|
Article Excerpt
1. Introduction
In business-to-business markets, one observes a range of horizontal scope strategies. Some firms pursue a "one-stop shop" strategy and expand their scope of activities to encompass a broad set of products or services. Other firms take a different tack and specialize in a limited array of products, striving to be recognized as the best supplier of a particular type of service.
At least two opposing forces are at play in such scope decisions. On one hand, there are benefits of focus due to organizational tradeoffs. Porter (1996) argues that such organizational tradeoffs require firms to specialize if they are to achieve competitive advantage (see also Siggelkow 2002). On the other hand, offering a broad set of services allows a supplier to do more business with a particular buyer and hence develop a deeper relationship. Strong ties with buyers provide many possible benefits, such as improved cooperation and better sharing of information (Dyer and Singh 1998, Uzzi 1999). That is, strong ties may create client-specific scope economies that favor broad scope strategies.
Hence, there is a fundamental tension in many markets between specializing to reap the benefits of focus and being more of a generalist to reap the benefits of deeper relationships (Siggelkow 2003). The environments that are conducive to specialist and generalist organizational forms is an important theme in organizational theory coming out of the population ecology literature (Hannan and Freeman 1977, Carroll 1985).
In contrast to the management and organizational literatures discussed above, formal work on horizontal scope has emphasized scope economies in production. In their classic work on the topic, Panzar and Willig (1981) define scope economies as occurring when it is less costly to produce two or more product lines in one firm than to spread the production across specialized firms. The idea that scope economies might arise from interactions with specific buyers is missing from this theory. (1) Prior work on the effect of such supply-side scope economies on firm specialization finds that the number of products places an upper bound on the set of scope strategies (MacDonald and Slivinski 1987, Eaton and Lemche 1991). Thus, with two products in a market, there can be at most two scope strategies.
We develop a formal model of scope decisions that is closer to the management and organization theory literatures. We assume that there are organizational tradeoffs that create diseconomies of scope in production. However, we also allow for benefits from building stronger buyer-supplier relationships by offering a broad product range. That is, we allow for client-specific scope economies. We study how the use of different scope strategies varies with industry structure: When do industries support the use of generalist strategies? When do they support specialists? When do generalists and specialists coexist? We consider classic elements of industry structure, such as barriers to entry and bargaining power, as well as more novel elements, such as the size of client-specific scope economies, the extent of organizational tradeoffs, and the extent to which buyers have heterogeneous task requirements.
Our paper is one of the first to employ the formalism of biform games (Brandenburger and Stuart 2007) for the formal study of strategy. In a biform game there is an initial stage in which players make decisions that affect their ability to create value when working with other players. This stage is analyzed using the standard noncooperative game theory that is common, for example, in the modern industrial organization literature. In a second stage, cooperative game theory is used to characterize the outcome of bargaining among the players over how to split the total surplus. A biform game is well suited to problems such as ours, where suppliers make initial entry and organizational design decisions and then negotiate with buyers over the fees to be paid on a task-by-task basis. (2) The use of cooperative game theory in strategy has been advocated by Brandenburger and Stuart (1996) and Lippman and Rumelt (2003); MacDonald and Ryall (2004) are the first to explicitly develop formal foundations of strategy using cooperative game theory.
1.1. Discussion of the Phenomenon
One often observes a mix of scope strategies in a given industry. For instance, in investment banking, Lazard's business is mostly mergers and acquisitions (M & A) advice, while Morgan Stanley offers a comprehensive range of services. In the enterprise software industry, Oracle built its success on database management, but IBM has had an integrative approach of offering "solutions" to its customers. Some law firms position themselves as generalists, and others position themselves as "boutiques" focused on a single area of law. In between these extremes are hybrid strategies. For example, some London law firms with a leading reputation in a very specific field, such as real estate or intellectual property, complement that focus with a secondary capability in corporate law (see Lewis 2005).
The nature and extent of organizational tradeoffs vary across industries. In his classic work on professional service firms, Maister (1993) argues that different service projects are best executed by different types of organizations. For example, complex corporate strategy projects are best executed by "greyhair" consulting firms such as McKinsey, and information technology (IT) implementation projects are best executed by "procedural" consulting firms such as Accenture. More generally, firms with a broader scope often suffer from agency costs (Jensen 1986) and influence activities (Milgrom 1988).
We study buyers that seek to outsource two tasks. These tasks could be anything from designing and supplying different types of components, as in the automotive industry, to delivering different professional services such as M & A advice and bond placements. We assume that the tasks are not too different so that they can feasibly be offered by the same organization (e.g., both tasks involve financial services or legal advice). We make the simplifying assumption that firms are not constrained by capacity. In the case of professional service firms, this would hold if they could hire people to staff projects after winning the business. Similarly, this would hold for an automotive supplier that can build capacity after winning a contract to supply a component.
We assume that there may be a benefit from using a single supplier for both tasks. Such client-specific scope economies arise from the more frequent interactions between the buyer and supplier. For example, a supplier learning about the buyer from performing one task could lower costs or improve quality on the other task. The benefit could also arise from improved coordination across the tasks when they are done by the same supplier, which is an argument made, for instance, in the enterprise software market. (3)
2. An Extended Analytic Example
We introduce our formal treatment of scope strategies with an extended analytic example. Consider a situation where there is a single buyer and three available suppliers. We index the supplying firms by i = 1, 2, 3 and refer to the buyer as firm b. The buyer has two tasks, labeled A and B, that it cannot do itself. The buyer negotiates simultaneously with the suppliers to determine which supplier or suppliers to which to outsource the tasks and under what terms. All the firms seek to maximize their share of the surplus created by outsourcing the tasks.
The surplus created by giving a task to a particular supplier is the difference between the buyer's willingness to pay (WTP) for using that supplier and the supplier's cost of executing the task. All suppliers have a cost of 60 for executing one task and 120 for executing two tasks. The suppliers differ in their competencies in executing each task, which leads to differences in buyer WTP, as given in Table 1.
Thus, Supplier 1 generates a surplus of 160 - 60 = 100 when doing task A and a surplus of 120 - 60 = 60 when doing task B. In this example, one can think of Supplier 1 as a specialist in task A, Supplier 2 as a specialist in task B, and Supplier 3 as a generalist.
We follow Brandenburger and Stuart (1996, 2007) in using cooperative game theory to analyze the negotiations among the firms. Brandenburger and Stuart (1996) point out that a firm's added value places an upper bound on its payoff from the negotiations. A firm's added value is the surplus that is lost if the firm were to be removed from the negotiations. It requires that one calculate the maximum surplus with all firms and then subtract the surplus without the focal firm.
With all firms, the surplus is 200, which comes from giving task A to Supplier 1, task B to Supplier 2, and no task to Supplier 3. The surplus without the buyer is 0; hence, the buyer's added value is 200. The surplus without Supplier 1 is 190, which comes from using Supplier 3 for task A and Supplier 2 for task B. Hence, the added value of Supplier 1 is 10. The added value of Supplier 2 is also 10. Supplier 3 has no added value, because surplus is still 200 if it leaves the negotiations.
Although intuitively appealing, added value is only an upper bound on what a firm can negotiate. Notice that the sum of the added values in the example is 220, which is greater than the total surplus generated. A standard approach in cooperative game theory is to solve for the core. The core satisfies two properties: The maximum surplus is divided among the firms, and no subset of firms, can increase its share of surplus by withdrawing and just transacting among themselves. Two drawbacks of the core are that it can be empty and that it often yields a range of possible payoffs.
Not having a unique payoff is a problem if one wants to study strategy decisions prior to the negotiations. Brandenburger and Stuart (2007) suggest the following. Solve for the full set of core allocations, which yields a minimum and a maximum allocation for each player. Assume that players expect to get a convex combination of these bounds. That is, firm i's expected surplus from the negotiations is [[alpha].sub.i][[pi].sub.i.sup.max] + (1 - [[alpha].sub.i])[[pi].sub.i.sup.min], where [[alpha].sub.i] [member of] [0, 1] is the weighting, [[pi].sub.i.sup.max] is firm i's maximum core allocation, and [[pi].sub.i.sup.min] is its minimum core allocation. Brandenburger and Stuart (2007) refer to the [[alpha].sub.i] as confidence indices that reflect a player's subjective assessment of its bargaining skill.
In this example, we assume that [[alpha].sub.i] = 1/2 for all firms. The set of possible core allocations for each firm and its expected share of surplus is given in Table 2.
Note that competition among the suppliers drives most of the surplus to the buyer, because Supplier 3 is a close substitute for the other suppliers. A convenient property of our example is that a supplier's expected share of the surplus is proportional to its added value. That is, with [[pi].sub.i.sup.min] = and [[pi].sub.i.sup.max] equal to the added value (A[V.sub.i]), a supplier's expected surplus is just [[alpha].sub.i]A[V.sub.i]. This is not necessarily the case for biform games in general. (4)
We now consider initial entry decisions by the suppliers. Suppose that there is a fixed cost of F = 3 required to serve the buyer that must be incurred prior to the negotiations. Though Suppliers 1 and 2 expect to cover this fixed cost, Supplier 3 does not. Hence, all firms entering cannot be a Nash equilibrium of the biform game. What happens if only Suppliers 1 and 2 enter? Table 3 gives the new bargaining outcomes.
The elimination of competition from Supplier 3, which was the next best alternative for both tasks, has significantly raised the added value of the remaining suppliers, to the detriment of the buyer's ability to capture surplus.
We now introduce client-specific scope economies to the example. Specifically, suppose that some extra surplus is created when the same supplier performs both tasks. (5) Denote this extra surplus by R > 0. Thus, using...
|
