Criteria used in the selection of franchisees: an application in the service industry

Abstract

The service industry is one of the greatest driving forces behind the growth of franchising in many countries. The aims of this study are to identify the characteristics of the franchisee profile preferred by franchisors in the service industry and to show how the simulation of franchisees with conjoint analysis constitutes a powerful tool for the correct selection of franchisees by franchisors. Criteria that franchisors in the service industry look for in franchisee candidates are ranked by importance using conjoint analysis; a decomposition methodology that is rarely used in this field. 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.

Appendix

A wide variety of models exists, which assume that consumers have a multi-attribute utility function. Generally, a multi-attribute choice alternative x can be represented by means of the following vector annotation (Green and Krieger 1993):

$$ x = \left( {x_{1} ,x_{2} , \ldots ,x_{M} } \right) $$

where x ₁, x ₂,…,x _j ,…,x _M refers to the level of jth attribute that forms part of stimulus x. If an attribute is categorical, its codification will be made by means of p − 1 dummy variables if the number of attribute levels is p.

It is assumed that the decider does not give the same value to all attribute levels but values some more than others. This can be shown by the following utility function:

$$ U\left( {x_{1} ,x_{2} , \ldots ,x_{M} } \right) = f\left[ {u_{1} \left( {x_{1} } \right),u_{2} \left( {x_{2} } \right), \ldots, u_{M} \left( {x_{M} } \right)} \right] $$

where u _j is a function of the partial utility defined on all considered levels of the jth attribute and f[·] is a function that sums the partial utilities of all the attributes.

If we considered another alternative x′ that differs from x, then it can be assumed that:

$$ U\left( x \right) \le U\left( {x^{\prime}} \right) \Leftrightarrow x \,{\le}_{0}\, x^{\prime} $$

where ≤₀ denotes that ‘is not preferred to’.

The additive model that has been frequently used in multi-attribute utility theory is defined as:

$$ U\left( {x_{1} ,x_{2} , \ldots ,x_{M} } \right) = \sum\limits_{j = 1}^{M} {w_{j} u_{j} \left( {x_{j} } \right)} $$

where w _j and u _j (·) are obtained in two stages by means of a self-explanatory model.

In this study, we used SPSS v12.0. Syntax analysis was applied to the data through the CATEGORIES module of SPSS v12.0. Due to the categorical nature of the attributes, a partial utility function model was considered, which yielded a faithful reflection of the possible real structure of franchisor preferences (Azpiazu 1994). The obtained results provide information on the relative importance of the attributes in the study and on the levels that form the franchisee profile. The relative importance of the attributes was obtained from the part-worths of each level forming part of the attribute. An attribute will be more important when there is a greater difference between the highest and lowest part-worths. The importance of attribute i is given by the following equation:

$$ {\text{IMP}}_{i} = {\frac{{{\text{Rank}}_{i} }}{{\sum\nolimits_{j = 1}^{p} {{\text{Rank}}_{j} } }}} \cdot 100 $$

where Rank _i is defined by the following equation:

$$ {\text{Rank}}_{i} = \left| {{ \max }\left( {u_{ik} } \right) - \min \left( {u_{ik} } \right)} \right|\quad \forall i = 1, \ldots ,p,\;\forall k $$