استراتژی های خرید در ماتریس Kraljic ؛ چشم انداز وابستگی و قدرت

Kraljic's purchasing portfolio approach has inspired many academics to undertake further research into purchasing portfolio models. Refined models typically recommend one purchasing strategy for each portfolio quadrant. Yet, it has been shown that purchasers make a clear distinction between alternative purchasing strategies within each quadrant. The fundamental assumption of portfolio models seems to be that differences in power and dependence between buyers and suppliers exist. Still, little is known about how these concepts influence the choice for a specific purchasing strategy. In this paper, ‘relative power’ and ‘total interdependence’ for a number of portfolio-based purchasing strategies have been quantified empirically, using data from a comprehensive survey among Dutch purchasing professionals. The survey data largely confirmed the hypotheses that were deduced from the literature.

Purchasing portfolio models have received much attention in the recent literature about professional purchasing. One of the most famous portfolio models was introduced by Kraljic (1983). His model has had a broad influence on professional purchasing (see the evidence of Kamann and Bakker, 2004; Gelderman, 2003). It also inspired many academic writers to undertake further research into portfolio models (e.g. Gelderman and Van Weele, 2002, Gelderman and Van Weele, 2003 and Gelderman and Van Weele, 2005; Dubois and Pedersen, 2002; Zolkiewski and Turnbull, 2002; Nellore and Soderquist, 2000; Wynstra and ten Pierick, 2000; Croom, 2000; Bensaou, 1999; Lilliecreutz and Ydreskog, 1999; Olsen and Ellram, 1997; Wagner and Johnson, 2004; Dyer et al., 1998). According to Kraljic (1983) a firm's supply strategy depends on two factors: (1) profit impact and (2) supply risk. Other scholars have introduced variations of the original Kraljic matrix (e.g. Elliott-Shircore and Steele, 1985; Syson, 1992; Hadeler and Evans, 1994; Olsen and Ellram, 1997). The proposed matrices are very similar to the Kraljic matrix in that they employ comparable dimensions, and derive equivalent recommendations. Typically, these matrices give only one recommendation for each portfolio quadrant, namely: form partnerships for strategic products; assure supply for bottleneck products; exploit power for leverage products and ensure efficient processing for non-critical products. However, a recent study into the actual use of the matrix by professional purchasers pointed out that purchasers make a clear distinction between several strategies within each quadrant ( Gelderman and Van Weele, 2003). Purchasers identify: (1) strategies to hold their position in the quadrant and (2) strategies to move to another position. However, the conditions determining the choice for a specific purchasing strategy within a quadrant are yet unclear. The fundamental assumption of all portfolio models seems to be the occurrence of differences in power and dependence between buyers and suppliers (Dubois and Pedersen, 2002). Kraljic (1983) does not explicitly deal with issues of power and dependence. However, some of his recommendations obviously refer to the power structure (‘exploit power’). Others are aimed at reducing the dependence on suppliers (‘diversify’). Moreover, Kraljic (1983, p. 112) stated that the general idea of the portfolio approach is to “minimize supply vulnerability and make the most of potential buying power”. It seems that power and dependence play a significant role in the Kraljic approach. The relative power and dependence position of buyers and suppliers are therefore expected to be factors of importance in explaining the conditions that influence the choice of purchasing strategy within each quadrant. Until now little is known about the conditions that influence the choice of purchasers to either hold their position in the quadrant or move. Moreover, little is known about the way in which power and dependence in buyer–supplier relationships enter the Kraljic matrix (Gelderman and Van Weele, 2003; Dubois and Pedersen, 2002). Empirical research on the impact of power and dependence on buyer–supplier relationships is even scarcer. Therefore, it is critically important to examine the power and dependence positions of buyers and suppliers for the various purchasing strategies that have been identified in each quadrant of the portfolio matrix. The goal of this paper is to empirically test several hypotheses that are deduced from the literature on power and dependence. This will be done with respect to the purchasing strategies in each quadrant of the Kraljic purchasing portfolio matrix. In order to do this we have defined the concepts of power and dependence in terms of buyer's and supplier's dependence. Subsequently, we have developed constructs for buyer's dependence as well as supplier's dependence. Our empirical analysis is founded on a survey among 250 purchasing professionals. On the basis of the survey data we have assessed power and interdependence in buyer–supplier relationships. In general terms this study contributes to a better understanding of the (perceived) power and interdependence in buyer–supplier relationships. The organization of the paper is as follows. First we will give a brief overview of the Kraljic approach and, on the basis of recent literature, we will identify hypotheses with respect to power and dependence for each strategy used in practice (Section 2). Furthermore, we will operationalize power and dependence into measurable variables and formulate hypotheses (Section 2). In Section 3 we will present our survey design and design constructs for our key variables. The results of our survey are presented in Section 4. Section 5 will conclude and give suggestions for further research.

Kraljic's approach has inspired many academics to undertake further research into purchasing portfolio models. Refined models typically recommended one purchasing strategy for each portfolio quadrant. Gelderman and Van Weele (2003) found that professional purchasers make a clear distinction between alternative strategies within each quadrant. The conditions determining the choice for a specific purchasing strategy within a quadrant remain yet unclear, but might very well be associated with differences in power and dependence positions. In this paper, we have empirically quantified the concepts of ‘relative power’ and ‘total interdependence’ for each purchasing strategy, using data from a comprehensive survey among Dutch purchasing professionals. This study builds on the case studies of Gelderman and Van Weele (2003) by adding further insights into the power and interdependence structure of the various portfolio-based purchasing strategies. 5 The results of our study prove that there is a significant difference in the power positions between the purchasing strategies within each quadrant. Fig. 3 summarizes the results, showing the relative power positions that are associated with each purchasing strategy. Note that the diagonal line indicates positions in which buyer's dependence is equal to supplier's dependence, i.e. it indicates power balance. Strategies on the left-hand side of the diagonal line are characterized by buyer dominance, while strategies on the right-hand side are characterized by supplier dominance. The length of the perpendicular line from each data point towards the diagonal line indicates the size of the power difference between buyers and suppliers. Full-size image (13 K) Fig. 3. Power map. Figure options Comparing strategies aimed at moving towards other positions in the Kraljic matrix (scenarios 3, 5, 7 and 8) with strategies aimed at maintaining the current position (scenarios 1, 2, 4, 6 and 9) leads to the following conclusions. The strategies in the strategic and the bottleneck quadrant which are aimed at moving (strategies 3 and 5) imply a reduction of the supply risk. Their positions in Fig. 3 indicate that these strategies are pursued in the case that buyer's dependence is relatively low. The level of supplier's dependence seems to be less critical to the choice for these strategies. Furthermore, buyers usually prefer a strategy aimed at maintaining their position in the leverage quadrant of the Kraljic matrix (scenario 6). Our findings suggest that relatively high levels of both buyer's and supplier's dependence constitute conditions for engaging in a partnership (scenario 7) and thereby follow a strategy aimed at moving to another quadrant. Finally, no significant differences were found between scenarios 8 and 9 in the non-critical quadrant. Hence, the levels of buyer's and supplier's dependence do not seem to influence the choice for changing or maintaining the position in this quadrant. Note that the differences of purchasing strategies in the power map within each Kraljic quadrant are smaller than the differences between the quadrants. For instance, the strategies in the non-critical quadrant (scenarios 8 and 9) are located closely together in the power map, while the other strategies are at a relatively large distance. Therefore, we can conclude that the levels of buyer's and supplier's dependence largely determine the position of a purchasing strategy within the Kraljic matrix. A final conclusion from Fig. 3 is that positions in the bottleneck quadrant and the strategic quadrant of the matrix (scenarios 1–5) are associated with buyer-supplier relationships, which are characterized by supplier dominance. On the other hand, positions in the leverange quadrant and the non-critical quadrant (scenarios 6–9) are perceived to be of a more power-balanced nature. Apparently, from the buyer's perspective the level of supply risk is strongly associated with the perceived power balance between buyer and supplier. A comparison between our empirical findings and the theoretical expectations generate the following conclusions (see also Appendix B). All hypotheses regarding the power balance associated with each purchasing strategy within the quadrants (1a–9a) were confirmed, except hypothesis 1a. That is, we observed supplier dominance in the strategic quadrant when a strategic relationship is maintained (scenario 1), where one would expect a balanced power situation on basis of the literature. This provocative result sheds new light on the buyer's view on issues of power and dependence. It indicates that a relationship which is characterized by a high involvement of buyers and suppliers does not necessarily imply a balanced power position between the parties, but can yet be satisfactory, at least from the point of view of the buyer. All hypotheses concerning the relative power positions of the different strategies within each quadrant (1b–9b) and the level of total interdependence (1c–9c) were confirmed. Despite the rigor of our analysis, our study contains several limitations that might entice further research. One of the limitations concerns the sample, which was drawn from a list of members of the Dutch Association of Purchasing Management (NEVI). Although our sample included a wide range of industry sectors, the generalizability of the results would benefit from the inclusion of firms in the service sector, as well as other Dutch and international companies. This might be the object of a further study. Another limitation concerns the fact that the survey was confined to the perspective of the buyer. Yet, buyer–supplier relationships are dyadic in nature. Suppliers might have different opinions on the power and interdependence structure of the various buyer–supplier relationships. Therefore, there is a need to more fully study the supplier's perspective in order to establish whether or not both parties perceive each other's power position in the relationship in the same way. In addition, further research should take a more holistic supply chain perspective on buyer–supplier relationships and acknowledge the possible influence of critical external forces, such as regulation, standards and international trade barriers.