Keywords

1 Introduction

In February 2011, two major U.S. retailers Best Buy Inc. and Home Depot announced divesture from the Chinese market. While certain chains, such as Kentucky Fried Chicken or Tesco are for the most part thriving in emerging markets such as China, Thailand or South Korea, others are struggling to achieve profitability. In spite of substantial investments, extensive marketing research, prestigious locations and lengthy employee trainings, many large foreign retailers are still facing intense difficulties in emerging markets. Nevertheless, these markets keep attracting international chains ranging from hypermarkets to luxury specialty stores or quick service restaurants.

Even though international retailing has inspired a growing number of articles (Dawson 1994; Robinson and Clarke-Hill 1995; Reijnders and Verhallen 1996; Dawson 2001; Burt et al. 2002; Palmer and Owens 2006; Picot-Coupey 2006, 2009; Gielens and Dekimpe 2007; Suh and Howard 2009), specific retail internationalization theories are still in their infancy (Burt et al. 2002). Therefore researchers have heavily borrowed from the manufacturing literature whose validity may be questionable in the field of retailing. Additionally, although Hollander (1970) already stressed the importance of the topic, a few but growing numbers of authors have published articles on retail failures (Vida and Fairhurst 1998; Alexander and Quinn 2002; Burt et al. 2002; Burt et al. 2003; Burt et al. 2004; Costil 2003; Wrigley and Currah 2003; Dupuis and Fournioux 2006; Etgar and Rachman-Moore 2007; Hang 2009). Moreover Hang (2009) argues that a consensus over the definition and the conceptualization of retail failure has not been reached (Gielens and Dekimpe 2001; Alexander and Quinn 2002; Burt et al. 2002, 2003). This reflects the need for specific frameworks dedicated to this facet of international retailing.

Moreover most current research on international retail failure has focused on developed countries (Burt et al. 2003; Palmer 2004) in spite of the substantial attraction of emerging markets on foreign retailers. This article aims to contribute to the under-researched problem of international retail failures in emerging markets by focusing on the following factors that have been more or less investigated in the literature as potential antecedents of failure or success: market entry mode choice, chain size, relative entropy, store density, financial soundness, cultural and psychic distance and market attractiveness. The purpose of this article is to start developing a set of exploratory guidelines specific to international retailers in emerging markets.

This paper will first review the theoretical background on the topic and articulate a set of hypotheses that will be tested in a subsequent section. Archival data and store failure narratives obtained through journal articles helped in developing the empirical research. These findings will be followed by a section dedicated to managerial and strategic implications.

2 Theoretical Backgrounds

2.1 Definition of International Retail Failure

Burt et al. (2003) argue that the knowledge obtained from the analysis of international retail failures was instrumental in the success of current market leaders such as Carrefour or Tesco. But for obvious reasons retailers do not emphasize nor publicize much of these failed attempts and focus instead on their success stories. It is therefore cumbersome to collect unbiased material on the topic and that may explain the scarcity of the research. Additionally as stated by Burt et al. (2002) the definition of the concept of retail failure is rather poor in the literature. A few international retailing researchers (Burt et al. 2003; Alexander et al. 2005) have therefore attempted to identify the various facets of divestment activities.

In essence, divestment does not necessarily entail country exit but could take multiple other forms such as store closure, sales of the chain in that region, termination of business agreements designing governance modes such as alliances, franchise contracts, licensing or joint ventures (Baroncelli and Manaresi 1997), and organizational restructuring. However, Burt et al. (2003) warn that the metrics used to measure failure could be misleading: for example store closures in a certain region may contribute to the success of the overall chain by allowing more profitable investments in other geographies with better strategic prospects for the organization; similarly, the takeover of certain foreign operations by a rival may also be highly profitable for a firm and deemed a financial success. But what is driving failure or success in a foreign market entry?

2.2 Antecedents of Failure and Success

Burt et al. (2003) summarizing the work of Benito (1997) list four potential causes to divestment: (1) market failures or issues with the economic, social and political environment, such as country openness and country risks; (2) competitive failure due to stronger competitors, and lack of firm differentiation; (3) operational failure where under-performance may result from the lack of skill to transfer domestic experience into foreign markets; and (4) business failure where a retailer may face difficulties in its home market and may lack governance competencies and financial means to maintain international operations.

2.2.1 Market Entry Decisions and Governance

Research has often linked exit decisions in international retailing to market entry modes (Li 1995; Chang and Singh 1999; Gielens and Dekimpe 2001). According to the literature (Li 1995; Burt et al. 2003) international retail joint ventures (IRJV) are particularly risky, principally when they take place between culturally distant partners. Palmer (2004) emphasizes the fact that they are multiple forms of IRJVs. He urges researchers to distinguish between equity joint ventures (EJV) and contractual joint ventures (CJV). Among EJVs, acquisition of an existing structure versus the creation of a greenfield entity may also have different outcomes. Alliances such as licensing or franchising are deemed safer than IRJVs (Burt et al. 2003). However the literature (Li 1995; Chang and Singh 1999; Gielens and Dekimpe 2001) recommends full ownership, through acquisition or greenfield investment as a preferred mode of entry. Additionally, Benito (1997) stresses that greenfield wholly owned subsidiaries (WOS) have a lower propensity to fail than acquired structures.

2.2.2 Other Antecedents of Failure and Success

Several authors have chosen to concentrate their efforts on other antecedents of failure or success: for example Couturier and Sola (2010) have investigated the impact of regulations and concluded that a country with few legal and environmental constraints that allows investors to freely choose their mode of entry is very favorable to success. The impact of cultural and psychic distance between the domestic and the host country is also actively researched (Sousa and Bradley 2005; Chen et al. 2009). Chen et al. (2009) believe that large cultural distance discourages firms to select joint-venture or fully owned acquisitions as their mode of entry, encouraging instead more flexible arrangements that allow for faster withdrawal in case of failure.

Domestic issues can also be potential antecedents for failure or success. According to Alexander and Quinn (2002) the reasons for divesture may reside in negative circumstances in the retailer’s domestic market that may call for a portfolio adjustment. Additionally, it is interesting to consult the survival literature. One of its key theories is that market entry timing plays a significant role in the longevity of a firm in a foreign market (Golder and Tellis 1993; Pan and Chi 1999; Cui and Lui 2005; Johnson and Tellis 2008).

Retailers’ characteristics in terms of industry, age and size may also influence the outcome. Burt et al. (2003) believe that the older and larger a corporation is the lower its risk to fail in a foreign market entry. Last, Bradach (1997) outlines the first challenge for a chain: its ability to add new units and Cliquet (1998) explains spatial strategies “as the need for growth by addition of new stores”. Hence as the notion of territory coverage appears to be a valid measurement of retailer’s success domestically spatial strategies can be a potential antecedent of failure or success in international retailing. Relative entropy will assess territory coverage (Cliquet 1998, p. 219) in this research in order to measure its impact on retailers’ propensity to fail.

2.3 Emerging Markets Specificities: The Case of China

According to Beamish (1985) joint-ventures in emerging countries are problematic endeavors characterized by higher instability rate and managerial dissatisfaction than developed countries’ ventures. Beamish stresses the need to consider these differences for governance purposes. According to Chen et al. (2009) firms entering transitional economies such as China should focus on reducing institutional risks instead of maximizing efficiency. Therefore the authors believe that cultural and institutional perspectives such as rules and norms developed by the government but also by society are more relevant in assessing success and failure than the traditional transaction cost theory.

Due to its size and the structure of its central government that empowers local governmental entities to make economic decisions (Lubman 1999), Chinese institutional laws along with cultural norms vary greatly from region to region. Liu (2008) partly attributes the success of Kentucky Fried Chicken (KFC) in China to the ability of their joint-venture partner in Shanghai to understand Chinese politics and to establish a strong relationship with the Chinese government including provincial, city and local government units. In 1992, China authorized foreign retailers to enter the Chinese market through IRJVs. Foreign equity was initially limited to 50 % in 1992 and increased to 65 % in 2002. This restriction was completely removed when China joined the world trade organization (WTO) in 2004 which benefited to Wal-Mart and Carrefour (Wang 2009). According to the National Statistics Bureau of China, newly appointed foreign retailers shifted their market entry strategy from the mandatory IRJV model to wholly owned foreign enterprise (WFOE) when the government lifted the investment restrictions in 2004.

3 Methodology

3.1 Measurements and Development of Hypotheses

Considering the current literature (see Table 1) and the availability of data the following factors were selected for this research: company size, financial performance (global level), market entry mode choice such as franchising, licensing, IRJV(minority or majority), greenfield investment or acquisition, time (order of entry, operation duration in the country), territorial coverage (entropy, store density), brand penetration in the host country, cultural and psychic distance with the host country, and country attractiveness (regulatory, political, economic environment in the host country).

Table 1 Review of the variables selected for the empirical assessment
Full size table

3.1.1 Outcome of Market Entry

The dependent variable used in the regression and as a grouping factor in the independent sample t-test refers to the outcome of market entry, failure or success. The outcome variable was defined as follow: 0-failure and 1-success. For the purpose of this article, failure was equated to country exit for a specific chain and success with its continuing presence in the country at year seven. Consistent with the results obtained in previous studies (Gandolfi and Strach 2009), the original statistical analysis of the 77 cases of failure in this research indicated that the average length of operation between market entry and country exit was 7 years and justifies therefore the following method of data collection: (1) at divestment time for the 77 cases of failure; and (2) at year seven of operation counting from the market entry year for the 35 cases of success. This procedure was particularly essential to follow in order to obtain meaningful results in terms of territorial coverage, and brand penetration.

3.1.2 Company Size

Firm’s size may influence international expansion success for several reasons: (1) access to more resources, (2) wealth of knowledge and experience, and (3) ability to sustain losses (Parsons 1996; Luo 2002; Johnson and Tellis 2008). However the literature debates vigorously about this topic. Johnson and Tellis (2008) for example mentioned the failure of Wal-Mart in South Korea and conclude that a large company size may facilitate success but is not a guarantee of success. Burt et al. (2003) argue that older and larger retailers have less risk to fail when expanding internationally. The following hypothesis was therefore established:

Hypothesis 1 (H1)

Larger retailers have lower risks to fail.

Using the method described by Johnson and Tellis (2008), the company size measurements were obtained from retailers’ annual reports. Annual revenue at time of divesture or on the seventh year of activity in the host country was converted into U.S. dollars and transformed into logarithms.

3.1.3 Financial Performance

According to Alexander and Quinn (2002) worsening financial performance is a key trigger for divestment activities in international retailing. This is particularly true for public companies that are facing market pressure for short-term profitability. Hypothesis two below reflects this observation:

Hypothesis 2 (H2)

Retailers with strong financial performance have lower risks to fail.

Financial strength was assessed using the Moody’s rating developed for large corporations. The 18 levels of ratings were coded the following way: Aaa, the highest rating for a firm was computed as 18, while the lowest rating of Caa-C was computed as 1 and withdrawn rating computed as 0.

3.1.4 Territorial Coverage: Entropy and Store Density

Territorial coverage has received minimal consideration in international market entry studies. However one may expect that a retailer is more likely to maintain operations in a given country if it reaches a level of territorial coverage that is relatively consistent with its other foreign locations. According to Cliquet (1998) growth through addition of new units is essential to the success of a retailer. One of the key measurements of spatial coverage is relative entropy. The following hypothesis was therefore formulated:

Hypothesis 3 (H3)

It is proposed that higher relative entropy in a host country lowers the risk of failure in that specific country

Relative entropy is typically calculated the following way (Cliquet 1998):

$$ \mathrm{ E}=-\sum\limits_{{\mathrm{ i}=1}}^{\mathrm{ k}} {{{\mathrm{ f}}_{\mathrm{ i}}}\;\log{{\mathrm{ f}}_{\mathrm{ i}}}} $$

Where

  • E = entropy

  • k = number of levels (1 level in our case. The host country)

  • fi = frequency of stores (for this specific chain) in area I (fi = ni/N)

  • ni = number of stores in area I (the host country in our case)

  • N = total number of stores for the chain (worldwide in our case)

Store density, (per type of retailer) in the host country is a good complement to the entropy measurement since it allows the assessment of the degree of retail development and the level of competitive pressure in the host country. According to the “organizational ecologists” defined by Mellahi and Wilkinson (2004) as scholars who explain retail failure through external factors such as density, the higher the store density in a given retail category the higher the competition and the risk of failure (Pal et al. 2007). Hence hypothesis four was derived.

Hypothesis 4 (H4)

Higher store density in the host country is positively correlated with failure.

Store density is measured this way: a ratio between the total number of hypermarkets, the only format considered in this research due to dataset limitations, at time of divestment (or 7 years after market entry for successful retailers) and the total population (in 100,000) of the host country that specific year and expressed as a density index. For example, an index of 2 means that there are two hypermarkets for 1 million inhabitants.

3.1.5 Brand Penetration

According to Arkolakis (2010) there is a strong negative relationship between overall brand penetration in a specific territory and relative marketing costs. This has a significant impact on the bottom line. This observation is expressed in hypothesis 5:

Hypothesis 5 (H5)

Higher brand penetration in the host country is negatively correlated with failure.

The following formula calculates brand penetration rate in the host countries for each hypermarket chain:

$$ {{\mathrm{ f}}_{\mathrm{ i}}}={{\mathrm{ n}}_{\mathrm{ i}}}/\mathrm{ N} $$

where:

fi = frequency of stores (for a specific chain) in area I (host country in our case)

ni = number of stores (for a specific chain) in area I (hypermarkets in this research)

N = total number of similar stores in area I (Hypermarkets in this paper)

3.1.6 Cultural and Psychic Distance

Authors frequently use the measurement of cultural and psychic distance in their research on foreign market entry (Kogut and Singh 1988; Wilcox and O’Callaghan 1999; Mitra and Gloder 2002; Huang and Sternquist 2007; Johnson and Tellis 2008; Chen et al. 2009). Hypothesis 6 reflects the general consensus of these authors:

Hypothesis 6 (H6)

The stronger the cultural distance between the domestic and the host country the higher the risk of failure.

This variable utilizes the four dimensions of the Hofstede scores for both domestic and host countries by following the method developed by Kogut and Singh (1988):

$$ \mathop{CD}\nolimits_{smt }=\sum\limits_{j=1}^4 {\sqrt{{{{{(\mathop{D}\nolimits_{jst }-\mathop{D}\nolimits_{jmt })}}^2}}}} $$

Where “CDsmt represents the country distance score between host country s and home country m in year t, Djst is the score on dimension j for host country s, and Djmt is the score on dimension j for home country m both measured in year t” (Johnson and Tellis 2008).

3.1.7 Country Attractiveness Assessment

Restrictive regulation, political risks and economic issues may lead to failure (Huang and Sternquist 2007; Myers and Alexander 2007; Chen et al. 2009). Multiple indexes are available. The market potential index (MPI) developed by Michigan State University has been thoroughly tested through academic research (Cavusgil 1997, 2004). The MPI was specifically designed for emerging countries and covers most of the time period that pertains to the dataset. The index uses the list of emerging countries annually published by the Economist as a frame for its analysis and measures market attractiveness according to seven dimensions: (1) market size, (2) market growth rate, (3) market intensity such as consumption expenditures, (4) market consumption capacity, (5) commercial infrastructure, (6) economic freedom such as trade policy, regulatory climate, taxation policy, (7) market receptivity to imports, transforming raw results into standardized data in order to compare to each other. The MPI values rank from 1 to 100 (100 being the perfect score).

The impact of the market attractiveness will be tested in hypothesis 7:

Hypothesis 7 (H7)

The host country overall attractiveness is negatively correlated with failure.

3.1.8 Market Entry Mode Choice

The literature has often researched market entry mode as a potential antecedent of failure or success in market expansion. Palmer and Owens (2006) have focused on the IRJVs while other authors have chosen to investigate international store acquisition or franchising (Quinn 1998; Quinn and Doherty 2000; Quinn and Alexander 2002).

This research investigates the influence of the following modes of entry on the outcome of failure or success: (1) franchising or licensing; (2) IRJV without a controlling position; (3) IRJV with controlling position; (4) acquisition of a wholly owned subsidiary; and (5) greenfield organic growth.

There is no consensus in the literature in regard to the risk of one entry mode choice versus the other. While Li (1995) and Burt et al. (2003) consider that IRJVs present the highest risk, Benito (1997), along with Barkema et al. (1996) and Hennart et al. (1998) claim that acquisitions have the highest propensity to fail. Therefore hypothesis 8 will test the nature of this relationship.

Hypothesis 8 (H8)

Market entry mode choice has a significant impact on the market entry outcome of success or failure.

3.1.9 Entry Timing and Order of Entry

The debate is raging between proponents of early market entry and late market entry: Pan and Chi (1999): argue that early entrants receive multiple benefits such as government incentives that are not available to late entrants. On the opposite Golder and Tellis (1993) believe that early entrants may face a long learning curve in emerging countries and Cui and Lui (2005) obtained mixed results in their research on China. Hypothesis 9 was therefore derived:

Hypothesis 9 (H9)

The order of entry into the host country influences the outcome of success or failure.

Due to the limited size of the present sample, it was not possible to test this variable in a quantitative manner, hence the choice of a qualitative approach consisting in selecting and analyzing the cases of a few retailers. Results are presented in Table 6.

3.2 Data

This study uses a set of data compiled from multiple sources such as annual reports, statistical offices, and other reliable governmental offices such as the USDA as well as ratings and ranking tools such as the Hofstede’s (1991) four cultural dimensions, Moody’s credit scores and the Market Potential Index (MPI) developed by Michigan State University.

The 112 cases collected for the empirical research pertained to emerging countries and represented various sectors of retailing with an over-representation of food retailers (see descriptive properties of the data in Table 2). The sample comprised 77 cases of failure and 35 cases of success, spanning 37 years from 1974 to 2011. It is however important to keep in mind that this database includes missing or incomplete data. For example the Moody’s rating was only available for 68 out of the 112 cases. Most retailers in this sample originated in Europe, and Asia was one of their main destinations for market expansion. A majority used joint-venture as their original mode of entry and this is certainly partly due to the restrictive regulation on full ownership in certain countries such as China. One may also note that most retailers acquired their joint-venture partner or obtained a majority control position after several years of operation in the host country.

Table 2 Descriptive statistics
Full size table

4 Findings

Using independent sample t-tests for each selected variable the authors investigated differences between failed and successful market entries (see Table 3). Significant results were obtained in terms of relative entropy score, density, brand penetration and cultural distance. Calculated for descriptive purposes, the t-test on the operation duration variable revealed significant differences that are linked with the research procedure adopted in this paper. As an average “failed” retailers spent 7 years in the host country before divesting. The other retailers in the sample have been operating in their host country for an average of 14.7 years.

Table 3 t-Test results differences between failures and successes
Full size table

Additionally, as indicated in Table 4, Pearsons correlation tests yielded significant results: as expected there are significant relationships between cultural distance and failure or success; high cultural distance is negatively correlated with success. There is also a positive correlation between relative entropy and success. In addition, as anticipated a higher level of relative entropy occurs in successful chains and store density appears to be negatively correlated with success. Last, country attractiveness seems to be positively correlated with success.

Table 4 Correlation results
Full size table

The correlation between company size and the financial performance scores is indirectly relevant for this research. It links the company size with a higher Moody’s score that could also connect to market entry success by helping the company to obtain additional funding. It is also interesting to note the negative correlation between relative entropy and company size: the larger the company, the lower the relative entropy level.

Last, a binary logistic regression help to understand the antecedents of market expansion failures.

$$ Y=\alpha +{\beta_1}{X_1}+{\beta_2}{X_2}+{\beta_3}{X_3}+{\beta_4}{X_4}+{\beta_5}{X_5} $$

Where:

  • Y = Outcome (failure or success) where failure = 0, success =1.

  • α = Constant of the equation.

  • β = Coefficient of the predictor variables.

  • X 1  = Company size

  • X 2  = Financial performance

  • X 3  = Market entry mode choice where licensing/franchising = 1, IRJV (no majority) = 2, IRJV (majority) = 3, acquisition (full ownership) = 4, and greenfield (full ownership) = 5.

  • X 4  = Territorial coverage (relative entropy)

  • X 5  = Cultural and psychic distance (with the host country)

Results are presented in Table 5.

Table 5 Regression results
Full size table

The full model displays a reasonable explanatory power of 71.7 and a pseudo R-square of 0.220. Although not significant, the results show a positive relationship between financial performance and success. Significant results include a positive relationship between entropy level and success, and a negative one with high-control market entry modes such as acquisitions.

Due to the number of missing cases the following variables were not considered in the full regression model: store density (hypermarkets) and country attractiveness. Separate logistic regressions were conducted for both variables and yielded significant result (p < 0.1) for store density whose estimated coefficient of −0.400 shows a negative relationship with the market entry outcome.

However there is no evidence of potential impact of the host country attractiveness (MPI score) on propensity to fail or succeed.

Considering the limited size of the sample, one may be cautious in drawing conclusions despite nevertheless significant statistical results.

H1 was not confirmed, there were no statistically significant differences between larger and smaller companies in terms of failure or success. However one may still note that successful retailers seem slightly larger than their counterparts.

H2 was not confirmed. Both the t-test and the regression results were not statistically significant. H3 was confirmed through the t-test, the Pearsons correlation test and the regression. It appears that the chains with the highest rate of relative entropy for one specific country are the most successful: for example when Tesco divested from Taiwan its entropy level was very low compared to the ones in Thailand and in South Korea after 7 years of operation. Hypothesis 4 was also supported. Store density (H4) varies dramatically between the two groups of retailers (“failed” or “successful”): as depicted through the t-test result, in average the store density was two times higher in the host countries where retailers divested. Similarly the Pearsons correlation test identified a negative correlation between store density and success. That may be due to the fact that successful retailers were often early entrants and therefore even after 7 years of activity, store density, and therefore competition were still fairly low in the host country. The results obtained in this research, in term of brand penetration (H5) are exploratory, due to the limited number of cases that were considered (n = 36) and the fact that only hypermarkets were included in the calculation. It appears however, as indicated in the t-test results that successful chains have a much higher penetration rate that their counterparts. Considering this observation in light of the previous comment on store density one may hypothesize that successful retailers may have benefited of a low store density that allowed them to establish themselves as market leaders and gain substantial spatial coverage and market-share. H6 was also supported, through the t-test, the Pearsons correlation but not through the logistic regression and indicated a negative correlation between high cultural distance and success. Although the t-test, measuring the differences according to the MPI score (H7) was not statistically significant it is counter-intuitive to realize that the MPI score was actually lower for the successful market entries than for the failures. However the correlation results presented in Table 4 indicated an opposite result that seems to indicate a potential impact of market attractiveness on the outcome. The cross-tabulation between failure and success and the market entry mode choices (H8) did not yield significant results but the logistic regression indicated a negative relationship between high-control market entry modes such as full ownership and success. It is important to note however that this research took into consideration the original mode of entry and not the governance mode in effect at divestment time. Most organizations seem to move from an IRJV model to full ownership as soon as the country’s regulations soften. This may have impacted the results.

Analyzing a few select cases allowed the authors to assess the impact of order entry on failure or success (H9). It is of course preposterous to reach definite conclusions with this limited number of observations but it appeared that in seven cases out of eight, early entrants were rewarded with success. It is however interesting to note that a few chains such as Casino in Brazil and Thailand and Auchan in China seem to have benefited from their late entrance. Detailed results are available in Table 6.

Table 6 Order entry as driver of failure or success
Full size table

5 Discussion, Managerial Implications and Research Limitations

This article contributes both to the international retailing literature and the research on emerging markets. More specifically studies on the topic of market entry failures are very scarce in the international retailing literature. Moreover most of the existing articles are based on qualitative data and not on a quantitative dataset. In spite of a small sample, this research allows us to identify potential new avenues to investigate in further research: time of entry, store density in the host country, brand penetration and relative entropy seem to play a significant role in terms of success or failure of international retailers in emerging countries. Market entry mode, cultural distance, global financial soundness, and host country attractiveness should also been taken into consideration to a lesser extent.

The debate is raging in regard to the time of entry. Exploratory results in the present article tend to demonstrate that early entrants are more successful. Entry timing seems to be important in this matter since more mature markets with higher store density that may in fact reflect stronger local competition, seem to negatively impact the outcome of the new market entry. Early entrants get the opportunity to increase their brand penetration and reach a higher entropy level due to a lower level of competition. Moreover it appears that retailers that have not reached a certain level of entropy and brand penetration in a given country by year seven are more likely to divest than others. It is a race against time. That would of course be particularly true if the retailer encounters at the same time financial tensions in its domestic market. The relative entropy score may in fact be reflective of the profitability level and overall relative contribution of the local venture as original market entry costs such as national advertising, public relation or other awareness building expenses would not be followed by a sufficient number of openings in the host country and may therefore compare negatively with other international locations across the chain. In that extent this may be why the size of the firm was an irrelevant antecedent to success or failure since the entropy rate is a relative measurement. In other words a smaller chain will reach a high relative entropy score in a given country much faster than a larger retailer. The large retailer would need to make very substantial investments in the host country in order to attain their “acceptable” entropy threshold in the first 7 years. Additionally large firms have a tendency to prefer governance modes that give them high control in their local subsidiaries. The downside of this strategy is that they have to fund their market expansion themselves and the costs may be prohibitive before they reach their critical mass threshold. It would be of great interest to develop research in international franchising trying for instance to show to which extent plural form organizations can help franchisors in managing chains (Dant et al. 2008) as it seems to be the case in domestic markets (Bradach 1997).

One may also argue that the mode of entry choice becomes therefore a significant component of success or failure. Franchising or licensing and to a lesser extent, minority IRJVs may help speed up spatial coverage in the host country and therefore increase both brand penetration and entropy score. Moreover a positive correlation between low control mode of governance and success was identified in this research. This exploratory finding that should be tested more extensively in further research slightly contradicts certain authors in the current literature who consider IRJVs as the riskiest mode of entry (Li 1995; Burt et al. 2003). Furthermore most retailers tend to consider IRJVs for example, as a necessary but temporary step when entering a new market. As soon as China entered the World Trade Organization and loosened the regulations on market entry, many international retailers chose to become majority shareholders or sole owners of their foreign business ventures.

Data availability is one of the major limitations for this type of research. Although limited, our dataset seemed nevertheless sufficient to assess our hypotheses by using quantitative methods. The key contribution of this paper has also been to investigate a new potential antecedent of failure or success of international retailing: the entropy threshold as a measure of territory coverage. This is a promising and fascinating field of research that may yield substantial managerial implications. Relative entropy may become a recognized metrics for evaluating foreign subsidiaries for a specific chain and to help assess progresses in terms of store openings in a given country.

6 Conclusion

At time of economic turbulence it is more essential than ever for organizations to accurately forecast and assess their international expansion progresses. It is also crucial to make the right choices when prioritizing their investment. The “me” too approach that may lead a company to enter apparently promising markets such as Brazil, China or Russia may just not work for all retailers. Strategies may require adaptations for each situation by taking into consideration criteria such as retailer’s characteristics, host market specificities but also key learning from past experiences. Franchising, licensing and potentially minority IRJVs may become a primary choice for market entry as a way to limit risks in high uncertainty markets and an opportunity to quickly reach a suitable relative entropy level and brand penetration in a host country.