The Impact Of Market-Related Factors On The Choice Of Foreign Market Entry Mode By Service Firms

The topic of internationalization has received an increased amount of attention due to globalization and growing amounts of international trade. One of the most important factors for the success of foreign market operations is the choice of the entry mode. This paper investigates the key determinants for the choice of foreign market entry mode. Specifically, the research examines the effects of market-related factors of the selected entry mode on service companies. This study contributes to the existing knowledge of internationalization of service companies by analyzing market-related factors of entry modes. Moreover, it provides managerial implications that might be applied by companies and governmental agencies to promote investment and internationalization of local companies.


INTRODUCTION
One of the most important factors for the success of foreign market operationsbesides market selection -is the choice of the entry mode (Lee, 2010;Kennedy, 2005). The choice of foreign market entry mode is explained by several key theories such as the Internationalization Theory (Canabala & White, 2008), the Transaction Cost Theory (Erramilli & Rao, 1993), Dunning's eclectic paradigm (Dunning, 1993) and the resource-based theory (Barney, 1991). However, these studies lack focus on the internationalization of service companies.
Frequently, scholars apply industry specific research to describe the factors of the choice of the export mode in certain fields: engineering consultancy (Winsted & Patterson, 1998), hotel industry (Gannon & Johnson, 1997), financial services (Grosse, 1997). Most of the earlier authors focused more on the analysis of manufacturing firms rather than service firms, whose importance nowadays is increasing (Andersen, Ahmad, & Chan, 2014). Studies of entry modes have been mostly focused on multinational enterprises (Kumar & Subramaniam, 1997). Due to globalization, the relatively well protected environment of SME's domestic markets was influenced by increasing competition, as a result such companies had to internationalize to survive (Lee, 2010). This was followed by an increased focus from researchers analyzing SME entry modes to foreign markets.
A number of research papers have focused on the analysis of factors influencing internationalization: Market attractiveness was analyzed by Agarwal & Ramaswami (1992) also Nakos & Brouthers (2002), demand uncertainty by Capel (1992), Kogut (1991) and McGrath (1997), country risk by Brouthers, Brouthers, & Werner (2002) and López-Duarte & Vidal-Suárez (2010) and finally cultural similarity by Canabal & White (2008). Even though the above research explored specific factors, there was no systematic attempt to assess a series of these factors and how they impact the entry mode selection of a service firm. This paper is designed to investigate the relationship between market-related factors and the choice of foreign market entry mode in the context of service providers. This research could add to the theoretical knowledge on the subject by improving and verifying the previous research in the field. Moreover, it will include a dataset from Lithuania (examined sample in Eastern Europe), as much of the research on entry modes has been conducted by analyzing companies from the USA and other developed economies. Furthermore, most of the studies of the determinants of foreign market entry mode have focused on the international expansion of multinational corporations (Kumar & Subramaniam, 1997). SMEs have distinct characteristics from MNCs which might have an impact on the choice of entry mode (Knight, 2001;Nakos & Brouthers, 2002;Fillis, 2001). Moreover, most research on the choice of entry mode examines companies from the highly developed regions such as the USA and Western Europe and has focused less on Eastern European countries (Ekeledo & Sivakumar, 2004;Canabal & White, 2008).
The goal of this paper is to analyze the relation between the key external market-related factors (market attractiveness, demand uncertainty, country risk, legal restrictions and cultural similarity between home and host market) and the entry mode selection of Lithuanian service SMEs.
The research question of the paper is how market-related factors (market attractiveness, demand uncertainty, country risk, legal restrictions and cultural similarity between home and host market) influence the choice of the foreign market entry mode, depending on service firms' asset specificity, size and business experience. Specifically, this particular question is raised in the context of Lithuanian services SMEs.
The research objective is to conduct quantitative analysis on the influence of market-related factors on the choice of foreign market entry modes, depending on service firms' asset specificity, size and business experience.
Methodological approach. The quantitative method was chosen for the empirical research. Specifically, this method includes logistic regression analysis on the data gathered via an online survey platform, together with the required validity, normality, and correlation analysis on the data used in the study. A survey of 52 Lithuanian service SMEs that have expanded abroad between 2005-2014 was conducted.

LITERATURE REVIEW ON FOREIGN MARKET ENTRY MODES
Theories of internationalization and transnational business activities date back to the 1960s and 1970s when Hymer and Dunning presented their pioneering works (Dicken, 2011;Buckley, 2011). Sharma & Erramilli (2004) provide a classification using three major paradigms: market imperfection, behavioral paradigm and market failure paradigm. The above mentioned paradigms and modern approaches are summarized in Table 1.
The question of choosing the right entry mode has been an ongoing discussion since the early evolvement of international business studies. Sharma and Erramilli (2004, p.2) defined an entry mode as a "structural agreement that allows a firm to implement its product or service market strategy in a host country either by carrying out only the marketing operations (i.e., via export modes), or both production and marketing operations there by itself or in partnership with others (contractual modes, joint ventures, wholly owned operations)". Several studies have classified possible entry modes by popularity, hierarchy, ownership or control (Agarwal & Ramaswami, 1992;Kumar & Subramaniam, 1997;Canabal & White, 2008;Erramilli & Rao, 1993). For example, Pan and Tse (2000) suggest a classification which is based heavily on the equity factor by including the importance of control. The resulting four groups (export, contractual agreements, equity joint venture and wholly owned subsidiary) incorporate the most common types of entry modes. The overall challenge for each decision maker is to balance the need for control with the level of uncertainty in foreign markets and the desired commitment.
In general, high control entry modes are chosen by companies that need to build up personal relationships in foreign markets, execute on-site analysis and adapt to foreign customers and markets (Hastings & Perry, 2000). Moreover, high control modes are preferred by companies owning high value brands where the high control is necessary to secure its consistent application in international markets (Klein & Leffler, 1981). Low control modes are more likely to be opted for by companies which estimate that the market uncertainty is high (or not acceptable) and the amounts of investment demanded by the high control modes cannot be dedicated to the particular entry mode at such risk levels (Gatignon & Anderson, 1988). This would imply that small companies  (Andersen, Ahmad, & Chan, 2014) Paradigm Theory Explanation for foreign market entry Author of theory Market Imperfection Paradigm

Monopolistic advantage Theory
If market imperfection is high, FDI will be preferred. Otherwise, licensing is adopted. Hymer (1960) International product Life Cycle Theory In the early stage of product life cycle, exporting is preferred. In later stages, FDI is adopted. Vernon (1966) Behavioural Paradigm

Internationalization Theory
Due to market uncertainty, firms use sequential stages entry modes from indirect exporting to direct exporting and full ownership Johanson & Wiedersheim-Paul (1975) Networks Theory If network relations provide strong competitive advantages, FDI is preferred. Otherwise, low control modes are adopted. If home market has (L), exporting is preferred. If (L) resides in the host market, high (I) will result in FDI. Otherwise, licensing is adopted. Dunning (1977) Transaction Cost Theory (TC) If transaction-specificity of an asset is high, firms prefer high control modes. Otherwise, low control modes are adopted. Anderson & Gatignon (1986) Current Approaches

Resource-based View (RVB)
Firms with strong firm-specific resources prefer high control modes. Otherwise, low control modes are adopted. Wernerfelt (1984) Contingency Theory Choice of entry mode depends on the internal and external environmental factors. Inseparable service firms prefer franchising or WOS. Okoroafo (1990) would opt for low control modes in general. However, in practice it can be observed that some SMEs choose high control entry modes. For this reason, the present research aims to investigate market-related factors that influence a decision to choose a foreign market entry mode.

Internationalization of Service SMEs
Most of the earlier authors focused more on the analysis of manufacturing firms rather than services firms, the importance of which nowadays is increasing (Andersen, Ahmad, & Chan, 2014). An increasingly competitive environment is where service companies have to choose their entry modes, where technological advances alter options, impacted by government policy and complicated by the speed of changes. Simultaneously, increasing trade liberalization and the increasing demands of businesses and consumers stimulate the growth of internationally traded services.
The specifics of services and production companies vary in terms of business expansion abroad. Several authors (Agarwal & Ramaswami, 1992;Gatignon & Anderson, 1988;Kogut & Singh, 1988) have analyzed the entry modes for manufacturing companies. The question remains whether findings of their research can be applied to service companies as well. Some authors argue that the determinants of entry mode choice for manufacturing companies are generalizable to service firms without much modification (Agarwal & Ramaswami, 1992;Terpstra & Yu, 1988), while the other group state that those determinants need substantial modification when applied to services (Erramilli, 1990;Erramilli & Rao, 1993).
Moreover, the study of entry modes has been mostly focused on the multinational enterprises (Kumar & Subramaniam, 1997). Due to the ongoing Globalization, the relatively well protected environment of SME's domestic markets has been influenced by increasing competition, therefore a number of such companies had to internationalize to survive in the market (Lee, 2010). These factors resulted in an increased focus from researchers analyzing the entry modes into foreign markets. Another reason for this interest relates to the ever-growing importance of SMEs in the World's economy and especially in international trade. There is still a question of whether the internalization processes differ for SMEs and MNCs (Laufs & Schwens, 2014).
It is considered that SMEs face higher challenges when they have to cope with more barriers (e.g., government regulations such as local ownership requirements, local content specifications, financial limitations, limited staff, less availability of resources; differential taxes and tariffs; higher transaction costs; competitive pressure) in the process of internationalization and possess, or have access to, fewer resources (Knight, 2001;Nakos & Brouthers, 2002;Fillis, 2001). Despite such difficulties, SMEs have several characteristics that potentiate and simplify their internationalization processes, such as high adaptability, reduced response times, low bureaucracy, creativity, innovations, recognition of opportunities, focalization in niche markets, and a global vision (Fillis, 2001;Knight, 2001). In this study it is presumed that the specific characteristics of a service SME which allow higher flexibility to internationalize are related to asset specificity, size of the company, as well as business experience. Further, in the research the latter are set as control variables exploring how the market-related factors (market attractiveness, demand uncertainty, country risk, legal restrictions and cultural similarity between home and host market) influence the choice foreign market entry mode by service SMEs.

Determinants of Entry Mode
In this paper factors related to the host market as well as the distance between home and host markets are included in the research. The below discussed set of factors has not been specifically researched in the context of service companies in the SME category. Additionally, market characteristics have less generalizable implications as determinants of entry mode (Kennedy, 2005). Besides the market related factors, firm specific factors such as asset specificity, firm size and business experience are discussed. The latter factors are included as control variables later in the empirical research part.
Market Attractiveness. One of the most important factors that influence the choice of entry mode is the growth and the potential size of the market (Agarwal & Ramaswami, 1992;Nakos & Brouthers, 2002). The transaction cost theory predicts high control modes in more attractive markets since firms are able to get the benefits of economies of scales and potential expected outcomes are of a higher value (Agarwal & Ramaswami, 1992). In the context of service firms, Agarwal and Ramaswami (1992) examined equipment leasing firms and Terpstra and Yu (1988) focused on advertising agencies. Both studies confirmed the positive effect of market attractiveness on high control entry modes.
Demand Uncertainty. The market size and its potential are important for the choice of foreign market entry mode, however, the market size by itself does not guarantee success for a particular company. Companies put their effort on estimating the net present value of their investment when making strategic decisions (Folta & O'Brien, 2004). Naturally, authors acknowledge that it is not always realistic to calculate the future demand for the product or service in a new foreign market before taking investment actions (McGrath, 1997). Consequently, the higher the level of demand uncertainty in foreign markets exists, the more low control (low investment) entry modes are preferred by companies, -as the demand uncertainty reduces potential benefits of the investment and encourages to opt for low control entry modes (Capel, 1992;Kogut, 1991;McGrath, 1997). Brouthers (2008) proved this while analyzing firms of Central and Eastern Europe, and Taylor (1998) confirmed the importance of demand uncertainty for companies in Japan and USA.
Country risk. Host market risk is one of the most common factors in studies regarding selection of entry mode (Canabal & White, 2008). It is evident that the risk is a significant factor for the choice of entry mode (Brouthers, Brouthers, & Werner, 2002;López-Duarte & Vidal-Suárez, 2010). This paper takes into account the list of country risks applied by Brouthers & Brouthers (2003). Services are considered to be human capital intensive, and as proposed by Erramilli & Rao (1993), are perceived to be inseparable, which means that production and consumption happen at the same time.
Legal Restrictions. Legal restrictions regarding business operations and entry mode choice is another factor that defines host market and is considered to have an effect on the selection of market entry mode (Brouthers & Brouthers, 2003). Legal restrictions impose not only additional barriers for entry but increase future operational costs and risk (Brouthers & Brouthers, 2003). A number of authors (Brouthers & Brouthers, 2003;Anderson & Gatignon, 1986;Anderson & Coughlan, 1987) have showed the significance of this factor: the negative relationship between the level of legal restrictions and high control entry modes was confirmed.
Cultural Similarity. Cultural similarity (some authors use the term "cultural distance" instead) between the home and host countries is another widely applied factor in the research of foreign market entry mode, the relevance of which has been accepted (Canabal & White, 2008;Agarwal & Ramaswami, 1992;Erramilli & Rao, 1993). Mitra & Golder (2002) argued that the cultural similarity is considered to result in the need for various changes in conducting business in foreign market, for example, the adaptation of business processes or changes in marketing strategy.
Empirical tests of the cultural similarity factor show controversial results. Morschett et al. (2010) and Tihanyi et al. (2005) showed the positive impact of cultural similarity on high control entry modes. On the contrary, a list of empirical studies confirmed that companies opt for low control modes in culturally distant markets (Kogut & Singh, 1988;Anderson & Gatignon, 1986;Kim & Hwang, 1992). Additionally, this was proved with the sample of service companies by Erramilli & Rao (1993). Boeh and Beamish (2012) suggest that the dyad travel time (and not geographical distance) has an impact on location choice and entry mode.
Asset specificity. Asset specificity is another variable which is considered to be significant in the selection of foreign market entry in various studies on this topic (Erramilli & Rao, 1993). Asset specificity is predicted to have a positive effect on the choice of high control entry modes.
Asset specificity is defined by the incorporation level of specific (or even unique) human and capital investments in the provision and delivery of services (Erramilli & Rao, 1993). Naturally, if the asset specificity is high and services require more knowhow, skills and competences, then there is a need for sharing valuable information with partners in a foreign market. Consequently, according to the transactional cost theory, companies that have high asset specificity are more likely to opt for high control entry modes in foreign markets (Williamson, 1985).
Firm size. Empirical studies in literature define that larger firms are more likely to opt for high control entry modes (Agarwal & Ramaswami, 1992;Erramilli & Rao, 1993;Gomes-Casseres, 1990;Kwon & Konopa, 1993). The possible rationale for this relationship between firm size and entry mode type can be explained by the transaction cost theory and the factor of resources. Smaller companies possess lower level of resources, as a result, are less likely to take high commitment risks in foreign markets.
Business experience. Another factor commonly applied to the choice of foreign market entry is the international business experience of a company. In some studies it was confirmed that international business experience is positively related to the choice of high control entry modes (Agarwal & Ramaswami, 1992;Terpstra & Yu, 1988). Companies having a high level of international business experience are considered to have additional advantages (Agarwal & Ramaswami, 1992). Consequently, companies with business experience are likely to have fewer problems in business operations management. To be more effective in foreign markets, therefore, high control modes are more probable for the internationally experienced companies than for the ones with less international experience (Gatignon & Anderson, 1988).

Research Problem and Gaps in Literature
After conducting the research on literature regarding the topic of foreign market entry modes, a few gaps were identified: 1. Most of the studies of the determinants of foreign market entry mode were focused on the international expansion of multinational corporations (Kumar & Subramaniam, 1997). SMEs have distinct characteristics from MNCs which might have an impact on the choice on entry mode (Knight, 2001;Nakos & Brouthers, 2002;Fillis, 2001); 2. Most research on the choice of entry mode are examining companies from the highly developed regions such as the USA and are less tested in Eastern European countries (Ekeledo & Sivakumar, 2004;Canabal & White, 2008); 3. The set of market-related factors on entry mode seems to be little tested with a sample of service firms of SME category, while taking into account certain firm characteristics, such as asset specificity, firm size and business experience.

METHODOLOGY
Theoretical Framework. A conceptual model was constructed in order to evaluate the effect of market-related factors (market attractiveness, demand uncertainty, country risk, legal restrictions, and cultural similarity) as well as firm specific characteristics (asset specificity, firm size and business experience) on the choice of foreign market entry mode (see Figure 1). Hypotheses Development. The goal of the research is to test the hypotheses which are grounded on the insights of the literature review. The model includes 5 hypotheses where each presumes the importance of market-related factors on the choice of foreign market entry mode. H1: Host market attractiveness is significantly positively related to high control entry modes. H2: Demand uncertainty in the host market is significantly negatively related to high control entry modes. H3: The risk of the host market is significantly negatively related to high control entry modes. H4: Legal restrictions in the host market are significantly negatively related to high control entry modes. H5: Cultural similarity between host and home markets is significantly positively related to high control entry modes.

Research Design.
To implement the defined research aims and objectives, the quantitative method was chosen, which allows the testing of the proposed hypotheses in order to investigate the relationship between market-related factors and the choice of foreign market entry mode. According to Canabal & White (2008), the quantitative method is the most commonly applied method in entry mode researches. The list of authors that grounded their studies on this type of research includes Brouthers et al. (2002;2003;2008;2013), Erramilli & Rao (1990;1993) and others.
Data Collection and Sample. The sample includes 52 Lithuanian service SMEs that have made a decision on the selection of foreign market entry mode. Eventually, 58 questionnaires were filled out and processed. After the screening process, the answers with missing values, non-logical values or where a person filling out the questionnaire had indicated that one had not been directly involved in the decision on foreign market entry mode selection issues, were removed from the final data file.
Research is particular focused on companies satisfying three criteria: 1. The company falls under the category of micro-enterprise or small-medium-size enterprise; 2. The company provides services; 3. The company operates internationally and has made a decision for selecting foreign market entry mode in the period from 2005 to 2014. The primary data for this research was collected by using self-administrating online questionnaire platform. The questionnaire was constructed of four sections: 1 3. Entry mode selection factors. This section includes 16 Likert scale questions regarding the market-related entry mode factors such as market attractiveness, demand uncertainty, country risk, legal restrictions and cultural similarity. 4. Final remarks. The final section is built of two questions that define the job position of the person completing the questionnaire and the clarification whether the respondent had taken part in the entry mode selection process. Both questions are designed to test the general validity of the responses for the survey.

Measures
This research encompasses the type of entry mode as a dependent variable, and foreign market factors as independent variables. In this section, the list of dependent, independent and control variables together with their measures are described. Dependent Variable. In this study, the dependent variable type of entry mode is dichotomous (includes values 0 or 1). The first group of entry modes (value equal to 0) consists of high control entry modes such as an office branch, wholly-owned subsidiary and joint ventures (company owning equal or majority equity share). The second group (value equal to 1) includes low control entry modes such as joint venture (a company owning minority equity share), alliances, direct-to-customer export, agent / distributor export and licensing / franchising. This method is adopted from Erramilli & Rao (1993).
Consequently, the companies were asked to identify the entry mode which they had chosen at the time of market entry from a full list. The answers were divided into two groups of entry modes: high or low control.

Independent Variables
Market attractiveness. In this research, market attractiveness is defined by two characteristics: the size of the market and market growth. Two 5-point Likert scale questions (where 5 stands for strongly agree; and 1 -for strongly disagree) (Cronbach alpha -0.771) were applied from Taylor (1998), and respondents were asked to express their agreement level on the potential market size and the growth of foreign market.
Demand uncertainty. In order to measure demand uncertainty, a single-question measurement was adopted from Taylor (1998). Respondents were asked to identify their perceived evaluation of demand uncertainty in foreign market on a 5-point Likert scale (where 5 stands for strongly agree; 1 -for strongly disagree).
Country risk. Country risk was assessed by using four Likert scale questions that assessed managerial perception of risks in the host country such as economic stability, political stability, repatriation of profits, expropriation of company (Agarwal & Ramaswami, 1992;Brouthers & Brouthers, 2003;Brouthers K. D., 2008). In the research of Agarwal & Ramaswami (1992) the reliability coefficient was 0.9, while Brouthers & Brouthers (2003) and Brouthers (2008) found the coefficient to be 0.67 and 0.73 accordingly. Questions asked to identify the importance of the risk (where 5 stands for a significant risk; 1 is not considered to be a risk at all).
Legal restrictions. The factor of legal restrictions was measured by using a single Likert type questionnaire adopted from Brouthers & Brouthers (2003) and Anderson and Coughlan (1987). The 7-point question asked to evaluate the level of legal restrictions at the time of foreign market entry (where 7 stands for many restrictions; 1 -for no legal restrictions).
Cultural similarity. In order to measure cultural similarity, a set of four Likerttype questions were adapted from Taylor (1998). Questions were constructed to ask respondents to evaluate their level of agreement to the provided statements which included the similarity of value systems, norms, practiced customs and general cultural similarity between the home and host countries (Cronbach alpha -0.816) on a 5-point scale (where 5 stands for strongly agree; 1 for strongly disagree).

Control Variables
The firm specific characteristics that were discussed in literature review are included as control variables, namely, asset specificity, firm size, and business experience.
Asset specificity is measured by three Likert-type questions that asked the respondents to evaluate the specificity of their training programs (indicator of human capital know-how), ability to introduce new services (indicator of service uniqueness) and the level of technological resources that companies have to handle for the foreign market entry. The reliability coefficient (Cronbach alpha) for this measure is found to be 0.7 (Brouthers, 2008). All three questions are constructed as a 7-point Likert scale (where 7 stands for the characteristic to be well above average; 1 -well below the average).
The firm size was measured by the total number of employees in the company. Similar measuring was applied by other scholars (Brouthers, Brouthers, & Werner, 2002;Erramilli & Rao, 1993;Gatignon & Anderson, 1988).
International business experience was measured by the number of years of international business operations (conducting business in any other than home country).
Companies were asked to identify the number of years in an open question adapted from Brouthers, Brouthers, & Werner (2002).

Data Analysis
Raw data was checked for any missing values as some questions in the questionnaire were not mandatory in the online survey. Validity for each set of items for each construct was checked by evaluating the Cronbach alphas. Additionally, the correlation of all variables was evaluated. The hypotheses were tested by building a binary logistic regression model.

EMPIRICAL RESEARCH
Companies that answered the questions were divided into 8 separate service industry groups (see Table 2). This distribution of groups is similar to the samples of other authors that analyze service companies on the topic of the selection of the market entry mode. The data indicates that 31 companies opted for low control and 21 chose high control entry mode (see Table 3). The distribution is acceptable (no clear focus on a single entry mode type). Therefore, the dichotomous dependent variable can further be tested in the empirical research.

Reliability.
Cronbach alphas were calculated for the independent variables in order to test the reliability of data. This method helps to evaluate whether the questions put in a set are internally consistent and can be used to explain a measure. The Cronbach alpha coefficient exceeding 0.7 is considered to be acceptable for further research steps. The calculated Cronbach alpha coefficients for measures of asset specificity, cultural similarity, country risk and market attractiveness indicate that the questions are correctly structured -each set of items generates a coefficient higher than 0.7 (see Table  4). Consequently, the questions for the measures of asset specificity, cultural similarity, country risk and market attractiveness remain unchanged and can be used for further research.
Descriptive Statistics. The normality tests (Kolmogorov-Smirnov and Shapiro-Wilk) were conducted for the data of all independent and control variables. To confirm normal distribution, significance coefficient should exceed 0.05. The results have shown that generally none of the variables (firm size, business experience, demand uncertainty, legal restrictions, asset specificity, cultural similarity, market attractiveness and country risk) comply with a normal distribution. Only cultural similarity shows controversial results according to the Kolmogorov-Smirnov and Shapiro-Wilk tests. To sum up, variables in the model do not fall under the normal distribution. However, further analysis can be performed despite having datasets that do not satisfy normal distribution. Consequently, the logistic regression is a chosen analytical method, which does not require a normal distribution of data. The full picture of descriptive statistics is provided in the table below (See Table 5).

HYPOTHESES TESTING: LOGISTIC REGRESSION
The logistic regression model was constructed by including all the independent variables (market attractiveness, demand uncertainty, legal restrictions, country risk and cultural similarity) and control variables (asset specificity, firm size, and business experience). Then, each variable was eliminated (one at a time) which proved to be insignificant by the output of models. Variables were not significant and were eliminated in the following order: firm size, country risk, demand uncertainty, asset specificity and, finally, business experience. The final model was constructed by including the following variables: The final regression model was found to be statistically significant explanation of the choice of entry mode. The p value is equal to 0.01364, which shows the model to be significant. Moreover, the classification table supports the validity of the model (see Table  6). The model correctly classified 75 % of entry mode choices by companies, while the chance rate was able to explain approximately 59 % of total cases. Lastly, the general model deviance is 1.239182, which is considered to be acceptable and the results of the model can be further analyzed. During the process of final model construction, a number of variables were found to be insignificant. Consequently, all the three control variables (firm size, business experience and asset specificity) and two out of five independent variables (demand uncertainty, country risk) were eliminated since they produced no statistical significance. The elimination of independent variables from the model resulted in the rejection of the second hypothesis (H2: Demand uncertainty in the host market is significantly negatively related to high control entry modes) and the third hypothesis (H3: The risk of the host market is significantly negatively related to high control entry modes). However, both demand uncertainty and country risk had negative coefficient signs (as predicted in the hypotheses). The results for each variable of the final regression model are presented in Table 7. The first hypothesis (H1) suggests that host market attractiveness (market size and market growth) is positively related to high control entry modes. The findings of the research supported this hypothesis: the variable of market attractiveness was found to be significant (p = 0.069 < 0.1) and the coefficient sign was positive (coefficient = 0.736).
The fourth hypothesis (H4) stated that legal restrictions in the host market are negatively related to high control entry modes. Although the empirical results provided the predicted sign of the coefficient (coefficient = -0.479), the hypothesis could not be confirmed due to the lack of significance of the variable of legal restrictions (p = 0.151 > 0.1). However, the significance level is close to the 0.1 limit.
The statistical analysis showed that cultural similarity is significant (p = 0.067 < 0.1) and has a positive effect on the choice of high control entry modes as it was predicted in the model (coefficient = 1.176). These results support the fifth Hypothesis (H5), which states that cultural similarity between host and home markets is positively related to high control entry modes.

FINDINGS AND DISCUSSION
The findings of this paper indicate that the two market-related factors, namely market attractiveness and cultural similarity between home and host market are significant when selecting the foreign market entry modes by service firms.
First, market attractiveness is measured by how the potential market size and market growth influences Lithuanian service companies to opt for high control entry modes. The results might be linked to the arguments of the transaction cost theory which indicate that companies are able to reach economies of scale and are more likely to take higher risk by choosing more investment-intense entry modes for bigger future benefits (choose long-term investment strategy). The findings of this study support the results of other research in the field (Agarwal & Ramaswami, 1992;Taylor, Zou, & Osland, 1998). Moreover, results match the findings of Nakos & Brouthers (2002), which showed the importance of market attractiveness in the context of SME.
A second variable which was proved to be important in the entry mode selection process is cultural similarity. The research findings indicate that service companies tend to choose high control entry modes when they enter more culturally similar foreign markets. This effect can be explained by the argument that similar markets make it easier for companies to acquire information, understand the market and to estimate the required inputs and outputs. As a result, companies in more culturally similar markets have lower level of perceived risk and are more likely to choose high control entry mode. The results of this study go in line with the findings of other researchers (Kogut & Singh, 1988;Anderson & Gatignon, 1986;Kim & Hwang, 1992). Additionally, Erramilli & Rao (1993) proved cultural similarity to have the same effect in the context of service firms. However, the results of this paper contradict with the findings of Morschett et al. (2010) and Tihanyi et al. (2005) who showed opposite results: Companies go for high control modes in order to deal with the risk due to the cultural similarity. Consequently, it is important to notice that the findings of this paper support the ones in the field of service companies.
In general, there was a lack of empirical evidence to support the importance of demand uncertainty, country risk and legal restrictions. The logistic regression model failed to show a significant link between these factors and the choice of foreign market entry mode.

LIMITATIONS AND IMPLICATIONS FOR FURTHER RESEARCH
This research has several limitations that need to be mentioned. First, there is an issue of non-representative sample size. As it is noticed by Kennedy (2005), this is a common problem faced by many researches in the field, for example, Erramilli and Rao (1993) acknowledged the same problem. The reason is that it is challenging to collect a large amount of specific data from companies that generate low response rates. Secondly, this study does not take into account all factors that are considered to be important in the literature. This challenges the validity of the model as other important variables in the selection of foreign market entry mode might influence the empirical results. Thirdly, there is a potential problem that the answers regarding the entry mode are provided by a single person from a company. Answers from a single respondent might include personal bias regarding the company`s selection of entry mode. Finally, data is gathered from companies in Lithuania. To make more generalizable findings for the region of Eastern Europe, dataset from more countries should be included in the research. Besides, further research could focus on longitudinal effects of entry mode selection by service SMEs.