عملکرد و وعده انتخاب رسمی تامین کننده: مطالعه چهار مورد تجربی

In this paper an experimental study of formal decision-making models for supplier selection is described. Attention is paid to all phases of the supplier selection process: the recognition of the need for a new supplier, the formulation of decision criteria, the qualification of suitable candidates and final selection. It appears from the experiments that formal decision models may prove to be useful in various ways throughout the whole supplier selection process and in different purchasing situations.

Academic attention for a more systematic approach to decision-making in purchasing and especially supplier selection has increased steadily over the last three decades; see Weber et al. (1991) and De Boer et al. (2001) for extensive literature overviews on methods and tools for supporting supplier selection. Almost invariably, authors on this topic justify their efforts in developing decision support tools and methods by pointing to the increased importance and complexity of purchasing and supply management in general, the crucial role of supplier selection decisions within the purchasing process and a lack of available decision tools—at least for certain specific purchasing settings (see De Boer et al., 2001). Still, while the number of decision tools seems to grow steadily, there is little empirical scientific evidence of the practical merit of such tools in the supplier selection practice. Usually, the decision tools for supplier selection are only provisionally tested on a fictitious example for illustrative purposes although usually based on input data that were gathered in practice. The few real empirical applications (see for example Liu et al., 2000; Vokurka et al., 1996; Karpak et al., 1999) appear without a systematic and comprehensive analysis of such aspects as user-appreciation, costs of building the model, the availability of data, the integration in existing systems and procedures and so on. In addition, while Lamming et al. (1996) discuss the perceived benefits and problems of vendor assessment systems, their research essentially focuses on how the buyer and supplier may jointly assess their relationship rather than how the buyer makes individual decisions regarding suppliers. Therefore, the study presented in this paper attempts to assess more firmly the perceived merit of using formal decision tools and approaches for supplier selection in practice. More specifically, we aim to address the following basic questions: “To what extent and how can decision models provide useful support to purchasing decision makers when it comes to supplier selection?”. As reported by De Boer et al. (2001), a supplier selection problem typically consists of four phases, namely (1) problem definition, (2) formulation of criteria, (3) qualification of suitable suppliers and (4) final selection of the ultimate supplier(s). In each phase a different set of models is appropriate (see Table 2 in the sequel). We investigate to what extent decision-makers are receptive to the use of formal tools for each of the phases by letting them apply one or more tools in the different phases of a previously solved supplier selection situation and letting them evaluate the possible merits and demerits of the use of these tools. So, the goal of our research is not to evaluate specific models or to develop an evaluation method, but to investigate the receptivity of decision-makers to the use of formal decision tools. Table 2. Overview of the literature on supplier selection models used in the experiments New task Modified rebuy Straight rebuy (routine items) Straight rebuy (strategic/bottleneck) Problem definition None found Formulation of criteria Mandal and Deshmukh (1994) Mandal and Deshmukh (1994), Vokurka et al. (1996) Qualification De Boer et al. (1998), Barbarosoglu and Yazgaç (1997) De Boer et al. (1998) Timmerman (categorical) (1986) Timmerman (categorical) (1986) De Boer et al. (1998) Hinkle et al. (1969) Grando and Sianesi (1996) Holt (1998) Gregory (1986) Papagapiou et al. (1996) Hinkle et al. (1969) Timmerman (categorical) (1986) Holt (1998) Li et al. (1997) Liu et al. (2000) Min (1994) Narasimhan (1983) Ng and Skitmore (1995) Nydick and Hill (1992) Papagapiou et al. (1996) Soukup (1987) Thompson (1990) Timmerman (categorical) (1986) Vokurka et al. (1996) Williams (1984) Willis et al. (1993) Choice Barbarosoglu and Yazgaç (1997) Akinc (1993) Albino and Garavelli (1998) Barbarosoglu and Yazgaç (1997) De Boer et al. (1998) Albino and Garavelli (1998) Barbarosoglu and Yazgaç (1997) De Boer et al. (1998) Grando and Sianesi (1996) Barbarosoglu and Yazgaç (1997) De Boer et al. (1998) Grando and Sianesi (1996) Gregory (1986) Bender et al. (1985) Grando and Sianesi (1996) Gregory (1986) Holt (1998) Benton (1991) Gregory (1986) Holt (1998) Li et al. (1997) Buffa and Jackson (1983) Holt (1998) Li et al. (1997) Min (1994) Chaudhry et al. (1993) Khoo et al. (1998) Min (1994) Narasimhan (1983) Current and Weber (1994) Li et al. (1997) Morlacchi (1997) and Morlacchi (1999) Nydick and Hill (1992) Das and Tyagi (1994) Min (1994) Narasimhan (1983) Sarkis and Talluri (2000) De Boer et al. (1998) Narasimhan (1983) Nydick and Hill (1992) Thompson (1990) and Thompson (1991) Degraeve and Roodhooft (1998), Degraeve and Roodhooft (1999) and Degraeve and Roodhooft (2000)) Nydick and Hill (1992) Sarkis and Talluri (2000) Vokurka et al. (1996) Degraeve et al. (2000) Soukup (1987) Thompson (1990) and Thompson (1991) Morlacchi (1997) and Morlacchi (1999) Weber and Ellram (1992) Thompson (1990) and Thompson (1991) Williams (1984) Williams (1984) Ganeshan et al. (1999) Williams (1984) Willis et al. (1993) Willis et al. (1993) Ghoudsypour and O’Brien (1998) Willis et al. (1993) Yoon and Naadimuthu (1993) Yoon and Naadimuthu (1993) Grando and Sianesi (1996) Yoon and Naadimuthu (1993) Gregory (1986) Holt (1998)Morlacchi (1997) and Morlacchi (1999)Karpak et al. (1999)Khoo et al. (1998) Li et al. (1997) Min (1994) Monczka and Trecha (1988) Narasimhan (1983) Nydick and Hill (1992) Pan (1989) Petroni and Braglia (2000) Ronen and Trietsch (1988) Rosenthal et al. (1995) Sadrian and Yoon (1994)Turner (1988) Smytka and Clemens (1993) Soukup (1987) Thompson (1990) and Thompson (1991) Timmerman (cost-ratio) (1986) Vokurka et al. (1996) Weber and Current (1993) Weber and Desai (1996) Weber (1991) Weber et al (1991), Weber et al (1998) and Weber et al (2000) Williams (1984) Willis et al. (1993) Yoon and Naadimuthu (1993) Table options The paper is organised as follows. First, we further specify the problem statement and the methodology applied. Next, we describe four experiments in which we applied and subsequently evaluated several decision tools suggested in the existing literature. Finally, we discuss the findings from the experiments, draw general conclusions from these and give recommendations for further research.

Based on the foregoing, the answer to our question “To what extent and how can decision models provide useful support to purchasers when it comes to supplier selection?” is twofold. First of all, in general the response resulting from the interviews can be described as ranging from neutral to positive with regard to the overall usefulness of decision models for supplier selection. This leads us to suggest that using decision models in supplier selection—provided that they are carefully selected and given certain conditions—may prove useful in various ways throughout the whole supplier selection process in fundamentally different purchasing situations. In our view, this underlines the usefulness and importance of research in this area. It also shows that the wide range of methods and techniques that has emerged during the past 50 or so years in Operations Research seems to provide the required variety of approaches required for coping with the diversity found in the purchasing practice. Furthermore, it implies the need for a shift in emphasis from developing decision models for the final selection phase to the earlier phases of problem definition and the formulation of criteria. Secondly, the analysis reveals the importance of considering the flexibility of a decision model in relation to the function of providing structure throughout the process. An extremely flexible decision model may limit the ‘structuring’ function. Depending on the respective importance of these two aspects for the decision-maker, different decision models—or ways of using them—may be considered. Also, the nature of the input information provided by the decision-makers requires specific attention. On the one hand, the acceptability of the outcome relies on using ‘unpolished’ personal judgements while objectivity of input information is important, as the subsequent aggregation mechanisms are not always clear to the users of the decision models. The results indicate the importance of a differentiated approach when applying decision models in supplier selection processes. For example, differences regarding the quantity and quality of available decision information and the financial impact of the supplier selection decision clearly require different types of decision models. This implies that in future research, the applicability of a novel decision model should be specified more clearly than has been the case so far. Finally, the analysis of the evaluation of the experiments suggests that organisation-wide implementation is not a straightforward matter. Clearly, most decision-makers need to be trained in order to be able to effectively use most of the decision models. Another challenge is to strike an appropriate balance between (a) the variety of decision models required to cope with the many different purchasing situations in terms of importance and complexity and (b) the overall cost of implementing and maintaining a set of decision models in an organisation which clearly will rise as the number of different decision models used in the firm increases. This points to two important and interrelated directions for further research on decision support for supplier selection. The first direction concerns the problem of trying to identify and develop and empirically investigate a number of small sets of different decision models per phase in the supplier selection process that may meet the challenge pointed out above. Secondly, this challenge also requires us to investigate the strong emergence of web-based provision of decision models and related know-how (see for example Geoffrion and Krishnan, 2001), which increasingly offers opportunities for buying decision support for supplier selection rather than developing all support in-house.