Relationship Portfolios - Past, Present and Future

Since the 1980s the concepts of portfolio management have been adopted within the context of understanding business markets. Academics on both sides of the Atlantic have conceptualized and empirically tested a variety of customer and supplier portfolio models. More recently the concept of an indirect portfolio of relationships and its management has also been introduced. This paper reviews the role of relationship portfolios: customer, supplier and indirect in the context of business-to-business markets. We critically assess the most significant models, map their evolution and consider what the future holds for relationship management. We then discuss the manner in which portfolios and network can be integrated in order to provide a practical guide for marketing management. In conclusion we recognize the importance of conceptualizing the network as a set of portfolios (customer, supplier and indirect) and suggest that, in the context of business-to-business marketing at the very least, portfolio analysis provides the key to successful relationship management and important inputs to strategic management. Introduction The origins of portfolio theory lie in financial investment (Markowitz, 1952 and Sharpe, 1963). The concept has also been widely adopted in other spheres of management such as strategic management and marketing, as a mechanism for aiding decisions about resource allocation. No strategic management or marketing text appears to be complete without the inclusion of the Boston Consulting Group (BCG) growth share matrix or the McKinsey model. When used effectively, these models provide guidance for resource allocation and the BCG model, despite its inherent weaknesses, is probably one of the most widely used management decision aids. The concepts of portfolio management have also been adopted within the context of understanding business markets. Portfolios provide a mechanism for conceptualizing and managing the customer, supplier and indirect sets of relationships which surround a firm. The growth of interest in 'relationship' marketing has also put greater emphasis on the study of these relationship portfolios. This paper is written very much from a business marketing perspective. Specifically, because in this arena the role of relationship management is acknowledged to be critical to gaining competitive advantage (Håkansson, 1982, Turnbull and Valla, 1986, Ford, 1990, 1997 and Sheth and Sharma,1997). The paper reviews work on both sides of the Atlantic in this area. It critically assesses the significant models, maps their evolution and considers what the future holds for relationship portfolio management. Relationship Portfolio Models The bulk of models that have been conceptualized are based in either customer or supplier relationship modelling. However, Zolkiewski and Turnbull (1999) also raised the importance of modelling the set of indirect relationships which surround a firm. (Indirect relationships are those with actors who influence the operation of the organization and can include Government (national and local), institutions such as universities and lobby groups, for example.) The models which have been developed include both two and three-dimensional axes along with single, two and three-step analysis phases. They are listed in Table 1 below. The most significant of these models are then reviewed in the following sections. Table 1 Summary of Portfolio Models Year Customer Portfolio Models Supplier Portfolio Models Indirect Portfolio Models 1982 Cunningham and Homse Cunningham Fiocca 1983 Campbell and Cunningham Dickson (Distributor portfolio analysis) 1984 Dubinsky and Ingram 1987 Shapiro et al. 1991 Krapfel, Salmond and Spekman 1992 Rangan et al. 1994 Yorke and Droussiotis (modified Fiocca model) 1997 Turnbull and Zolkiewski (composite of Shapiro et al and Krapfel, Salmond and Spekman) Olsen and Ellram 1999 Zolkiewski and Turnbull Fiocca (1982) Two-Step Customer Portfolio Analysis Fiocca (1982) proposed a two-step customer portfolio analysis, see Figure 1, and argued that the selling organization should firstly analyze customers at a general level, according to the strategic importance of, and the difficulties in, managing the relationship with each customer (account). The second step of analysis requires another two-dimensional matrix to be constructed for the key accounts identified in step 1, with the customers ' business attractiveness' on one axis and the 'strength of the supplier customer relationship' on the other. Fiocca (1982) suggests a number of mechanisms for assessing the proposed axes: 'Difficulty in managing the customer' is a function of the level of competition for the customer, customer buying behaviour and the characteristics of the product bought by the customer. 'Strategic importance' is determined by the value/volume of purchases, the potential and prestige of the customer, customer market leadership, and the overall desirability to the supplier in making strategic improvements and adaptation to customer specifications. This mixture of subjective and actual values makes these calculations difficult especially when the main point of using such analysis should be to produce data which can be used for comparison. 'Business attractiveness' is determined by considering a number of factors that are related to the customer's market (growth rate, competition, maturity, changes in the environment, etc.) and the status/position of the customer's business within the market. Such calculations are particularly difficult to assess and Fiocca (1982) does not take into account factors which can be critical in doing business internationally such as distance and cultural factors. The strength of supplier/customer relationships is again measured by applying a mix of objective, judgmental or subjective factors: length of relationship, importance of the customer, friendship, co-operation in product development, social distance, etc. A criticism of the Fiocca (1982) model put forward by Yorke and Droussiotis (1994) is that it does not recognize the importance of considering customer profitability and, in fact, simply assumes that different cells can be associated with different levels of profitability. This assumption that customers are profitable simply because management perceive them to be was identified by Turnbull and Zolkiewski (1997) as a general problem in much analysis. In reality, customers were often found to be not as profitable as managers believed them to be (once full account was taken of real selling costs). These reservations have been confirmed by Turnbull and Topcu (1994) who tested the Fiocca (1982) matrix using detailed case study data from a Turkish industrial minerals manufacturer. They identified a number of problems with data calculation. Firstly, the axes scales (high and low; high, medium and low; and, strong, medium and weak) tend to be subjective. Secondly, they used different interpretations of 'Difficulty in Managing the Account' (one based upon the amount of problem solving needed and the other based upon relative service requirement) and 'Relative Buyer Seller Relationship' (where they used Ford's (1982) life-cycle stages as a proxy measure). Finally, they used a proportion of sales revenue to indicate the strategic importance of the customer. This is questionable because there are many other factors which can contribute to strategic importance, such as technical leadership, which need consideration. Figure 1, Two-Step Customer Based Portfolio Analysis (Fiocca, 1982, page 56) Not surprisingly, Turnbull and Topcu (1994) found that the use of different measures for each of the two dimensions, suggested by Fiocca (1982), resulted in significant disparities in 'categorization' of customers. This was particularly marked in relation to the definition of non-key customers. However, they demonstrated that a number of non-key customers required disproportionately high levels of service from the supplier. It could also be postulated that some of the key customers were not receiving enough service. They also demonstrated that the matrix provides a great deal of valuable information which managers should consider when forming strategy. Step 1: Key Difficult Non-Key Difficult Key easy Non-Key Easy High