insider (Johanson & Vahlne, 2009). Hence, network entry speed can be defined as how quickly a firm gains a position in the foreign market network ; therefore, the speed at which the firm's relationships are developed in the focal network indicates its speed of network entry. Since a changing network position where the firm goes from outsidership to insidership reflects a process over time, this means that speed to internationalization covers a period characterized by outsidership, and speed of internationalization covers a period starting from outsidership and moving towards insidership . We argue that the timing at which such an entry into a foreign network takes place, plays an important role in determining the firm's subsequent performance.
It takes time and financial resources to build long-lasting relationships, to learn about the resources in the foreign network, and how to combine these resources with the SME's resources to develop an insidership position Johanson & Vahlne, 2009). Thus, by entering the network early, firms will have more time compared to their competitors, to get acquainted with their customers, and to learn from interactions in these collaborations incrementally to develop their networking capabilities. These networking capabilities include the ability to coordinate the network, use their existing relationships to build new relationships, and combine their resources with their customers' resources to create new opportunities (Pinho & Prange, 2016;Sarasvathy et al., 2014).
The SME needs to engage in various marketing activities that involves time and financial resources to conduct business in foreign markets (Eriksson et al., 1997;Martín Martín et al. 2021). Before it starts exporting, the SME expands its resources by acquiring information from its domestic network and builds legitimacy through the reputation of this network (Boehe, 2013;Haddoud et al., 2021). The resources from the domestic network trigger export opportunities and also enables the SME to reduce its exporting costs. The SME needs to invest financial resources, time and effort to collect information, and to learn about the foreign market, such as customer, competitors and other actors in the network (Hilmersson & Jansson, 2012). Furthermore, the SME has to learn about the ways of doing business in other cultures and the rules and regulations in the foreign market (Eriksson et al., 1997). There are administration costs for exporting to a wide range of countries with different geographic, cultural and institutional contexts (Abdi & Aulakh, 2018; Martín Martín et al. 2021) such as setting up a separate unit to deal with exports (Beamish et al., 1999). Furthermore, the firm may need to invest in adapting its product and production processes to satisfy the requirements of its customers (Martín Martín et al. 2021). Interdependence in the relationship is a result of mutual commitment and investment, more interdependence means that it is more costly to terminate (Johanson & Vahlne, 2009;Yamin & Kurt, 2018).
In addition, there are travel costs because the firm needs to visit its customers and other network partners in various foreign markets to build and maintain relationships. In particular, the foreign markets in our study such as China and Russia need more time and investment to build networks and legitimacy because their institutional environments at the time of our study were undergoing dramatic changes and their styles of building networks are very different from Sweden (Jansson et al., 2007). In Russia and China, the changes in both the networks and institutions occur concurrently, and this involves high uncertainty.
We build on Johanson and Johanson's (2021) reasoning and argue that insights from first mover advantages (Fosfuri et al., 2013;Suarez & Lanzolla, 2007) are important in order to tease out the relationship between the timing at which a firm enters a network and the firm's subsequent performance. Research on first mover advantages has identified two main reasons why earliness generates positive outcomes (Suarez & Lanzolla, 2007). Firstly, "isolating mechanisms" (Rumelt, 1987), which enables first movers to prevent competition from imitators, for example, because of switching costs relating to terminating one relationship to form a new relationship. Secondly firm-level resources, capabilities, and "complementary assets" (Teece, 1986) enable firms to benefit from early mover advantages because they have the potential to choose from a large pool of complementary network partners that are available.
By acting early, firms will be able to make choices from a superior and larger pool of resources to subsequently develop a unique bundle of resources (McNamara et al., 2008). Using this first mover advantage logic, we reason that when firms enter a network early, they are able to be more selective and quickly choose the most compatible network partners in terms of complementary resources and capabilities. Firms can gain an advantage over competitors if they can effectively leverage these complementary resources from their network to make their product or services unique (Vidal & Mitchell, 2013) and thereby making it hard for others to imitate their offerings. In addition, early entrants have the resources and capabilities to immediately identify new opportunities during the narrow time frame available for the opportunity to be of value to the firm and to make informed decisions about whether to pursue these new opportunities. Therefore, we hypothesize that: H1 Earliness of network entry will positively affect firms' performance in the foreign market network.

Speed of Network Entry and Firm Performance
For our first hypothesis, we established the rationale to expect that early entrants will benefit from first mover advantages. Whereas the first hypothesis concerns the timing of the network entry, we also believe that there is reason to expect that the speed at which the network is entered influences performance in the network. Thus, in the section below, we outline why and how the speed at which relationships in the focal network is developed affects firm performance in the network.
Earlier research on the relationship between internationalization speed and firm performance has suggested a curvilinear relationship between the speed at which sales and operations are spread between country markets and firm performance (Autio et al., 2000;Sapienza et al., 2006). This relationship is considered to be driven mainly by two mechanisms. First, the positive slope in the inverted U-shaped relationship is regarded as an outcome of learning advantages of newness (LAN) (Autio et al., 2000;Hilmersson & Johanson, 2016;Sapienza et al., 2006). Research on LAN has suggested that early internationalizing firms absorb knowledge quickly because they are not stuck in old routines that must be unlearned to absorb new knowledge (Autio et al., 2000). Thus, they can learn quickly from their initial internationalization and expand into other foreign markets very quickly. Hilmersson & Johanson (2016) showed that the same pattern can be observed for the speed at which the internationalization unfolds. Second, the negative slope in the inverted U-shaped relationship is regarded as an outcome of TCD (Dierickx & Cool, 1989). Research on TCD showed that, if the capability development process is drastically accelerated, mistakes, misunderstandings, and failures occur (Jiang et al., 2014;Sadeghi et al., 2018). TCD is an outcome of human and organizational limitations. If too much information or experience is to be processed during a limited period of time, mistakes will happen and there is a risk of overheating the organization. Further, there is insufficient time or resources to reflect and learn from these mistakes (Vidal & Mitchell, 2013) and to replenish resources Crossan et al., 1999). As a result, firms continue to make expensive blunders (Jiang et al., 2014), which subsequently impedes their performance.
We rationalize that these underlying mechanisms are of importance in the network entry process. Entering a network is a relationship development process where a position is to be established. Network outsiders need to learn about the network, possibly adapt to the network, and build trust with potential partners in the network (Fraccastoro et al., 2021;Saleh et al., 2014;Verbeke et al., 2019). Therefore, the entrant firm will have to commit resources to its entry process. Research on relationship develop-ment in business markets (Anderson et al., 1994;Håkansson, 1982) has shown that relationship development is resource and time consuming (Menzies et al., 2020).
To develop a position in the network, the firm needs to develop resource ties, actor bonds, and activity links with actors in the foreign business network (Anderson et al., 1994;Håkansson, 1982), while maintaining its relationships in the domestic market. We expect that firms committing significant time and energy to this entry process during a shorter period of time will experience a positive network entry performance.
However, if the period during which the entry takes place is shortened significantly, we argue that there is a greater risk that TCD will set in, and the entry process fail. This is partly because early entrants are outsiders (Johanson & Vahlne, 2009). The fast pace of the learning process will be highly challenging in developing the level of trust and commitment with local actors needed to become insiders in the foreign market network (Hilmersson, 2011). It is difficult for firms to absorb a large amount of knowledge within a very short space of time (Chen et al., 2012;Jain et al., 2019) and then assimilate it quickly within the firm (Jiang et al., 2014). The fast pace is costly and overstrains the firm's resources because it must concurrently develop multiple new capabilities alongside the essential ones to overcome the liabilities of foreignness and newness (Freixanet & Renart, 2020;Jain et al., 2019;Sapienza et al., 2006) and outsidership (Johanson &Vahlne, 2009).
Early entrants particularly in emerging markets experience high environmental uncertainty mainly because of rapid changes in regulations and investment opportunities (Bowman & Hurry, 1993;Jiang et al., 2014). When environmental uncertainty is high, early entrants experience TCD because they have ambiguous and insufficient information with which to adapt quickly to the constantly changing institutional environment, which results in failure. In addition, prior successes in a particular foreign market could also lead to subsequent failures, because the environment might change rapidly, leaving the firm without the relevant resources and capabilities to adjust quickly to these changes (Jiang et al., 2014).
Furthermore, as the firm's experiential knowledge increases, it might recognize several opportunities, but have difficulty deciding which specific opportunity to pursue . The firm might also have enough experience to recognize that it lacks the appropriate resources to pursue a new opportunity (Vidal & Mitchell, 2013). Taking up new opportunities might result in entering new networks or strengthening existing ones. Since network partners must make quick decisions, they might initially commit to the relationship with good intentions and then subsequently realize that they have overcommitted because they lack the time or resources for the collaboration (Fraccastoro et al., 2021). Another perspective is that network partners could behave opportunistically because the firm did not have sufficient time, resources, and capabilities to build trust and mutual commitment in the relationship (Fraccastoro et al., 2021;Oliveira & Johanson, 2021;Verbeke et al., 2019). Since they have not had enough time to build trust with their network partners, firms might perceive some risk and uncertainty in their decision to commit resources to the relationship (Johanson & Vahlne, 2009). Therefore, we hypothesize that: H2: There is a curvilinear (inverted U-shaped) effect of the speed of foreign network entry on firm performance.

Sample and Data Collection
We tested our theoretically generated hypotheses with data collected on-site from 198 small-and medium-sized manufacturing firms (SMEs) from Sweden. We report on a sample that we followed for 8 years, during this period we visited the firms on-site in 2008, we conducted an e-mail based survey in 2012 and we compiled data from their annual financial statements (covering the period 2006-2014) as well as their entries in European patent registers up until 2015. In this study, we are interested in studying the role of timing and speed of entry into foreign business networks, a question that  suggested to be important for empirical analysis.
To enable an analysis of the speed of network entry, we needed a fixed startingpoint for when business networks were entered. For this purpose, we found the research design developed by Jiang et al., (2014) very inspiring. Whereas Jiang et al., (2014) studied the establishment of subsidiaries by foreign MNCs after the opening up of the Chinese economy, we decided to use data from entries into foreign business networks that were opened for Swedish firms to enter at an almost specific point in time. Hence, we focused on the Russian, Estonian, Latvian, Lithuanian, Polish, and Chinese market. These markets were considered major potential growth regions for Swedish firms because of liberalisations and openings in the early 1990s. Prior to that, very few Swedish SMEs had any contact with the business networks in these markets. Therefore, the liberalisations taking place in the early 1990s offered a unique opportunity to, in retrospect, understand the consequences of the timing and speed of entering a business network. The data we report on covers developments of Swedish SMEs in these business networks between 1990 and 2008, and is based on our on-site survey conducted in 2008. Since these markets are unfamiliar for the Swedish SMEs and they lack knowledge about these markets, we adopt Johanson and Vahlne's (1977) assumption that they are high uncertainty environments for these SMEs.
For our sample, we ordered data on all exporting manufacturing SMEs in Southern Sweden with a total export turnover exceeding €1 million falling under the EU definition of SMEs (headcount < 250) with exports exceeding €100 000 to any of the markets accounted for above. We use manufacturing firms to test our theoretical model which is in line with the network perspective developed by Johanson & Vahlne (2009). In total, a sampling source of 692 SMEs was returned to us. Subsequently, we followed two steps to exclude firms that did not belong to our population. First, we screened the list from Statistics Sweden, and second, we made personal phone calls, leaving us with 277 SMEs as a representative sample. Once the sample was identified, our data collection followed an on-site design. This strategy provided three advantages. First, it offered us an important opportunity to talk to the respondents. Considering that we are developing new measures on the network entry process we believed that a faceless survey would have been problematic. Instead we realized that we needed the opportunity to clarify some of our measures and to ensure that our questions where understood as intended. Second, we could ensure full attention from experienced and well-informed respondents while answering the questionnaire. Third, it ensured a high response rate. As for the latter, we achieved a response rate of 73%, which meant that 203 of the 277 SMEs were visited on-site for the interviews. We had 74 SMEs that did not participate because they were unable to invite us for a visit during the time allocated for data collection, they referred to policies of not participating in research, or they could not be reached after four attempts. After the data collection, we tested for non-response biases by comparing the size, turnover, and industry between the responding and the non-responding firms in the sample. From a variety of such tests, we were unable to identify any problematic patterns or biases in the dataset.
To identify the individual respondents, we followed an intra-firm snowball technique by asking for the person with greatest insight into the international activities of the firm in general, and particularly in the specific markets that we are interested in studying. We ended up with respondents in the following positions: 55% were CEOs, 17% marketing/sales managers, 8% area sales/marketing managers, and the remaining 20% were business development, key account, and product managers. Each meeting with a respondent lasted around 1.5 h, where half of the time was dedicated to a general semi-structured interview to understand the international activities of the firm, and the other half of the time was used to guide the respondent through our structured questionnaire. The on-site study was followed by an e-mail based survey in 2012 and compilation of financial statements and patent registers for each firm.
Both prior to our data collection and during our data analysis, we made several efforts to minimize potential method biases related to a common rater of dependent and independent variables. To do so, we carefully followed the advice by Podsakoff, MacKenzie and Podsakoff (2003). First, during the interviews, we prevented the respondents from searching for implicit theories by contextually and temporarily separating our variables tested. The statements were answered in different sequences, and we did not ask directly for the independent variable. Instead, we calculated the independent variable from two variables in the questionnaire. Therefore, the respondents could not possibly identify the model to be tested while they filled in our questionnaire. We also held discussions with respondents between the different sections of our questionnaire. Before each section of the questionnaire was answered, the interviewer made clarifications and responded to potential comments from the respondent. Once the questionnaire was completed, we held a general discussion with the respondents to ensure that no answers were problematic and to reduce the number of missing values. After the data collection was completed, we ran one factor tests and a marker variable test with several different variables in the dataset without identifying any biases related to the single respondents. Since we collected new data via an e-mail based survey in 2012, we could also validate the values reported on many of our variables. Consequently, we are confident that there are no severe biases in our data related to single respondents. While preparing the data for the test of our hypotheses, we removed 5 SMEs from the dataset due to missing values in the dependent variable. Hence, we tested the hypotheses on 198 SMEs out of the 203 that we interviewed.

Measures and Control Variables
Our dependent variable of the research model is market entry performance. Measuring performance on the (international) market level for SMEs is a difficult task, since it is rarely available from objective sources. Therefore, we were forced to access this data with reflective items (Katsikeas et al., 2000). It has been suggested that attitudinal measurements of performance can be used as a reflective way to capture performance since respondents are equally inconsistent in their responses, but the responses have a central tendency and a variance (Zaller & Feldman, 1992). In this study, we measured performance by asking two questions addressing the firm's profitability level in the business network entered. First, we asked the respondent to answer the statement, "We are pleased with how the profitability of our business in X has developed over the last three years." Second, we asked the respondent to answer the statement, "We are not satisfied with the return on the resources and time invested in X." X in our questions represented the market entered by the firm, thus, it represented China, Russia, Poland, Estonia, Latvia, or Lithuania. The respondents were given a questionnaire with the market relevant for their firm. On both statements we used a scale stretching from 1, representing strongly disagree, to 7, representing strongly agree. The answers to the second statement were reversed prior to the analysis to ensure harmony in the direction of the answers. To create the construct for continued analysis, a factor analysis was performed with acceptable values. The factor loadings were above 0.8, the communalities were above 0.7, and the Alpha value was above 0.7, which meets most thresholds in the literature. We therefore decided to merge the items into a summated scale which was used for further analysis.
The independent variables were measured as follows. First, to address the timing of network entry, we measured the earliness of market entry by subtracting the year the firm had its first sales in the 'X' market network from the year of the study (2008) in line with Jiang et al., (2014). Consequently, we captured how early in time the firm entered the business network after the focal business networks opened up for foreign firms. We emphasize that our measure of earliness is inspired by Jiang et al., (2014) and therefore in line with how the concept of earliness has been treated in the first mover advantage literature. An SME entering soon after 1990 is considered to be an early entrant, whereas a firm entering long after 1990 is considered to be a late entrant. This measure should not be confused with how earliness is treated in the International New Venture/Born Global literature (Acedo & Jones, 2007;Rialp et al., 2005), which referred to earliness by studying the time elapsed from the firm's foundation to the firm's first international sales. Our reference point is the year 1990, when liberalizations took place in the focal markets of our study and not the year of firm foundation.With respect to the importance of the time elapsing between firms' foundation and their first foreign sales in explaining international performance among INVs, we included a measure of firm age as a control in our hypothesis's tests.
Second, we measure the speed of network entry in line with the suggestions of , who defined network entry speed as, "How quickly the firm can gain a position in the foreign market network." They suggested that network entry speed is the relation between the position gained and the time taken to reach this position in the network. We followed the same reasoning and build on the assumption that the position gained is an outcome of how well-developed the firms business relationships are in the focal network. This assumption, we anchor in Johanson and Vahlne's (2009) suggestion that network entries are accomplished when an insidership position has been reached. Insidership positions are characterized by low levels of uncertainty and that the firm is a position to recognized new opportunities in the network. Such insidership positions are gained from the development of experience and trust in busines relationships, that is in the entry node (Hilmersson & Jansson, 2012). Consequently, we asked the managers about their total number of relationships in the focal business network and asked them to inform about how large a share of their business relationships that were well-developed (with ongoing business transactions) and how large a share of the business relationships that are under development (negotiations taking place, pilots delivered, etc.). Whereas our question about the stage of development of the business relationships aimed at capturing the position of the firm in the business network, we also asked about which year the firm had its first sales in the local market network to be able to capture the speed of the network entry. Thereafter, we divided the measure of the position reached by the time taken to reach this position. An important benefit of the on-site methodology was that it gave the interviewing researchers the opportunity to describe what we considered to be a well-developed relationship in contrast to a relationship under development to the respondent before they responded to the question, for this purpose an interview guide was used to prevent interviewer biases. The interview guide was based on the research on the foreign network entry process presented in Jansson & Sandberg (2008). In dialogue with the respondents we informed about our view on relationship development and during the interviews we had a printed version Jansson and Sandberg's (2008) model to show the respondents. We believed that such an approach was needed to measure a construct as complex as the network position of the firm.
We included five control variables in our analysis. First, we controlled for the number of employees of the firm, since we expected that the number of employees would indicate the resources available for the entry process and thus influence how its speed affects performance. For this purpose we downloaded data for the year 2008 from official registers. Second, we controlled for the international spread of the firm (in terms of export markets), since earlier research has suggested that internationally experienced firms will outperform those with limited experience and that earlier experiences reduce the cost of international expansions (Eriksson et al., 1997). Third, we controlled for the share of the SMEs total assets that are located outside the domestic market since it would indicate if export is the main foreign mode of entry or if the firm has set up local subsidiaries in foreign markets. Earlier research (Jansson & Sandberg, 2008) has indicated that the entry process in business networks is influenced by the entry mode, and entry mode experience of the firm. Fourth, we controlled for firm ownership by including a dummy variable separating familyowned firms from firms with alternative ownership. For this dummy variable we gave family-owned firms a value of 0 and firms with alternative ownership a value of 1. Thus, the mean value of the variable in Table 1, shows the portion of the share that is not family owned. Fifth, since we are testing our model based on entries into business networks located in six countries of varying size, institutional-and cultural distance to the home market, we controlled for the market entered. For this purpose, we gave the smaller EU countries Estonia, Latvia, Lithuania, and Poland a value of 1, the larger Russian market which is more culturally and institutionally distant a value of 2 and China, the largest and most distant market both culturally and institutionally, a value of 3. In total, 92 firms reported on the Baltic states and Poland, 64 firms reported on Russia and 49 firms reported on China. Sixth, since the INV and Born global literature has shown that the time elapsing between firms' foundation and their first foreign sales influence the subsequent internationalization behavior of the firm, we included a control of the age of the firm. Seventh, the export literature makes a distinction between firms having a direct, or an indirect contact with customers in the foreign market. Sandberg (2013) refers to this as the entry node of the firm and based on her findings we expect that the entry process will be influenced by the characteristics of the entry node in the market. We expect that firms having an indirect contact with the customer, working with intermediaries in the foreign market would have a less resource demanding entry process than firms building direct relationships with the customers.We therefore controlled for an indirect relationship with the customers. We gave firms with an indirect customer relationship a value of 0 and firms with a direct customer relationship a value of 1. Table 1 reports descriptive data and correlations among the variables of the study.

Hypotheses tests
To test the hypotheses, Ordinary Least Squares Regression was performed in SPSS. We present the results in Table 2. To ease the interpretation of the effects revealed and to avoid multicollinearity in the model, we mean-centered the independent and the dependent variables. With this action, we kept all vertical inflation factor values below 2, indicating that multicollinearity should not bias the results. In Table 1, model 1, we see the effects of the control variables on the performance of the firm in the foreign business network. Here, we see that the age of the firm reduces performance in the business network, indicating that younger firms in our sample are more profitable in these markets compared to relatively older firms.
In model 2, we tested the hypotheses by including the independent variables in our equation. Here, we see that the adjusted r-squared value increased by about 9%. The effects of the independent variables are significant, and both our hypotheses have empirical support. First, the results show that the earliness of network entry positively influences firm performance in the foreign business network, providing support for hypothesis (1) For this hypothesis the β value is 0.209 and the effect is significant on the 5% level. Second, the results show that there is a positive effect of the speed of network entry on performance, indicating that, initially, the speed of network entry positively influences the firm's performance. The β value is 0.428 and the effect is significant at the 1% level. Thus, we see a positive performance effect generated by the speed of the network entry process. The positive performance effect, however, is reduced if the speed is drastically accelerated as shown in the results reported for the squared variable. Here we can see that the squared network entry speed variable returns with a significant but negative Beta value. The β value is -0.186 and the effect is significant at the 5% level. This indicates that there is an inflection point in the relationship between network entry speed and firm performance. If the speed of network entry is drastically accelerated the speed starts having destructive performance consequences as we suggested in the development of hypothesis (2) Empirically we confirmed this inverted U-shaped relationship. The data therefore confirm that the speed of network entry positively influences firm performance, but the relationship is not linear. If the speed is drastically accelerated, then it starts to have destructive performance consequences. This effect seem to confirm our expectations that time compression diseconomies will weaken firm performance is the speed of network entry drastically is accelerated.

Discussion and Conclusions
In this study, we ask the question, how does the timing and speed of foreign network entry influence SME performance? We take a novel approach by connecting two parallel streams of literature that are rarely combined, namely, (1) research that highlights the importance of networks for SMEs' internationalization and (2) research on the timing, earliness and/or speed of SMEs internationalization. We draw on network perspective and TCD to better understand the process of capability development after a firm's initial start to exporting. The research on post-entry speed of internationalization (Acedo & Jones, 2007;Hilmersson & Johanson, 2016;Sadeghi et al., 2018) is gaining momentum. We add to this nascent literature by using a network perspective to examine the timing and speed of entry into networks in foreign markets and their impact on firm performance. Since we are interested in how quickly relationships are formed, our study is unique because we measure the speed of firms' movement into a network  instead of the extant trend of firm's position in the network (Johanson & Mattsson, 1988;Johanson & Vahlne, 2009;Schweizer et al., 2010). Thus, our study offers an alternative measure for speed of internationalization compared to the prevailing literature, which uses entering a number of countries as a dimension of speed of internationalization (Chetty et al., 2014;Sadeghi et al., 2018;Batsakis & Theoharakis, 2021). Consequently, we address the timing and speed involved in reducing liability of network outsidership in contrast to earlier research, which hitherto mainly has addressed the timing and speed involved in reducing the liability of foreignness. Our findings provide a deeper understanding of what happens in networks and how the speed of forming a relationship with the customer in a foreign network can either enhance or hinder the firm's performance depending on the timing of the entry. This relationship with the customer would allow the firm to gradually become an insider in the foreign market network.
Indeed, becoming an insider in a foreign market network is critical for successful internationalization because it allows firms to acquire knowledge about the market and to identify new opportunities (Johanson & Vahlne, 2009). Our study provides a better understanding of insidership by considering the speed of entering a network as a capability development process. In particular, we offer novel insights into the timing and speed of entry into the network in a foreign market and its impact on performance. Our findings have implications for theory because we contribute to the debate in the literature regarding rapid versus incremental internationalization (Barkema & Drogendijk, 2007;Gassmann & Keupp, 2007;Sapienza et al., 2006). We provide empirical support for the incremental approach to building networks and for the gradual accumulation of knowledge (Johanson & Vahlne, 2009). Based on Johanson and Vahlne's (2009) argument, we assume that by entering the network too fast, a firm is exposed to inefficiencies that occur because the firm has not had the time to learn and build trust with their customers. Furthermore, our results provide empirical evidence for Johanson and Johanson's (2021) conceptualization that speed of network entry is essential for gaining network insidership, and what happens in the network determines the speed of internationalization.
In addition, our results provide new insights for the "first mover advantage" literature (Suarez & Lanzolla, 2007;Teece et al., 1997) by connecting the timing of enter-ing networks and country contexts that have high uncertainty, such as the countries covered in our study. Based on our results, we deduce that firms that enter a network early, benefit from the early mover advantage, subsequently leading to greater performance than firms that enter later. This is in line with LAN, where early entrants are open to new knowledge that subsequently leads to successful internationalization (Autio et al., 2000;Freixanet & Renart, 2020). It seems that early entrants performed better because they had more time to develop trust and to learn gradually, thus enabling them to assimilate new knowledge into the organization. However, earliness has its downsides, as we found that the early mover advantage declines when the entry into the network occurs too quickly.
Our contribution to the TCD literature is that we offer deeper insights into the speed of entering foreign networks by confirming that when firms enter foreign networks too quickly, they experience TCD, which has a negative impact on the firm's performance. While this result is consistent with findings in other studies relating to TCD and the speed of internationalization (Hilmersson & Johanson, 2016;Jiang et al., 2014;Sadeghi et al., 2018), we add to this literature by including a novel perspective relating to speed of entry into networks. We advance the literature on TCD and SME internationalization by providing unique insights about how the time available in the capability development process when entering new networks has an impact on performance. Furthermore, we consider that TCD could be more pronounced in markets with high uncertainty, such as those in our study (Russian, Estonian, Latvian, Lithuanian, Polish, and Chinese markets) because of a rapidly changing institutional environment. In the context of such unstable environments, we assume that past success does not guarantee future success because of the swift changes in resources and capabilities that are constantly required. SMEs that need to engage in networks in such environments would have to be discerning in their timing of entering the network because if they move too quickly, they could suffer the effects of TCD, which seems to ultimately impede their performance.

Future Research Implications
One implication for future research is to develop better measures for speed of entering the network. In this paper we developed our measures based on Johanson and Johanson's (2021) suggestions as we believe that customer relationships are central for the exporting firms in our study. Yet, we acknowledge that networks are complex, and network positions difficult to measure, which might be one reason why the predictive power of our model is relatively low. We measured the relationship level rather than the network level by only focusing on customer relationships. We see an opportunity for future research to measure more relationships in the network thereby studying the network level. In this paper, we have taken several steps in developing a sound theoretical platform for such attempts and we have shown that network entry speed and timing have significant performance effects. Future measures should seek to have a more inclusive measure of networks and acknowledge that SMEs relationships include suppliers, distributors, competitors, and institutions such as banks and industry associations. However, future research should note that the network structure and interactions within a network would vary depending on the dynamics of the industry context. Since our study focuses on manufacturing firms, future research could examine other types of industries, such as service industries and digitalized industries to compare the speed of entry into networks in other industry contexts. A question to raise is whether some industry contexts are more conducive to rapid entry into networks and thus TCD may not apply to them. Such an approach could help us to examine the role of technological innovations and R&D intensity in explaining the speed of network entry. The lack of inclusion of such a variable is a shortcoming in our study, which we suggest future research to address.
In addition, future research could connect the timing of entering networks and other high uncertainty country contexts as well as the countries covered in our study. While we have used Johanson and Vahlne's (1977) internationalization process theory to make our assumption about a high uncertainty context, future research could develop a measure for this context. Once this measure is developed, it could be used to examine the similarities and differences regarding the timing of entering networks in high uncertainty contexts and low uncertainty contexts and the impact on firm performance.
Since we focused on Swedish SMEs and their entry into the Russian, Estonian, Latvian, Lithuanian, Polish, and Chinese markets, our study provides a foundation for replication in other country contexts. Our results are, therefore, applicable for a context where an SME from a small, developed country with low uncertainty enters emerging markets with high uncertainty. This provides an opportunity for future research to test the boundary conditions of our results by comparing SMEs from other country contexts, such as large emerging markets like Brazil and India, where uncertainty is high in both the home market and in some of their host countries, for example, an Indian firm doing business in South Africa or in Brazil. The findings will answer the question: How will this country's context affect the early mover advantage and TCD and their speed of entry into the network in foreign markets and performance? Such a study would help us gain new insights about SMEs from emerging markets. For instance, will they perform better or worse when transferring their resources and capabilities than the Swedish SMEs when they rapidly enter foreign networks in environments with high uncertainty? Researchers could also explore whether the country context matters for speed of entry into networks or whether networks have no borders.
Since our study focused on SMEs that mainly use exporting, future studies could consider other modes of doing business, such as company employees located in foreign markets, foreign sales subsidiaries, joint ventures, and foreign direct investment. It is possible that firms use multiple modes of doing business and so the question to raise is whether the mode has an impact on the speed of network entry and firm performance. Furthermore, it would be useful to understand whether single or multiple modes of doing business increase the speed of network entry.

Practical and Policy Implications
SMEs need to be aware of the importance of timing and speed when entering foreign networks and be able to balance the speed of the network entry process. They also need to closely monitor their resources and capabilities and consider them when making decisions about entering networks in new foreign markets. Emphasis should be on the capability development process rather than on their current resources and capabilities. SMEs need to gain insidership positions in networks in foreign markets and develop the capability to recognize situations where they can benefit from the early mover advantage. However, they also need to consider that an advantage can soon become a disadvantage due to susceptibility to TCD if they move too fast. While SMEs may need to enter foreign networks quickly to work with the best network partners, they also need to balance this with trust and commitment rather than blind trust. Entering foreign market networks too fast puts pressure on SMEs' existing resources and capabilities and may also result in entering unsuitable networks that hinder their capability development. SMEs need to be patient and allow time for incremental learning because entering the network too fast could mean that they cannot absorb all the new information quickly. Therefore, our main message to SME managers is to "move first to grow slow." Our results have several implications for public policy agencies. First, policy agencies could benefit from an alternative perspective regarding speed of international expansion by considering the speed of entering networks in foreign markets (depth) rather than the speed of entering a number of countries (diversity). Instead of considering the number of countries as a measure of success for the SME, policy makers need to consider the speed of network entry as a capability development process and how this enhances or impedes performance. Second, rather than vigorously promoting rapid growth, policy agencies could offer workshops and training courses for SMEs to understand the pros and cons of rapid entry into networks and the benefits of incremental growth. While several countries (e.g., Sweden, Finland, and New Zealand) encourage SMEs to grow through rapid internationalization, policy agencies need to consider the negative consequences of growing too fast by entering networks too quickly. Third, SMEs could be trained through workshops and seminars about the dynamics, timing, and speed of networking processes to become insiders in foreign networks, and the concepts of the early mover advantage and TCD. Such training would give SMEs the ability to recognize their readiness to enter foreign networks and the benefits of gradual accumulation of resources and capability development.
Funding Open access funding provided by University of Gothenburg.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/ licenses/by/4.0/.