IDENTIFYING THE FRANCHISEE PROFILES FRANCHISORS PREFER

. This study identifies the franchisee profiles that franchisors prefer. Specifically, the goals of this study are: (1) to identify the characteristics that form the franchisee profile preferred by the franchisors operating in the Spanish market and (2) to contrast diverse hypotheses related to this profile. These hypotheses are related to the ideal fran chisee, sector, financial level, number of franchisee units and age of franchisors. Criteria that franchisors look for in potential franchisees are ranked by importance. The value of the paper is significant as it provides a practical framework for franchisors for the selection process of franchisees when choosing from a group of potential franchisees. The franchisee profile was obtained using conjoint analysis, which is a decomposition methodology that is rarely used in this field.


Introduction
Effects of franchising in the world are extensive and important. Franchising helps as a business development and growth strategy in numerous contexts around the globe (Rahatullah, Raeside 2009). With sales accounting for over three percent of the United States GDP, business format franchise systems represent a significant segment of the retail sector of the U. S. economy (Lafontaine, Shaw 1999). In other countries, franchising has also experienced a rapid growth in the last decades (Tormo and Associates 2010).
Franchising has been one of the most popular means of engaging in business in recent years (Peretiatko et al. 2009). The development of franchising in many countries is due to the expansion objective generated by many companies that have suitable positioning. Also, an initial domestic presence is viewed as crucial for successful international expansion (Quinn, Alexander 2002).
Franchising's longevity and success may be due to the fact that, organizationally, it represents a collaborative alliance (Hoffman, Preble 2003) that depends on the cooperation of two entrepreneurs (franchisor / franchisee) in order to be successful (Shane, Hoy 1996). Selecting the right partner is vital for the success of relationships (Leiser 2002;Sanders 2002;Rahatullah, Raeside 2009). This selection should not to be based solely on experience, intuition or chance. Robust studies are necessary to help franchisors with this topic (Ramirez 2007).
This paper evaluates franchisors preferences about franchisees. Specifically, the objective of this study is to identify the characteristics that form the franchisee profile preferred by the franchisors that operate in the Spanish market. Several hypotheses regarding this profile are contrasted. These hypotheses are related to the ideal franchisee, sector, financial level, number of franchisee units and age of franchisors. In order to answer all these questions conjoint analysis is used.
This work differs essentially from the majority of previous works in two aspects. First, in the study of the franchisee profile a specific set of defined a posteriori variables by experts in franchising is carried out. Secondly, the franchisee profile was obtained using a methodology that is rarely used in this field: conjoint analysis. The article proceeds with a brief review of the literature on selection of franchisees, after which we develop our hypothesis about the franchisee profiles. We then describe the data, methods and results, after which we offer our conclusions.

An overview of the literature
Franchising is a business format in which franchisors compete in at least two markets: (1) candidates to be franchisees, where they commercialize their concept of business, and (2) consumers, where they market their final products or services (López, Ventura 2001). Franchising is a type of collaborative business with the existing of a profit sharing problem (Stein, Ginevičius 2010). Contrary to other types of retail channel arrangement, franchise networks represent a type of coordination that is much deeper than simple buyer-sellers relationships (Gauzente 2010). In this sense, cooperation is very important because it mainly affects the profitability of an enterprise (Ginevičius 2010).
Selection of franchisees is perceived as one of the most important operational problems for franchisors. An appropriate selection of franchisees may bring about remarkable results. On the contrary, a poor selection of them may cause problems in the net (Justis, Judd 1989). The future success of franchisors depends, to a great extent, on a systematic process of selecting good potential franchisees (Olm et al. 1988). Therefore franchising system can be more appropriate for some franchisees than for others (Jambulingam, Nevin 1999).
A degree of conflict prevails in any franchise system as the partners have different behaviours, approaches, and goals. Also, the different governance levels of franchised intangibles profit sharing problems between the partners (Hendrikse, Jiang 2005;Ginevičius, Krivka 2008;Stein, Ginevičius 2010). Franchisees have certain expectations of participation in the process of which they are an integral part (Stanworth et al. 2004). In this sense, the formation of unrealistic expectations increases the potential for future relational conflict (Weaven et al. 2010). Given the underlying potential for conflict in the franchise relationship, an essential challenge for the franchisor is to coordinate the franchise system without escalating conflict (Tikoo 2005).
To avoid relationship conflicts, franchisors should select franchisees with compatible characteristics (Hoy, Shane 1998). Hoy and Shane contend that when a franchisor recruits a franchisee with the right characteristics, the latter will be less likely to shirk and resort to opportunism.
The capacity that some franchisors have in order to correctly select their franchisees means a competitive advantage (Jambulingam, Nevin 1999). This is a remarkable weapon that differentiates leader franchising chains from the rest (Carroll, Bassuk 2002).
The Franchising Task Force (1991) identified the wrong selection of franchisees among the main causes of franchising failure in Australia. Clarkin and Swavely (2006) indicated that little is known about how franchisors evaluate these characteristics in franchisees. More specifically, the relative importance that franchisors grant to the different criteria used in franchisee selection has not been examined sufficiently.
To sum up, there is a generalized agreement about the importance of a suitable selection of franchisees for the success of the franchisor-franchisee relationship. Therefore, the analysis to determine which are the personal and entrepreneurial characteristics of adequate franchisees is a key aspect for the development of franchising.

Franchisee profile characteristics
The study of the characteristics of franchisees can allow franchisors to develop strategies according to the knowledge of the main features that affect the success and failure of franchisees (Falbe, Welsh 1998). Equally, these characteristics allow the differentiation of good franchisees from the rest, so that the company results generated by them can also differ to a quite high degree. In a study carried out in Spain, Bordonaba (2003) showed that the third cause for which a franchisor would not renew the contracts with some franchisees is the lack of a suitable profile. Previously, Gonçalves and Duarte (1994) found the difficulty of recruiting enough qualified franchisees as a major problem in franchise system in Portugal. As Bergen et al. (1992) point out, the onus of success largely depends on the franchisor' ability to select franchisees with the right profile.
In Table 1, we show the main publications on the franchisee profile. These studies have been focused on franchisor' financial requirements, franchisee experience and skills, franchisees' personal traits, and demographic factors.  Tatham et al. (1972) They show several criteria used in the profile by order of importance Edens et al. (1976) To determine which biographical, past and personality characteristics are more important Knight (1986) To detect which are the personal characteristics of the franchisees required to achieve success Withane (1991) To determine which personal and business characteristics of the franchisees are necessary for the success of the franchise Castrogiovanni et al. (1993) To identify the characteristics of the more qualified franchisees

DeCeglie (1993)
To identify the ideal franchisee profile Forward and Fulop (1993) To determine what characteristics the franchisors consider as more important in the franchisees Kahn (1994) To identify the ideal franchisee profile Hing (1995) To determine what characteristics of the franchisee can contribute to their satisfaction Kaufmann and Stanworth (1995) To determine various characteristics of potential franchisees interested in becoming franchisees Shubart and Bennett (1997) To determine the average franchisee profile by means of a descriptive study Falbe and Welsh (1998) Perceptions of the franchisees about the characteristics associated with their success and failure Fenwick and Strombon (1998) To verify if several characteristics of the past of the franchisees are determinants for the operation of the franchise Jambulingam and Nevin (1999) To relate the criteria used in the selection of franchisees to the results desired by the franchisors Berni (2002) To determine the basic values that are common to all franchisees Kasselmann et al. (2002) To identify the personality characteristics of franchisees successful in the fast food sector Clarkin and Swavely (2006) To measure the importance of six criteria used in the franchisee selection process Rahatullah and Raeside (2009) To evaluate franchisor strategies for selecting the franchisee from the perspective of the franchisor' business age, size and type Source: own elaboration. This list does not claim to be exhaustive.
Two aspects may be highlighted analyzing the existing literature about the characteristics that franchisees must have (Ramirez 2007): a) Although franchising is a system that has undergone an important growth in recent years at a worldwide level, the academic studies that have analyzed the franchisee profile are very scarce. Therefore, the advance of research in the matter has not moved in line with that experienced by franchising in the economy.
b) The majority of the existing studies of the franchisee profile are descriptive, which does not allow its generalization. In this sense, the existing literature has followed the positive scientific approach (what it is). Usually, the literature has not carried out the normative scientific approach (what should be). That is to say, to identify the characteristics that franchisees should contain.
Reviewing previous literature about the franchisee profile, the main characteristics that make it up have been brought together (Table 2). In spite of the studies analyzed, there is apparently little consensus between franchisors about the ideal franchisee (DeCeglie 1993). For example, many franchisors prefer candidates experienced in businesses or management and, nevertheless, others do not give much importance to this aspect. Many franchisors will select franchisees having no previous experience in the industry (McCosker, Frazer 1998;Mendelsohn 1993) and some even actively recruit inexperienced franchisees because they feel they will be easier to indoctrinate into the system (Frazer 2001). For these reasons, despite concessions offered by some franchisors to franchisees as to entry fees, royalties or training, franchisors must acquire the means to evaluate candidates (Gauzente 2003).

Hypotheses formulation
In order to obtain a franchisee profile, interviews from a representative sample of franchisors operating in Spain are carried out. To obtain the best franchisee profile, called ideal franchisee profile, it is necessary to analyze solely the opinion of "the more profitable franchisors". We understand as "more profitable franchisors" those that present high standards in some variables. The suitable variable is yield or profits of each franchisee unit, nevertheless, this variable is not usually facilitated by the companies. Because of the availability of data the criteria of Tormo and Associates (2005) consultants are followed. They set up a ranking of franchising chains based on their incomes (own plus franchised units). Therefore, to obtain the ideal franchisee profile only the opinions of franchisors that present high incomes are taken into account. This is compared with the opinion of the rest of franchisors. Consequently, a priori the ideal franchisee profile and the franchisee profile obtained with the opinion of the rest of the franchisors will be different.
The lack of works that have investigated the franchisee profile by means of using conjoint analysis indicates the innovation of this paper and, therefore, the scarce literature about its theoretical justification. As a result of the previous commentaries, the first hypothesis of this research is worded in the following form:

H1:
The franchisee profile obtained from the franchisors with higher incomes differs from that obtained from the rest of franchisors.
Franchising in Spain has expanded in very varied and different sectors, from fashion retailers to hotels and restaurants or specialized services. In some cases the experienced franchisee candidates are preferred in technically complex sectors. In other cases the inexperienced franchisee candidates are preferred to avoid acquired practices that can complicate their adaptation to the commercial philosophy of franchisors (Forward, Fulop 1993). It is noted that some operations tend to be more complex systems than others (Frazer, Winzar 2005). At present, there are still some differences among the chains, according to the sector of activities (Perrigot 2006). Edens et al. (1976) study includes 184 companies from the United States that represent 5 sectors (fast food, automobiles, retailing, travel agencies and employment agencies).
The study reveals that significant differences by sector exist in the perception of the ideal franchisee profile, so that different work conditions require different characteristics for franchisees. Rahatullah and Raeside (2009) show that the requirements for potential franchisees depend on business type as well. Some franchisors are focused on franchisees with cultural compatibility, technical competence, and financial ability and flexibility, whereas others franchisors require franchisees with managerial ability, trustworthiness and references.
Falbe and Welsh (1998) measure the perceptions that franchisors have about the factors associated with the success and failure of franchisees. Conversely to the study of Edens et al. (1976), the results show that the type of franchising chain did not have any effect on the success or failure characteristics of franchisees. Following the franchisors classification of Tormo and Associates (2005) we have divided the population in three wide groups: Retailing, Hotels and Restaurant and Services. In view of these different results, the following hypotheses is considered:

H2:
The preferences of franchisor about franchisee profile do not vary based on the sector in which the chain is developed.
A very important aspect at the time of integrating a candidate as a franchisee is his / her financial level. This is going to allow the franchisee to be able to face the initial costs (entry fee) as well as the periodic costs (royalties on sales, advertising). In franchise system franchisor receives contractual fees from franchisee and franchisee receives profit after the fee is paid to franchisor (Stein, Ginevičius 2010). In the same way, to have a high financial standing is going to allow the franchisee to be able to resolve any unforeseen costs that may arise. Tatham et al. (1972) carry out a study on characteristics of franchisees in which they indicate that the financial level is the first criterion used by franchisors in the selection of its franchisees. Bordonaba (2003) concludes that the lack of payment of royalties and fees is one of the main reasons why a franchisor would not renew some of the contracts of their franchisees. Both franchisees and franchisors consider that to have high financial support is an essential characteristic required for franchisees to be successful (Knight 1986). Furthermore, Hing (1995) identifies financial status as one of the four characteristics of franchisees that influence their satisfaction. DeCeglie (1993) also indicates that the financial resources constitute one of the characteristics that franchisors look for in franchisees. In addition, Tormo and Associates (2005) consider the economic capacity as a very important quality within the franchisee profile. Franchisees will have to contemplate not only the capital necessary to start up the business, but also, to take it to the point in which it begins to be profitable. The opposite can lead to failure and the consequent closing of the activity. It is important to indicate that if a franchisee invests a substantial amount of money in the business then he or she will be motivated to work hard to retrieve it (Frazer, Winzar 2005). In view of all the previous results, the following hypothesis is considered: H3: Financial level of a franchisee is the most important attribute for a franchisor at the time of carrying out the selection.
Previously, the importance of franchisee as a key element in the determination of success of the franchisor-franchisee relationship is analyzed (Grünhagen, Dorsch 2003;Howe 2003;Sanders 2002;Sturgis 1993). Some franchisors have a high number of franchisees. Adding new franchised units may help franchisors to overcome growth constrains (Ginevičius et al. 2007). To a certain extent, this fact allows them to work hard in the selection and consolidation of future franchisees. Therefore, this is a remarkable feature that can be used to differentiate franchising chain leaders from the rest (Carroll, Bassuk 2002). This means that the franchising chain leaders possess a series of tangible and intangible qualities that many other franchisors wish to follow. In general, a high degree strategic planning and a better competitive position are positively related to firm size (De Jorge et al. 2010). In addition, many franchisors change the characteristics asked for possible candidates, due to the experience they have accumulated over the years (Forward, Fulop 1993). The most unstable period is when the first franchisees are selected. It is a period in which franchisors do not know who are good franchisees, because they are not experienced enough. These authors even affirm that some franchisors would not nowadays select franchisees they selected at the beginning of the franchising period. It seems logical that franchisors with a great number of franchisees will have a franchisee profile different from those franchisors with a minor number of franchisee units. Nevertheless, Falbe and Welsh (1998) find that size of a franchising chain does not affect the success or failure of franchisees. On the other hand, Rahatullah and Raeside (2009) show that size of firm plays an important role in future franchisee selection.
The experience accumulated in a certain period of time depends on the size of franchisor (Castrogiovanni et al. 1993). Franchisors with many franchisee units gain experience since they must deal simultaneously with different markets, whereas franchisors with few franchisee units only gain experience in one or a few markets. As franchisors gain experience, they are more trained to identify qualified franchisees. In this sense, the existence of significant differences in the franchisee profile between those franchisors that have a greater number of franchisees and those that have a smaller number of them is contrasted. In order to determine the franchisors with a smaller number of franchisees (group 1) and a greater number of them (group 2), the first and the third quartile of the variable number of franchisee establishments are considered, obtaining in this way an effective discrimination for this variable. The hypothesis raised on this matter is:

H4:
The preferences of franchisor about franchisee profile vary depending on the number of franchisee units that each one has.
According to Forward and Fulop (1993) experience accumulated by each franchisor should have a good effect not only in obtaining a better selection of franchisees but also in many other aspects: improvement of the learning process that is carried out with franchisees, a more effective fixation of payments, location of the stores in strategic geographic zones, etc. Peterson and Dant (1990) indicate that the longer franchisees have become part of the chain positively influences the whole system. Also, Castrogiovanni et al. (1993) consider that the experience gained by franchisors over the years improves their capacity to identify qualified franchisees. In Frazer and Winzar (2005) study about franchising failure, all franchisors with franchisee failure admit to choosing unsuitable franchisees in their early days, sometimes out of a need to quickly start up the system and other times due to incorrect selection criteria.
Nevertheless, the accumulated experience not only depends on the size of franchisor, but also experience is related to their age. For example, a franchisor that quickly opens 500 units in its first year probably has accumulated less knowledge than a franchisor that has gradually obtained a size of 500 franchisee units in ten years. Surprisingly, Castrogiovanni et al. (1993) reveal that the age of a chain is positively related to its failure rate that is to say that the older the franchisors are, the higher their failure ratio is. The explanation of this fact is that success conditions for a franchising chain can change over the years. Probably, a franchisee unit that was successful for 10 years, perhaps it is not now because market conditions have changed. Possibly the argument of learning is valid for the first 3-5 years of existence of franchisor, but after this period, the age is more associated with obsolete units than with learning (Bruderl, Schussler 1990). Therefore, the ratio of franchisor failure is high at the beginning, due to the little experience accumulated, and at the end, due to the effects of the obsolete units. As a consequence of that, this study of Castrogiovanni et al. (1993) show a positive relationship between the age of franchisor and the ratio of failure because the sample is biased by including the oldest franchisors. Rahatullah and Raeside (2009) show that like size of firm, age plays an important role in future franchisee selection.
Franchising is in a maturity stage in Spain, in which most of franchisors are more than 3-5 years old and in which the existence of obsolete units is insignificant. There are just a few franchisors in Spain with less than 1 year old. Therefore, most franchisors have accumulated a similar experience that allows them to select qualified franchisees, for this reason it can be thought that the older franchisors will select franchisees in the same way than the newer ones. The boom of franchising in Spain began in 1986 (Alonso 2005), so in this paper the sample is divided in two groups: those chains that are founded before 1986 (group 1) and the rest (group 2). In relation to the previous ideas the following hypotheses is put forward:

H5:
The preferences of franchisor about franchisee profile does not vary based on the age of franchisors.

Methodology
In order to find the franchisee profile preferred by franchisors conjoint analysis has been used. The first stage to be able to apply conjoint analysis is to identify the attributes and levels that are going to make up the profile wished for. To obtain the attributes, a survey of representative experts in franchising has been used, by means of a directed non-random sampling method, in order to value the different characteristics that may conform the franchisee profile. One of the most used procedure for identifying and selecting attributes is the opinion of experts (Marzocchi et al. 2003). These representative experts were selected amongst members of associations of franchising chains, university professors researching in franchising, managers of franchising consultants, editors of magazines from this sector, managers of websites focused on franchising and managers of the area of franchising from financial firms specialized in franchising. In this selection of experts, franchisors were not considered to avoid the effect that their own experience could cause. A questionnaire with a set of features was sent to each of these experts that can be associated with the franchisee profile. The experts had to mark the seven characteristics they considered most desirable for a franchisee candidate 1 . The characteristics sent to the experts were obtained from existing studies that analyze the franchisee profile.
The sample size of representative experts was 73. Data collection was carried out between December 2008 and January 2009, by means of a questionnaire sent by email. The total number of questionnaires received was 39, which means an answer rate of 53.4%.
The ten attributes most frequently marked were: management capability, capacity for human relations, entrepreneurship, wish for success, willingness to work hard, previous experience in a related business, faculty of adaptation to changes, loyalty to franchisor, financial level and perseverance 2 .
Once the attributes have been identified, the next step is the establishment of levels.
Levels constitute the real measures in the construction of the attributes. The range of attribute variation used in a conjoint design influences the inferred attribute importance (Wittink et al. 1990;Verlegh et al. 2002). It is noteworthy that even if the range is held constant, the addition of intermediate levels can increase this importance. So, a balanced number of levels must be applied to the attributes in order to minimize the influence on the attributes importance. Therefore, in this work high differences about the number of levels have been avoided, establishing attributes by only two or three levels. Also, in this study, levels have been defined looking for similarity to reality. The levels that have come up appear in Table 3. Conjoint analysis is a multivariate technique whose theoretical assumptions are less than for other techniques. Nevertheless, the practical assumptions are higher, since the researcher must specify the design of the model, its estimation and its interpretation (Hair et al. 1999).
Twenty eight cards or combinations established in the orthogonal design 3 were shown to the respondents. Each person interviewed had to score on a scale of 1 (less favourite) to 7 (more favourite). Therefore, a metric scale was used, because this is the one that provides the most reliable results among the different scale possibilities of the dependent variable (Ramírez, Rondán 2004).
The questionnaire was sent by email and post, from April to June 2009, to the 792 franchisors that operate in Spain according to Tormo and Associates (2005). After a pretest of the questionnaire with 10 franchisors, we contacted the rest of the population. In all cases, we contacted the Expansion Director or Franchise Director of every chain and asked him / her to respond to the questionnaire. Of the 792 remaining firms, we received 192 total responses, of which 181 were usable (22.85% usable response rate), because 11 questionnaire were uncompleted.

Results
Results were obtained using SPSS v16.0. Syntax analysis was applied to the data through the CATEGORIES module of SPSS (Table 4). SPSS offers information of the relative importance of attributes and of partial utility of levels.
It can be observed that the most important attribute is loyalty to franchisor (19.1%), followed by management capability (16.4%) and willingness to work hard (14.2%).
Franchisors grant the smallest importance to previous experience in a related business (5.8%) and financial level of franchisee (7.8%). The importance given by franchisors towards the rest of attributes is between the values previously indicated. It also stands out that the most important attribute is almost four times more important than the least important one. Another noticeable aspect is the low importance that franchisors give to the attribute of financial level. This assumes that franchisee has the minimum financial requirement to undertake the necessary investment to join the franchising chain, since on the contrary the candidate could not become a member.
Other information obtained is the partial utility of each one of the levels associated with each attribute, which can be defined as the satisfaction that each level provides to franchisors. Therefore, the franchisee profile preferred by franchisors has to bring together levels that present a higher partial utility. As a consequence, the franchisee profile would have to be a candidate: with a management capability at a professional level, extroverted, with an entrepreneur character, with a very accentuated wish for success, working full time in the franchisee unit, having previous experience in a related business, that faithfully follows the indications of the franchisor, with a financial level over the required initial investment and perseverant.

Discussion
With hypothesis H1 we try to contrast if the importance granted to the attributes by franchisors with higher incomes (ideal franchisee profile) differs from the rest of franchisors. In order to contrast this hypothesis, Kendall tau and Spearman rho non-parametric tests are used (Table 5). Since the variable used is the relative importance, a disposition of ranks can be obtained, thus allowing the application of these tests.
Kendall tau and Spearman rho tests take values between -1 and +1 and they allow the contrasting of the null hypothesis of no correlation or no association between the variables. Under the null hypothesis, the statistic is zero. The results obtained when applying these tests indicate that the existing correlation between both variables is significant at a level of confidence of 99%. According to this tests an association between the aforementioned variables exists, which means that hypothesis H1 is rejected. That is to say, that the importance that high income franchisors grant to the attributes does not differ from that given by the rest of franchisors. So, the ideal franchisee profile does not agree with the franchisee profile obtained from the opinion of the rest of franchisors. With the purpose of contrasting hypotheses H2 the 181 franchisors of the sample were grouped in the retailing, hotels and restaurants, and service sectors. A total of 58 firms were included in the retailing sector, 35 were hotels and restaurants and 88 came from services. Next, conjoint analysis was applied to each of these sectors, obtaining the three highest score attributes in the three sectors. There are some different among the industries but they are not significant. In this case the most important attributes were loyalty to franchisor, management capability and willingness to work hard. Loyalty to franchisor is much more important in the sectors of hotels and restaurants and retailing than in the service sector. Also, it highlights that management capability is more noticeable for the service sector than for the other two. In addition loyalty is more valued in the sector of hotels and restaurants than in the others. Furthermore, the lowest scored attribute in the three sectors is the previous experience in a related business, given greater importance in the sector of hotels and restaurants, next in services and last in retailing.
The Kendall tau and Spearman rho tests were used to contrast hypotheses H2. The results deduce that correlation between the three sectors is significant at a level of confidence of 99%, confirming that the franchisee profile does not vary depending on the industry. The similarity between the profiles by sectors is very elevated. The sectors retailing and hotels and restaurants were those that show a greater correlation. So, hypotheses H2 is accepted.
In reference to hypothesis 3, the results shown in Table 4 indicate that for franchisors the financial level of the potential franchisee is not the most important attribute in the decision process. Therefore, this hypothesis is not verified. In general, this result is not consistent with the previous studies of Edens et al. (1976), Falbe and Welsh (1998), Hing (1995), Knight (1986), Tatham et al. (1972) and Tormo and Associates (2005). It is necessary to insist on this hypothesis because it is meaningful supposing that franchisee reaches the indispensable minimum financial level for its possible incorporation in a franchising chain. If the candidate does not reach the minimum financial level required, it will not be able to join any franchising chain, which is why it is not appropriate to analyze the importance of this attribute. This result can be explained due to the fact that the study was carried out in Spain, a country with a stable economy and a positive economic growth in the earlier 2009. Today financial status could be more important in a declining economic environment or in a context with higher uncertainty in the level of economic growth.
This study tries to confirm the hypothesis that the franchisee profile preferred by franchisors with a higher number of franchisee units is different from the franchisee profile preferred by franchisors with a smaller number of franchisees. In order to find this profile, the variable number of franchisee establishments of the 181 franchisors of the sample was ranked. Next, the sample was divided into two subsets. The first was composed of franchisors with a number of establishments inferior to the first quartile of the distribution and the second of franchisors with a number of establishments over the third quartile of the distribution. In this way, an effective discrimination of the variable number of franchisee units is obtained.
To analyze the data, two groups were taken. One set up by chains with 5 or less franchisee units (Group A) and another formed by chains with 47 or more franchisees (Group B). The relative importance of the attributes was calculated applying conjoint analysis to both groups, showing that it is very similar in both groups. For the two groups, the highest score attributes were loyalty to franchisor, management capability and willingness to work hard, whereas the less important attributes were previous experience in a related business and financial level, for group A, and previous experience in a related business and wish for success for group B. Anyway, essential differences do not exist with regard to the importance assigned to each of these attributes. The greatest difference is in the attributes perseverance (2.5) and capacity for human relations (2.0). Franchisors with a higher number of franchisee units scored higher capacity for human relations than franchisors with a smaller number of franchisees, whereas the latter scored higher perseverance of franchisees than the former. Furthermore, franchisors with a greater number of franchisee establishments evaluate more positively the financial level that franchisee candidates can have, whereas previous experience in a related business is more valued by franchisors with a smaller number of franchisees. As in the previous cases, the corresponding hypotheses was contrasted by the Kendall tau and Spearman rho tests, and the results of these tests reject hypotheses H4.
In order to contrast hypotheses H5, data of the oldest chains were separated from the most recent ones. The boom of franchising in Spain began in 1985-86 (Alonso 2005).
For this reason the sample was divided in two groups: those chains that were founded before the year of the boom of franchising in Spain and the rest 4 . It can be observed that for the oldest chains the three most important attributes were, by order of importance, willingness to work hard, management capability and loyalty to franchisor, whereas the least important attributes were previous experience in a related business and perseverance. With regard to the most recent chains, the most significant attributes by order of importance were loyalty to franchisor, management capability and willingness to work hard, whereas the least important were previous experience in a related business and financial level. In this case, the results of the aforementioned tests indicated that hypotheses H5 is accepted.

Managerial implications and conclusions
In this work the characteristics that should configure the franchisee profile are identified, so that their selection by franchisors were carried out fulfilling the objectives of both parts. A good selection of franchisees will facilitate a successful and long relationship between them and franchisors.
Every franchisee candidate should fulfil some minimum requirements. Some of these are demanded by franchisors and others would have to belong to the personality of the candidate in order to join a franchising chain. For a correct development of the selection process it is necessary to determine the personal and entrepreneurial characteristics of the ideal franchisee. In spite of the difficulty of determining this profile, as Díez, Navarro and Rondan (2005) indicate, since each franchisor usually has its own franchisee profile candidate from its own experience, a common basic criteria may exist that all franchisees should have.
From the results it can be highlight that among all these attributes, loyalty to franchisor is the most important by far. The second most essential attribute is management capability, followed by willingness to work hard. On the contrary, the lowest scored attributes are previous experience in a related business and financial level. The relative importance of the rest of the attributes is placed between the abovementioned groups.
Loyalty to the franchisor is linked to commitment to business, trust in the franchisor, cooperation, and communication. Loyalty suggests that franchisees are likely to be tolerant to the franchisor's influence. This allows franchisor to coordinate franchisees without creating conflict (Tikoo 2005). This result is consistent with other studies that have found that commitment, trust, cooperation, and communication are very important for the success of franchisor-franchisee relationships (Allen 1994;Chiou et al. 2004;Doherty, Alexander 2006;Clarkin, Swavely 2006;Edens et al. 1976;Falbe, Welsh 1998;Schultz 1999). Previous work has demonstrated that loyalty to the franchisor is seen as an important attribute, but this work yields a ranking of all attributes, thereby showing that fidelity to the franchisor is the attribute with the greatest relative importance. In the early years, franchisees are willing to be guided by franchisors and where few communication problems exist (Nathan 1993). Later, franchisees develop a greater sense of independence (Peterson, Dant 1990) and may be more inclined to deviate from system procedures to demonstrate this independence. Consequently, communication problems and dispute situations are more likely to occur in future (Frazer 2001). In this sense is very important the fidelity of franchisees.
In addition, a person with a management capability at a professional level is desirable. This characteristic means that the candidate should have knowledge of the necessary management tools required for the successful development of business.
Furthermore, franchisor demands from franchisee the availability to work hard at the head of his / her business. Therefore, people who have a strong character are required, ready to assume responsibilities and to manage difficult situations and, essentially, dedicated full time to the franchisee unit.
Franchising is a commercial system that is seen on many occasions as a self employment formula and, on other occasions, as an investment opportunity. In the first case, a franchisee will have a complete dedication to the business, whereas in the second case franchisee can delegate all work to other people. Therefore, the former will be a self employed and the latter will be an investor. In general, self employed and investors have different motivations (Grüner 2006). As a consequence, franchisors prefer entrepreneur franchisees to investor franchisees. In many occasions franchisors recruit franchisees to capture their entrepreneurial spirit and distribute ownership, encouraging franchisees to view the business as a personal venture (Gassenheimer et al. 1996).
Another desirable characteristic for the franchisee profile, although to a lesser extent than the three previous ones, is for the candidate to be an extroverted person. That is to say, open minded, who maintains a constant communication with franchisor as well as with his / her own employees. It is recommendable for this to be a perseverant person, since franchising is not an infallible system, which is why on many occasions dedication and motivation are required to obtain profits. Also, a person with an entrepreneurial character is wanted, that is to say a person who looks for a self employment formula in franchising. Other desirable characteristic in franchisee is to have a noticeable wish for success, that is to say a person who has a great necessity to prevail in businesses, which would mean his / her having a positive attitude towards doing what is best.
Finally, two desirable attributes exist but to a much lesser extent than the previous ones. They are financial level and previous experience in a related business. All franchisees should have a financial level superior to the required initial investment and, in addition, they should have previous experience in a related business. In relation to the financial level, it is necessary to indicate that the analysis of this feature has to be done carefully.
From this research it is deduced that financial level is an attribute of little importance, but this idea is based on the assumption of franchisees reach the minimum level required to be enrolled into the chain. On the other hand, previous literature showed that for some researchers previous experience in a related business is an important characteristic (Kaufmann, Stanworth 1995), whereas in this study and others it is not (Edens et al. 1976;Fenwick, Strombom 1998;Forward, Fulop 1993;Hing 1995;Knight 1986).
In short, the success of the franchise is strongly influenced by the quality of franchisees involved in the network and their characteristics. Therefore, the selection of franchisees is a challenge for franchisors. In this sense, franchisors should follow some important guidelines: I. Franchisors must incorporate to the chain a system defined for recruiting franchisees in order to strengthen the network of franchisees as a whole. An agreement with a consultant or company specialized in personal recruiting should be used if it is not possible the first option. II. Franchisors must look for a person with an elevated level of loyalty as the most desirable feature. Nevertheless, to detect the degree of loyalty of franchisee to franchisor before the beginning of the relationship is quite complicated, since this degree is detected when the relation between both is developing. It would be necessary to apply some psychological tests that allow the detection of the degree of fidelity of a potential franchisee to franchisor. Management capacity must be the second feature that franchisors should look for in franchisees. This implies that the person should have a minimum level of training to know those tools. Thirdly, franchisor should look for franchisees with the availability to work hard at the head of his / her business. III. Franchisors must seek persons who will use the franchise as a self employment formula instead of an investment opportunity. Investor franchisees can be aggressive and hard working in their efforts to strike a good deal. IV. Financial level and previous experience are not very important features in the selection of franchisees. All these conclusions are obtained from the analysis of the franchisee profile preferred by franchisors operating in the Spanish market. In addition, the ideal franchisee profile was obtained analyzing the answers of "the more profitable franchisors". The results show that the franchisee profile agrees with the ideal franchisee profile. Therefore, the profile obtained analyzing the opinion of all franchisors is essentially the same as that obtained solely the higher income franchisors. Only small insignificant differences in the ranking of the attributes between both profiles exist.
Several hypotheses regarding the franchisee profile were contrasted, and the franchisee profile agrees overall with the ideal franchisee profile. In addition, it is deduced that the configuration of the franchisee profile does not vary across the industries analyzed, either based on the number of franchisee establishments, or based on the age of the chains.
The main limitations of the study are related to the application of conjoint analysis. Thus, one of them is the objective definition of attributes and levels by researchers, which may bias results. Another one is the arduous work of valuation that respondents must develop in a process of conjoint analysis, which can also distort results to a certain extent.
In the world, there are franchise markets with different characteristics and stages of development. In this sense, the results of this research, due to specific characteristic of Spain as a franchise market, must be carefully generalized to other markets.
In relation to future research, it is necessary to say that the study of the franchisee profile by means of other multivariate techniques would be interesting in order to compare results, as well as from the point of view of other agents different from franchisors. Obviously, it would be very interesting to replicate the study in other countries, and analyze the franchisee profile wanted by franchisors in those countries.