A Study of Sourcing Channels for Electronic Business Transactions. - Byungjoon

Journal of Management Information Systems / Fall 2011, Vol. 28, No. 2, pp. 145–169.

© 2011 M.E. Sharpe, Inc.

0742–1222 / 2011 $9.50 + 0.00.

DOI 10.2753/MIS0742-1222280206

A Study of Sourcing Channels for Electronic Business Transactions

ByuNgjOON yOO, VIDyaNaND ChOuDhary, aND TrIDaS MukhOpaDhyay

Byungjoon yoo is an associate professor at the graduate School of Business at Seoul National university. Before that, he was on the faculty of korea university and hong kong university of Science and Technology. he received his ph.D. in information systems from Carnegie Mellon university. his research interests are in business-to-business e-commerce, online auctions, and pricing strategies of digital goods such as software products and online games. his research on these subjects have been published in journals such as Journal of Management Information Systems, International Journal of Electronic Commerce, Management Science, and Decision Support Systems.

Vidyanand Choudhary is an associate professor at the paul Merage School of Business at the university of California, Irvine. he received his ph.D. in management from purdue university. his research interests are in the areas of economics of informa-tion systems, marketing strategy for technology products, and governance. Specific topics of interest include information goods pricing strategy, product-line strategy, economics of cloud computing, impact of a software-as-a-service business model, and outsourcing strategy.

Tridas Mukhopadhyay is the Deloitte Consulting professor of E-business at Carnegie Mellon university. he received his ph.D. in computer and information systems from the university of Michigan. his research interests include business value of information technology, strategic use of information technology, business-to-business commerce, and software development productivity

aBsTraCT: There are two popular forms of business-to-business (B2B) marketplaces: public marketplaces and private channels. We study why firms choose either or both of these sourcing channels. using a framework of decision making under uncertainty, we explain firms’ choice of B2B channels as a hedging strategy and as a method of obtaining greater managerial flexibility for the future. We show that greater uncertainty can lead to higher investment with firms more likely to invest in both public and private channels. We find that the level of information technology (IT) capability and spending is an important factor in firms’ decision making. When a firm chooses its level of IT investment simultaneously with the decision about which sourcing channels to use, the firm choosing both channels selects the highest level of IT capability and the firm implementing only one channel selects lower levels of IT capability.

key words and phrases: analytical modeling, B2B e-commerce, decision making under uncertainty, economic theory, IT capabilities, managerial decision making, private channels, public marketplaces.

146 yOO, ChOuDhary, aND MukhOpaDhyay

The gloBal supply Chain faCes Many unCerTainTies as illustrated by several recent events in 2011, such as the earthquake and tsunami in japan, popular uprisings in several Middle Eastern countries, and civil war in Libya. prior research has addressed most sources of uncertainty but has not examined the uncertainty due to the high failure rate of electronic business-to-business (B2B) marketplaces such as MetalSite [7]. B2B marketplaces are susceptible to coordination failure for several reasons. Some larger firms may prefer not to share the benefits of a common marketplace with competi-tors and therefore divert their transactions to a private channel. In addition, buyers or sellers may not trust the public marketplace, especially when it is owned by the other party [37]. Both of these factors played a role in the failure of the original business model of a high-profile public marketplace for the automotive industry—Covisint [15]. The resulting high failure rate of public marketplaces imposes uncertainty about the future availability of this sourcing channel. In this paper, we examine firms’ choice of B2B sourcing channels under uncertainty when they have a choice of using either a B2B marketplace or a Web services–based private channel or both.

The uncertainty about B2B marketplaces was highlighted recently when american airlines stopped providing its fare data to Orbitz, an online travel agency. By early 2011, the battle had grown more intense and several other players had become em-broiled in the dispute, including Expedia and Sabre [12]. american airlines and other carriers seek to have greater control of their retail relationship with customers and have greater ability to compete not just on prices and schedules but also on services [25]. Thus, airlines are weighing the costs and benefits of using a public marketplace such as Orbitz and Expedia versus a private portal—their own Web site. Many firms face similar choices about their sourcing channels: (1) they can choose to join a public marketplace, such as ariba’s public marketplace for procurement, which has numerous clients, including Staples; (2) they can develop a private channel, such as getSupply, which was built by hewlett-packard (hp) for transacting with its supplier base; or (3) they can do both. For example, DaimlerChrysler and general Motors (gM) were founders and partial owners of Covisint in 2000, yet both were simultaneously devel-oping their private channels [15]. Similarly, British airways participates in ariba’s public marketplace while also using private channels for e-procurement [10], and the university of California, Irvine, uses a private purchasing portal called “uCIBuy” while at the same time using public marketplaces, such as amazon.com and Expedia [28].

The choice of B2B sourcing channels is becoming increasingly important, with firms focusing on their core competency and outsourcing nonstrategic parts of their business. The revenue from B2B electronic commerce is expected to grow from $0.8 trillion in 2002 to $5.1 trillion in 2013 [22]. B2B marketplaces facilitate the meeting of buyers and suppliers and provide infrastructure for B2B transactions. however, the uncertain-ties associated with B2B marketplaces are similar to those with business-to-consumer (B2C) e-commerce in the 1990s. Of the 1,500 marketplaces that existed in 2000, only about 200 marketplaces survived in 2008 [22].

Electronic B2B channels can be categorized into two types: public marketplaces and private channels based on Web services. a private Web service is typically owned by a firm and can be used only by its business partners. a public marketplace is owned

a STuDy OF SOurCINg ChaNNELS FOr ELECTrONIC BuSINESS TraNSaCTIONS 147

by a neutral third party or many firms, and used by them and their partners. While public marketplaces allow firms to save on procurement costs by drawing on a larger base of suppliers and buyers, the strength of the private Web service channel is the efficiency of customized transaction processes for only one supplier or buyer. Table 1 summarizes the features and differences among private and public sourcing channels discussed in many articles [9, 18, 22, 26].

prior research has analyzed several factors associated with B2B business models, such as the size of the firm and the fragmentation level of industries; however, strategic uncertainty about the future has not been considered in previous research on elec-tronic B2B channels. Even though B2B marketplaces are valued for their enormous potential to increase firms’ profits, opportunistic behavior by some firms has also been observed [26]. Some firms cooperate in creating public marketplaces while simultane-ously establishing and using their own private channels based on Web services, thus reducing their commitment to the success of public marketplaces. This behavior is partly due to the strategic uncertainty inherent in public B2B marketplaces. Because firms cannot predict whether their competitors will cooperate in using public B2B marketplaces or compete by using private channels, the innovative firm may choose to do both. This can be viewed as a hedging strategy to reduce risk.

This study seeks to examine firms’ choice of B2B sourcing and incorporate the uncertainty about future success of B2B marketplaces. There are other factors that also play a role in the choice of B2B sourcing channels. For example, firm size plays a role because building a private exchange requires significant investment of capital. johnson et al. [17] conducted an empirical study on electronic procurement and report that firm size is a significant determinant of the use of certain e-business technologies. another important factor is the level of information technology (IT) capability in the

Table 1. Comparison of public Marketplaces and private Web Services

B2B sourcing channels

public marketplaces private Web services

Characteristics in common

1. Both enhanced by general information technology2. Both use electronic technology in B2B transactions

Differences 1. Owned by neutral third party or consortium and open to everyone in the industry

2. Benefits from network effects3. Better for commodity-like

goods4. Enhanced by technology

that reduces the possibility of private information being released

5. Requires cooperation among many firms

1. Owned by a firm and its partners and used exclusively for them

2. More focus on customization3. Better for more complex and

specific goods4. Enhanced by technology that

increases efficiency through customization

5. Requires cooperation among few firms


firm because it may impact the firm’s ability to execute projects relating to electronic procurement. Mishra and agarwal [24] examine survey data on electronic procurement and find that technological capability is a significant moderating variable. The transac-tion cost economics view of the firm propounded by Williamson [35] argues that asset specificity is an important consideration in transactions between firms. Therefore, we include the following factors in our study: the size of the firm, its IT capability, and the specificity of the product. We develop a framework for understanding a firms’ decision to adopt a B2B marketplace by combining a game-theoretic analysis with the embedded option perspective [19] in the economics of decision making under uncertainty. We find that firms with higher IT capability make greater use of both public marketplaces and private Web services and firms that make use of both kinds of channels invest more in IT. When the level of IT capability is chosen simultaneously with the decision on B2B channels, firms that choose to implement both channels select the highest level of IT capability and firms that implement only the private channel select the lowest level of IT capability. We also show that the existence of risk aversion offsets the preference for using public marketplaces.

Next we provide a literature review of related studies. We then describe the model, and following that we present our results. Finally, the contribution of this study and directions for further research are discussed in the last section.

Literature review

There is signifiCanT liTeraTure on The eConoMiC forCes that affect organizational structure. Coase [3] points out that as transaction costs are reduced, the benefits from markets increase. Williamson [35] introduces the importance of transaction cost eco-nomics in firms’ organizational choices under different conditions. he explains the impact of transaction cost economics in determining the form and size of organizations with the concept of “asset specificity.” researchers studying B2B e-commerce have recognized the role of transaction costs in this field.

recent research into B2B e-commerce includes Walden and Subramani [33], who find empirical evidence to support the value of B2B e-commerce. They identify the extent of the positive effect of B2B e-commerce on different types of firms—pure Internet firms and brick-and-mortar companies—and on different types of products—digital products and tangible products. kauffman and Mohtadi [21] consider the size of firms to be an important factor in choosing between private marketplaces, which are exclusively owned by a supplier or a buyer, and public marketplaces, which are open to all suppliers and buyers in the industries. They expect larger firms to use private marketplaces. Dai and kauffman [5] find that the desired gains from lower search and operation costs, the importance of information sharing, and competition are important factors in deciding between public and private marketplaces. kauffman and Li [20] analyze the trade-off between the value of preemption with positive network effects and the value of waiting for technology uncertainty to be resolved.

While there are several studies of B2B channels, strategic uncertainty about the future has not been considered in research related to B2B channels. B2B channels are normally praised for their potential to increase the profit of firms, but opportu-


nistic behaviors by some firms have also been observed [26]. although opportunistic firms claim to cooperate in creating public marketplaces, the uncertainty about B2B public marketplaces leads them to also use private channels based on Web services. They do so because they are unsure whether their competitors will cooperate in the public marketplaces or act individually against each other. Firms cannot be certain of the liquidity provided by other firms in public B2B marketplaces. unlike the uncertainty of supply and demand, this strategic uncertainty has not been studied in B2B marketplaces.

a firm that simultaneously develops private and public channels can be viewed as adopting a hedging strategy to manage risk. prior literature has used this view to evaluate capital investments as real options, especially when there is significant uncertainty about the future and when investments open up more alternatives in the future [27]. Van alstyne [32] discusses the value of information itself as an option, with the cost of action as its strike price. an investment in IT is a good example of an investment that can be seen as an option because it involves a great deal of uncertainty at the time of the investment. The value of managerial flexibility can be interpreted as an option for hedging risk due to uncertainty. This value is called a “real option” to differentiate it from a “financial option.” Different categories of strategies can be seen as real options in different ways. roberts and Weitzman [30] consider investments at the early stage as a real option that is valuable because it allows firms to collect information from early investments and reduce uncertainty for further investments. McDonald and Siegel [23] regard the quick investment as a real option because it allows investors to gain great benefits when the investment is successful but lose just the amount invested if unsuccessful.

The real options approach has been used by several information systems research-ers. Dos Santos [8] asserts that the real options approach is better than traditional investment evaluation methods such as net present value (NpV) because it is able to consider the value of managerial flexibility. Fichman [11] applies the real options view to evaluate IT investments in IT platform technologies. he discusses characteristics of projects and organizations that affect the option value of investments in technologies. Dai et al. [6] use an analytical model to explain the value of IT investments as provid-ing potential leverage for further applications. Benaroch and kauffman [2] provide an example of the evaluation of IT investments using real options. herath and herath [16] show how to apply real option techniques in evaluating investments on information security. They develop an integrated real options model by using Bayesian statistics. Tallon et al. [31] review the status of IT research on the real options approach and discuss research questions for further study. They suggest that future research should consider decision, simulation, and game-theoretic models to provide insights into IT investment processes. kauffman and kumar [19] discuss the “embedded option” nature of IT investments because IT investments and options are not separable in many cases. In our model, the investments have this “embedded” nature because the investments on channels are not separable from the option.

By considering the uncertainty and network externalities in B2B channels as well as various possible scenarios, our research seeks to explain current phenomena in the real world and the likely direction of B2B channels in the future.


Model

following The CaTegorizaTion of kaplan and sawhney [18], in this study we focus on manufacturing-related products and processes in individual industries that con-stitute 90 percent of total procurement costs [4, 36] and not on supporting products and services for operating supplies. For common supporting products and services, we expect there to be horizontal marketplaces across industries instead of vertical marketplaces for each industry.

Our model considers several factors at three levels: at the product, firm, and industry levels. For example, the specificity of a product (θ) is a product-level factor. a highly specific product can only be used for particular purposes, whereas products with low specificity, such as commodity products, can be used for general purposes. Firm-level factors include firm size (n) and expertise in IT (λ) that can provide assistance in using and adjusting to the external environment. Customization and information security (e) are also firm-level factors that tend to be higher in private channels based on Web services and are key advantages of this channel. Since private channels are available only to the firm and its business partners, the focal firm’s business processes can be easily customized. also, information related to the business processes can be more secure with a private channel because the marketplace is used exclusively by the firm and its partners.

at the industry level, the current status of B2B marketplaces in the industry affects the firm’s choice of a business model. The expected size of the public marketplace (N) is an important industry-level factor. The greater the size of a public marketplace in an industry, the more likely a firm is to choose it since it can provide larger benefits to its participants with more liquidity. although there are some concerns about negative network effects resulting from competition especially for the supplier side [29, 34], it is assumed that the net effect for suppliers is positive. In this study, it is assumed that firms can be buyers or suppliers or both, and there is no distinction made between buyers and suppliers. all of them are regarded as participants in the channel. We model a firm’s choice of B2B channels using a simplified two-stage model. Table 2 summarizes the notation used in our model.

at the first stage, the firm decides which channel it will implement: private or public or both. at that point the firm does not know what level of cooperation exists among firms (and therefore N) in the public marketplace. at the second stage, when the uncer-tainty about cooperation in the public marketplace is resolved, the firm decides which channel to continue to use. however, if the firm had chosen to implement only one of the two channels in the first stage, it can use only the channel it had implemented at the first stage. The summary of the logic of the game is shown in Figure 1.

using a noncooperative game model, we examine which B2B channel, public or private, will be chosen by the firm. Let pb denote the “public” channel and pr denote the “private” channel. We assume that products can be ordered based on their speci-ficity (θ), which is uniformly distributed between 0 (commodity) and 1 (customized product):

θ ~ U(0, 1). (1)


Cost Factors

There are many costs incurred by buyers and suppliers in B2B procurement processes. prior studies have proposed different ways of categorizing the cost factors related to B2B transactions [9, 21]. We categorize several of these cost factors below.

Buyers face six types of costs associated with searching for products and business partners, collecting information, decision making, bargaining, writing contracts, and

Table 2. Summary of Notations

Notation Definition Comment

pb, pr Subscripts to differentiate public and private channels.

Identify the private and public channels.

bpb, bpr Benefit from the public and private channel normalized by firm size.

Gains from using the B2B channel.

λ General IT level of B2B channel. General level of IT such as the capability of IT personnel, hardware, and software.

e Level of IT customization and security provided in the private channel.

High customization and more secure environment are main benefits of the private channel.

θ Level of specificity of a product. 0 is the lowest (commodity), 1 is the highest.

s Scaling parameter, effects benefit from the public channel.

Reduces benefit from the public channel for products with high asset specificity (θ).

n Size of the firm. Used as a scaling parameter.

N| Ex ante expected size of the public marketplace at the second stage.

N is assumed to be uniformly distributed between N| – e and N| + e.

e Ex ante uncertainty about the size of the public marketplace at the second stage.

More uncertainty means that the expectation about the size of the public marketplace at the second stage is more uncertain.

fpb1, fpr1 Fixed costs at the first stage to set up a private or public channel.

These are costs used to build the IT infrastructure and set up the channel.

fpb2, fpr2 Fixed costs at the second stage to use and maintain a private or public channel.

These are costs to use the channel in the second stage.

k The size of the public marketplace at which the benefits of the public and private channel are equal.

When the costs of the private and public channels are equal, k = (λ + e)/λ.

Prob(pr) Probability of selecting a private channel over a public channel.

The expected probability that N is smaller than N *.

g The extent of risk aversion of the firm.

The utility of the firm is less when e is greater.


enforcing contract provisions. There are five types of suppliers’ costs associated with transactions, managing workflow, inventory, production, and shipping. These costs may further be categorized as fixed costs or variable costs depending on whether they change with the number of transactions. For example, search costs are incurred for each transaction and are considered to be variable costs. however, implementation and maintenance costs for the channels are fixed costs that are not affected by the number of transactions.

alternative 1: private Channel Only

This alternative is the case in which a firm uses its private channel only. For example, Volkswagen continued to use its private channel only while its rivals such as gM and Ford built and joined Covisint. It is assumed that the savings from the private channel, normalized for firm size, can be expressed as the sum of the benefits from customization and general IT, as in Equation (2). Compared to a public marketplace, the negative effect of product specificity on savings is marginal (little or no disutility) in a private channel and assumed as constant, 1:

bpr

(λ, e, θ) = (λ + e) · 1. (2)

Figure 1. Simplified Flow Chart of Choice of Strategies


Thus, the aggregate gain in profit (Π) from the private marketplaces can be calculated as in Equation (3), where C

pr(n, λ) denotes the total of fixed costs for implementing the

private channel and variable costs of using the channel, normalized for firm size (n). We assume that the private channel yields savings that increase with the level of IT capability of the firm (λ), as well as with the benefit of IT customization and informa-tion security (e). We also assume that the benefit is linearly related to firm size (n). Because we defined benefits (and variable costs) as normalized for firm size, the ag-gregate benefit from using the private channel is scaled by firm size. The profit from the private channel can be defined as in Equation (4) following our assumptions. as assumed in the first section of this paper, costs are categorized as fixed costs in the first stage for setting up and implementing the private marketplace and in the second stage for maintaining the marketplace (f

pr1, f

pr2), and as variable costs (v

pr · n).

Πpr pr prn e n b e d C, , , ,λ λ θ θ( ) = ( ) −∫0

1

(3)

= +( ) − + +( )n e f f v npr pr prλ 1 2 .

(4)

alternative 2: public Marketplace Only

This alternative is the case when a firm uses only the public marketplace without building its own private marketplace. Many small and medium-sized firms prefer to use a public marketplace and they do not want to invest in developing private channels. For the public marketplace, procurement savings can be calculated where the benefit decreases when the product specificity (θ) increases and the benefit increases when the size of the public market increases. We include IT customization and information security parameter, e, of the private channel in the benefit level of the public channel only to make the comparison of benefit levels of the private and public channels a fair one. Equation (5) was designed so that the benefit from the public marketplace is equal to the benefit from the private marketplace when the size of the public marketplace is at the level k = (λ + e)/λ, which makes the profits from using those two marketplaces equal when the costs of using the two markets are equal (b

pr = b

pb if N = k). Note that

for a given value of the parameter s, the level of benefit of the public channel reduces as the product specificity (θ) increases. as in the subsection “alternative 1: private Channel Only,” the benefit is normalized for firm size:

b N k s N kpb , , ,λ θ λ θ λ( ) = −

+

+ −( )1

21

(5)

where k = (λ + e)/λ.There is an inherent uncertainty about the level of cooperation in the public mar-

ketplace. Firms get savings when they use public marketplaces. Based on external conditions (the level of uncertainty), firms decide whether to invest in the public marketplace. For simplicity, we assume that the size of the public marketplace result-ing from the cooperation of firms is uniformly distributed around an average N|, as


in Equation (6), with the range of 2e. This scope parameter, e, represents the degree of uncertainty at the first stage about the size of the public marketplace at the second stage. It is assumed that the nonparticipation of a firm of size n reduces the level of cooperation:

N U N n N n

N U N N

~ ,

~ ,

− − + −( ) ( )− +( )

ε ε

ε ε

if the firm does not joinn

n if the firm joins( ) .

(6)

The aggregate gain in profit (Π) from public marketplaces can be calculated as shown in Equation (7). The total profit is the size of the firm (n), multiplied by the integral of profit for products of different specificity. C

pb denotes the sum of fixed costs and

variable costs of the public marketplace normalized for firm size. We expect the sav-ings to increase with the firm’s level of IT expertise and the level of participation of others. When more participants join the public marketplace, there is greater likelihood of finding a better match and therefore the value of the public marketplace increases as well. as assumed in the subsection “Cost Factors,” costs are categorized as fixed costs in the first and second stages (f

pb1, f

pb2), and as the variable costs (v

pb · n) of us-

ing the marketplace.

E N n n b N d C

nN f f v n

pb pb pb

pb pb pb

Π , , , ,

.

λ λ θ θ

λ

( )( ) = ( ) −

= − + +( )∫0

1

1 2

(7)

Thus, we can calculate the cooperation level, N *, where the profits from using the private and public channels are equal. When the profits of private and public channels are compared, there is a trade-off between more benefits from customization and se-curity (e) and the network effect from greater participation of business partners (N) in the public channel. If the expected level of participation in the public channel exceeds N *, then the firm will choose the public channel over the private channel.

N

e n C C

n

pr pbpr pb

* , .=+( ) − −( )

=λ

λΠ Π

(8)

alternative 3: Both Channels

We also consider the case in which a firm uses both channels at the same time. For example, gM and DaimlerChrysler used their own private channels called “gM Supplypower” and “Fastcar,” respectively, while simultaneously using Covisint as a public marketplace [15].

For simplicity, we start with the case in which a firm implements both channels at the first stage and then chooses to use a single channel in the second stage (the one that is more profitable after uncertainty about the level of participation on the public channel is resolved). This is shown in Equation (9). We call this strategy the “pick” strategy:


E E C C C

C Prob pr

pick pr pr pb pb pick

pr pr

Π Π Π

Π

( ) = + +{ }( ) −

= +{ }⋅ ( ) +

max ,

ΠΠpb pb pickC Prob pr C+{ } − ( )( ) −1 ,

(9)

where the probability of selecting the private channel is

Prob prN N( ) =

− −( )*

.ε

ε2

at the first stage, the firm invests in both channels if

E Epick pr pbΠ Π Π( ) > ( ){ }max , .

(10)

Thus, we can calculate the option value of implementing both channels and having the flexibility to choose the more favorable channel later, as in Equation (11):

v option E E E Epick pr pick pb( ) = ( ) − ( ) ( ) − ( ){ }

max min , , .Π Π Π Π 0

(11)

Equation (11) means that, by having both channels in the first stage, the firm can pick the channel that produces greater benefits at the second stage. This is like having an option that the firm exercises when it is profitable to do so.

The price of the option is the amount of added costs incurred to implement both channels at the same time rather than implementing either of the two channels exclu-sively, as in Equation (12):

p option C C f

C C f

pick pick pb pr pb pr

pick pr pb

Π Π Π( ) = − = >

= − =1

1

if or

pr pbif Π Π> .

(12)

at the second stage, the firm chooses to use only the channel that produces higher profits:

Π Π

Π Π

pick pb

pick pr

N N

N N

= >

= <

if

if

*

*.

(13)

as can be seen in Figure 2, when a firm chooses to create both channels at the first stage, it will choose the public marketplace at the second stage only if it produces more profit than the private channel. If it implements both channels at the first stage and thereby invests more, the profit in the private channel will be the lowest profit the firm could earn at the second stage. The dashed line in Figure 2 is the adjusted probability distribution of the profit from the public marketplace truncated at the profit level of the private channel. The managerial flexibility of having two channels in place gives the firm the option to choose the more favorable channel in the second stage.

Decision

The decision of the firm is among three alternatives: implementing only the public channel, implementing only the private channel, or implementing both and using


one exclusively later. The firm will choose the alternative that it expects to yield the maximum profit. The problem can be summarized as follows:

max , , , , , , , , , .Π Π Πpb pr pickN n n e E N n eλ λ λ( ) ( ) ( )( ){ }

(14)

results

we now presenT The resulTs of our analysis. We start with comparative statistics from a basic model that assumes that the level of IT spending is determined exogenously. Then we examine the case in which the firm implements both channels. The models presented in previous sections have assumed that the level of IT spending is determined exogenously. however, the level of IT spending at a firm is obviously an endogenous variable. Could the firm’s strategic choice of B2B sourcing channel affect its level of spending on IT? We examine this interesting question in the subsection “Endogenous IT.” In the subsection “risk-averse utility Function,” we incorporate a risk-averse util-ity function. While it is often assumed that large firms are risk neutral, it is important to consider the impact of risk aversion because supply chain disruptions can impose very high costs on the firm and its employees. For example, the shortage of a single essential component can idle a manufacturing plant.

Comparative analysis

proposition 1 shows the impact of the level of IT on channel choice:

Proposition 1 (Impact of the Level of IT on the Choice of B2B Channel): The impact of an increase in the level of IT (λ) on the choice of private or public channels depends on the implementation costs and customization technology levels. When

Figure 2. The Option Value in giving Managerial Flexibility


the fixed cost differential between public and private channels is small and there is a sufficiently high degree of customization available in the private channel, higher levels of IT will favor public channels.

an increase in the level of IT improves the efficiency of the private and public channels.

∂∂

>∂

∂>

Π Πpb pr

λ λ0 0,

(15)

∂∂

>N

e

*

0

(16)

∂∂

≤ ⋅ − −( ) >

∂∂

<

N e n C C

N

pr pb

*

*

.

λ

λ

0 0

0

if

otherwise

(17)

as shown in Equation (15), more advanced IT within a firm benefits both channels. The impact of advanced IT on customization (increases e) will favor the use of private channels as in Equation (16). however, the effect of general IT on B2B channels, such as better WaN (wide area network) technology and greater bandwidth (increases λ), is not deterministic. as shown in Equation (17), the impact depends on the cost structure and level of customization, as in the findings of gurbaxani and Whang [13] on the impact of IT on organizations. The impact depends on the trade-offs between internal and external economic conditions faced by the firm.

Proposition 2 (Value of Creating Both Versus Differences in Profits of Two Chan-nels): When the difference in the expected profit resulting from the two channels is greater, the value of creating both is lower.

The value of choosing both marketplaces at the first stage is lower when the differ-ence in expected profits between the two is greater. In other words, when the expected cooperation level (N |) is further from N * in either direction, the choice of one market-place will be preferred and will occur with a higher probability. as seen in Figure 3, we consider two factors in decision making: the expected cooperation level and the magnitude of implementation costs for the two marketplaces. The firm will invest in both when the implementation costs (f

pb1, f

pr1) are small and the expected cooperation

level is close to N *, and where the expected profits from the two channels are similar. Intuitively, when one marketplace is much more advantageous, the value of the option to choose between two channels decreases. N

h and N

l are the limits of the cooperation

levels. The firm need not implement both channels if the cooperation is higher or lower than these limits, even when there are no fixed costs (f

pr1, f

pb1) at the first stage:

Proposition 3 (Value of Creating Both Versus the Level of Uncertainty): When the level of uncertainty (e) regarding a B2B channel is higher, the benefit to the firm from implementing both channels is greater. Thus, the probability of implementing both channels at the first stage is higher.


as we saw in the subsection “alternative 3: Both Channels,” when there is greater uncertainty (e) at the first stage, there is a greater probability that firms will choose to implement both channels as in Equation (18). We assume that there is an interior solution of the network level (N *) that makes the profit from the public and private channels the same.

∂ ( )∂

=∂ ( )

∂>

E v optionpickΠ

ε ε0

(18)

if and f f f f v v N Npr pb pr pb pr pb1 1 2 2= = = − <, , , .ε

(19)

When the size of the marketplace is uncertain, the option value, as in Equation (11), increases as the uncertainty increases. It is difficult to measure the magnitude of uncertainty in B2B e-commerce. however, since we know that uncertainty in public marketplaces is very high because the environment of B2B e-commerce is unstable, firms that are able to increase their future flexibility by implementing both channels will do so based on their assessment of the uncertainty level. This explains the behavior of large firms that invest in both private and public channels at the same time [36]. These firms may intend to use both channels for certain activities, or they may be hedging their bets. They are likely to stop using the public marketplaces if conditions favor the use of private channels.

Simultaneous use of Both Marketplaces

The “pick” strategy above eliminates the possibility of the firm using both channels at the same time. The strategy of using both channels at the same time is called the “both” strategy.

as shown in Figure 4, it is assumed that the unit profit from the public marketplace decreases with an increase in product specificity. Meanwhile, the unit profit of the private channel stays the same. Thus, the firm uses the public marketplace only for

Figure 3. Strategy Chosen, Depending on the Magnitude of Implementation Costs (fpb1

, fpr1

) at the First Stage (x-axis) and the Expected Size of the public Marketplace (N|) (y-axis)


products whose procurement is more profitable via the public marketplace. For ex-ample, firms may use the public marketplace for less specific products (0 < θ < θ*) while it may use the private channel for highly specific products (θ* < θ < 1). The costs of using both will then be a combination of the two, as in Equation (20). The optimal use of the public marketplace (θ*) is decided at the second stage.

E b n d b N n d Cboth pr pb bothΠ( ) = ( ) + ( ) −∫∫ , , , , , ,

*

* λ θ θ λ θ θθθ 0

1

(20)

where bpr

(n, λ, θ*) = bpb

(N, n, λ, θ*).From Figure 5, by comparing the unit profits from the public and private channels,

we can see that the net gain of using the “both” strategy is the smaller of the two shaded areas (efficiency gain from 0 to θ* or θ* to 1) minus the price of the option to implement both channels. Thus, the value of the option, as in Equation (11), can be rewritten as in Equation (21):

v options s

( ) =⋅ ⋅ −( )

min , .

* *θ θ

2

1

2

(21)

Compared to the “pick” strategy, the “both” strategy incurs more costs by maintain-ing and using two channels at the same time. So, the price of the “both” option is as follows:

p option C C p option C C

f

both both pr both both pb

pb

Π Π( ) = − ( ) = −

=

or

2 iif or if f pr pb pr pr pbΠ Π Π Π> = <2 .

(22)

The decision depends on the level of cooperation (N). as in Figure 4, when there are greater benefits resulting from greater cooperation, more products will be procured via the public marketplace. When the “both” strategy of using two channels together in the second stage is also considered, the area of implementing both at the first stage is

Figure 4. Changes in unit profit from the public Marketplace, Depending on the Cooperation Level


larger, as in Figure 6, than the area of implementing both at the first stage when only the “‘pick” strategy of choosing only one channel in the second stage is considered. The probability of implementing both at the first stage (the shaded region in Figure 6) becomes larger (N ′

h > N

h and N ′

l < N

l). The analysis can be categorized into two cases,

depending on the size of the firm.

Exogenous N Case (Small Firm)

If the firm invests in both marketplaces at the first stage, the outer boundaries of the decision area is determined by either the “both” strategy or the “pick” strategy (see Figure 6). The decision depends on the benefit of the “both” strategy (s) and the costs of maintaining both (f

pb2, f

pr2) channels at the second stage. So, the range of Π

both can

be either larger or narrower than that of Πpick

.at the second stage, we denote the ex post profit by adding to the subscripts “| N.”

The firm chooses to use only one channel (Πpick

) if

Πboth | N

< Πpick | N

, (23)

where Πpick | N

= Πpr | N

– fpb1

if Πpr | N

> Πpb | N

, or Πpick | N

= Πpb | N

– fpr1

if Πpr | N

< Πpb | N

.In this subsection, we assume the firm is small, and therefore we disregard the

impact of the firm’s decision to use both channels or only one channel on the level of participation in the public channel.

Insofar as there is an area for the “both” strategy, that area is flanked by the “pick” strategy. In addition, the area of the “both” strategy can be so large that it usurps and therefore replaces the entire area of the “pick” strategy, as in Figure 7.

The size of the “both” strategy area in Figure 7 depends on the magnitude of the fixed costs at the second stage (f

pb2, f

pr2), while the size of the area of implementing

both channels at the first stage depends on the magnitude of the fixed costs at the first stage (f

pb1, f

pr1). The decision whether to implement both channels or only one

Figure 5. Comparison of unit profits procured by the public or the private Marketplace


channel at the first stage depends on the expected value from implementing both and implementing only one, as in Equation (24):

Π Π Π Π Π Πpr both pick pb both pickE E E> { }( ) ( ) > { }( )max , max , ,or

(24)

where

E dN

both pick bothUp N pick NN

N

both

max , max ,

max

*Π Π Π Π

Π

{ }( ) = { }+

+∫

ε

DDn N pick NN

NdN, .

*

Π{ }−∫ ε

More detailed mathematical explanations about this section are provided in ap-pendix a.

Endogenous N Case (Big Firm)

Compared to Exogenous N Case (Small Firm), when a firm is large, its decision to participate in the public marketplace has a meaningful effect on the size of the public

Figure 6. Expansion of the region of Investing in Both Channels at the First Stage When the “Both” Strategy Is Chosen

Figure 7. Ex post area of Optimal Choice


marketplace. Therefore, in this section, we examine the case when a firm is large enough to affect the size of the channel itself, and in turn, affect the level of coopera-tion and the benefit of the public marketplace from network effects. Thus, we define the cooperation level affected by the firm itself as N

e*:

N N ne e

* * * .= − −( )1 θ

(25)

This cooperation level (Ne*) means that when the “both” strategy is chosen, the use

of the public marketplace is reduced. If we compare the profit function of the “both” strategy to the single choices (public or private only), it can be arranged as follows:

Π Π

Π

both e pb e e pr e

pr

n n ns

f N N

n

= − −( )⋅ ⋅ + −( )⋅ − ≥

= −

1 12

1

2θ θ λ θ* * * *if

−−( )⋅ + ⋅ − <θ λ θe e pb en ns

f N N* * *.2 2 if

(26)

Proposition 4 (Impact of Firm Size on the Choice of Channels): The larger the size of the firm, the less it will use the public marketplace.

The decreasing network effects on the public market by a firm’s self-imposed de-parture reduce the overall power of the public marketplace, as well as the probability that both channels will be implemented. Furthermore, when the size of the firm (n) is greater, the use of the public marketplace (θ

e*) becomes less. The incentive for larger

firms to use public marketplaces is reduced partly because of their partial commitment to the public marketplaces. however, it is not deterministic whether the endogenous network effect increases the proportion of firms using the “both” or the “pick” strat-egy. The decision depends on the ex post network level (N). For example, when the ex post network is greater than N *, the reduction in network level results in a higher probability that firms will choose the “both” strategy.

θλ λ λ

λ

θλ λ

e

e

e

N N N n ns n e

nN s

N N N n ns

*

* *

*

*

* *

=− + +

− +( )

→ =− + +

12

12

− +( )

−

{ } −<

n e n

nN s n

λ

λθ

**.

(27)

Endogenous IT

until now, we have treated the IT factor, λ, as an exogenous variable. however, from the view of a firm, IT is an important strategic lever that it can control. We therefore redefine the IT factor with a convex cost functions, as in Equation (28):

CIT

= aλ2. (28)


It is assumed that the more IT capability the firm has, the more the implementation cost of the channel will decrease, as in Equation (29), because more advanced IT makes it easier for the firm to implement B2B channels. Meanwhile, the benefit from the channels increases with more IT capability.

f a b

f a b

pbi pbi pbi

pri pri pri

= − ⋅

= − ⋅

λ

λ,

(29)

where i = 1 or 2.The adjusted profit functions when the IT factor is endogenized are as follows:

Π Π

Π Π

Π Π

pb IT pb IT

pr IT pr IT

pick IT pick IT

C

C

C

= −

= −

= − .

(30)

Now we can derive some interesting results about the optimal level of IT.

Proposition 5 (Endogenously Decided General IT Level Depending on the Choice of Strategies): The optimal level of general IT (λ) is greater when the firm chooses to implement both marketplaces relative to the IT capability when the firm chooses to implement either the public or private channel only.

When we simplify the costs of implementing the channels (fpr1

= apr1

– bpr

λ, fpb1

= apb1

– bpb

λ, fpr2

= fpb2

= a2, where a

pr1 > a

pb1, b

pr > b

pb), the optimal level of general

IT is greater when the firm implements both than when the firm implements only one channel (λ*

pick > λ*

pr, λ*

pb if the “pick” strategy is chosen and uncertainty e is big

enough). The detailed proof is in appendix B.Thus, the general IT levels implemented by firms will be higher when they choose

to implement both channels as compared to the cases where firms choose to imple-ment only one channel. Because companies that implement both channels at the start have bigger gains from the investment on IT, they invest more than other companies that use only one channel. Meanwhile, when general IT level is an exogenous factor, it can be said that companies with higher general IT levels have more probability to choose to implement both channels.

risk-averse utility Function

until now, the utility function of the firm has been assumed to be risk neutral. Now we consider the effect of firm decision making with a risk-averse utility function. When the utility function of the firm is a risk-averse function, as in Equation (31), the result is as follows:

E Nn Nn Cpb pbΠ γ λ γε λ( ) = − ⋅ − .

(31)


Proposition 6 (Impact of Risk Aversion on the Choice of B2B Channels): Risk aversion tends to offset the benefit of implementing the relatively higher-risk public marketplace and also the benefit from implementing both channels.

We define Ng* as the cooperation level at which the expected benefit of the public

marketplace with the risk-averse utility function is equal to the benefit of the private channel:

N

e n C C

nN

pr pb

pr pbγ γ γ

λ

λ γε* *, .=

+( ) − −( )−( ) > =

1Π Π

(32)

risk aversion offsets the value of investing in a more risky channel, the public mar-ketplace, and IT spending, while the effect of the option value increases IT spending. When uncertainty (e) increases, a firm’s spending may increase, but the effects are offset by the firm’s risk aversion (g). So, firms’ greater skepticism about e-commerce in the public marketplace and higher risk aversion can be a cause for the public mar-ketplace to struggle and fail. This may explain the shakeout of public marketplaces in the early 2000s. When growth and profit expectations about B2B marketplaces were not met in the early 2000s and investors became more risk averse, public marketplaces failed at a very high rate.

Conclusion

The seleCTion and use of B2B Channels for industrial procurement is a key to supply chain management. Before the advent of the Internet, executives had spent decades honing various offline channels. With the outpouring of electronic channels, the B2B procurement puzzle suddenly became highly complex. Firms faced a slew of difficult questions. Should they stay away from public electronic marketplaces and develop their own private electronic channels? What if the public marketplaces grow in size and thus afford a high degree of network effects? What if they fail to grow? how much should they investment on IT? Our research attempts to develop an economic model of this complex phenomenon to address the uncertainty about the size of public marketplaces. The recent spat in the airline industry between public marketplaces and leading airline carriers reinforces the need to understand how firms should decide their involvement in public and private channels [25].

There are two key managerial implications of this paper. First, we carefully assemble a model that identifies the key forces management must consider in their attempts to solve the B2B puzzle. One force relates to the direct benefit and cost of developing a private channel. Managers can also estimate the benefits and cost of the public chan-nel, but they must also take into account the likelihood that the public channel may fail due to lack of support from other firms. Managers must also factor in their level of IT capability, the specificity of their procurement needs, and the extent to which their firms can influence the future success of the public marketplace by allocating their own purchasing to either the private or the public channel. Second, we not only


explain some behaviors that may seem irrational at first glance but we also develop specific actionable recommendations for management. how should managers react to increasing uncertainty about the success of the public marketplace? Instead of focusing only on the private channel, they should develop both channels. Such a strategy allows the firm to obtain the best of both worlds when the uncertainty vanishes, provided that the cost of developing the channels is reasonable.

Our model examines the impact of several factors such as firm size and the level of IT capability on the choice of B2B channel. proposition 1 shows that a greater level of IT capability within the firm does not always favor one type of channel over the other. however, when the cost of developing public and private channels is similar and there is a sufficiently high degree of customization available in the private channel, then a higher level of IT favors the development of the public channel. proposition 5 suggests that managers who decide to invest in both marketplaces should build strong IT capabilities to assist with these projects. proposition 4 examines the impact of firm size, showing that managers of smaller firms should focus more on public marketplaces, whereas larger firms should focus more on private channels. This is consistent with anecdotal evidence where large firms such as Walmart and hp have built tight link-ages with their suppliers through private channels. Next, we discuss how our research relates to some of the findings in the prior literature.

prior literature has examined B2B marketplaces and there is substantial literature on investment under uncertainty. The novelty in our model is that it combines firms’ choice of B2B marketplaces and the uncertainty surrounding future success of public marketplaces. We find several interesting results that contribute to the literature on B2B marketplaces as well as the literature on investment under uncertainty. proposi-tion 3 shows that higher levels of uncertainty can increase the likelihood of the firm investing in both public and private channels. This finding is counterintuitive because one would expect higher uncertainty to lower investment activity. For example, kauff-man and Li [20] and pindyck [27] suggest that increasing uncertainty reduces the level of investment, whereas abel [1] and hartman [14] have shown that increasing uncertainty can lead to greater investment. This paper shows that in the context of B2B marketplaces, increasing uncertainty about the success of B2B marketplaces is expected to lead to higher investment. We provided several examples where firms have chosen to invest in both private and public channels. The most prominent example is that of DaimlerChrysler and gM. They were founding partners in Covisint and both chose to simultaneously invest in public and private channels [15]. Our model offers one possible explanation for the high failure rate of public B2B marketplaces as large firms that initially invested in both channels subsequently shift to using only private channels.

We examined a two-stage model with uncertainty about the success of the public marketplace. Further research is needed to examine the impact of reversible invest-ments, the option value of small pilot projects, and the potential for investment defer-ral. The empirical validation of our analytical results is another potential extension of this research.


Acknowledgement: Byungjoon yoo appreciates the research support by humanity-Social Sci-ence Major research Fund for Overseas Study of Seoul National university.

referenCes

1. abel, a.B. Optimal investment under uncertainty. American Economic Review, 73, 1 (March 1983), 228–233.

2. Benaroch, M., and kauffmann, r.j. a case for using real options pricing analysis to evaluate information technology project investments. Information Systems Research, 10, 1 (March 1999), 70–86.

3. Coase, r. The nature of the firm. Economica, 4, 16 (November 1937), 386–405.4. The container case. Economist (October 19, 2000) (available at www.economist.com/

node/398030/).5. Dai, Q., and kauffman, r.j. To be or not to B2B: Evaluating managerial choices for

e-procurement channel adoption. Information Technology and Management, 7, 2 (june 2006), 109–130.

6. Dai, Q.; kauffman, r.j.; and March, S.T. Valuing information technology infrastructures: a growth options approach. Information Technology and Management, 8, 1 (March 2007), 1–17.

7. Day, g.S.; Fein, a.j.; and ruppersberger, g. Shakeouts in digital markets: Lessons from B2B exchanges. California Management Review, 45, 2 (Winter 2003), 131–150.

8. Dos Santos, B.L. justifying investments in new information technologies. Journal of Management Information Systems, 7, 4 (Spring 1991), 71–90.

9. Ehrens, S., and Zapf, p. The Internet business-to-business report. Bear Stearns, New york, 1999.

10. Explore our past: 2000–present. British airways, Waterside, uk, 2011 (available at www .britishairways.com/travel/history-2000-present/public/en_gb).

11. Fichman, r.g. real options and IT platform adoption: Implications for theory and practice. Information Systems Research, 15, 2 (june 2004), 132–154.

12. granados, N.; kauffman, r.L.; Lai, h.; and Lin, h. IT-supported a la carte pricing in air travel: a test of the strategic decommoditization hypothesis. Working paper, School of Informa-tion Systems, Singapore Management university, Singapore, 2011.

13. gurbaxani, V., and Whang, S. The impact of information systems on organizations and markets. Communications of the ACM, 34, 1 (january 1991), 59–73.

14. hartman, r. The effects of price and cost uncertainty on investment. Journal of Economic Theory, 5, 2 (October 1972), 258–266.

15. helper, S., and MacDuffie, j.p. Suppliers and intermediaries. In B. kogut (eds.), The Global Internet Economy. Cambridge: MIT press, 2004.

16. herath, h.S.B., and herath, T.C. Investment in information security: a real options per-spective with Bayesian postaudit. Journal of Management Information Systems, 25, 3 (Winter 2008–9), 337–375.

17. johnson, p.F.; klassen, r.D.; Leenders, M.r.; and awaysheh, a. utilizing e-business technologies in supply chains: The impact of firm characteristics and teams. Journal of Opera-tions Management, 25, 6 (November 2007), 1255–1274.

18. kaplan, S., and Sawhney, M. B2B e-commerce hubs: Toward a taxonomy of business models. Harvard Business Review, 78, 3 (May–june 2000), 97–103.

19. kauffman, r.j., and kumar, a. Network effects and embedded options: Decision-Making under uncertainty for network technology investments. Information Technology and Manage-ment, 9, 3 (September 2008), 149–168.

20. kauffman, r.j., and Li, X. Technology competition and optimal investment timing: a real options perspective. IEEE Transaction on Engineering Management, 52, 1 (February 2005), 15–29.

21. kauffman, r.j., and Mohtadi, h. proprietary and open systems adoption: a risk-augmented transactions cost perspective. Journal of Management Information Systems, 21, 1 (Summer 2004), 137–166.


22. Laudon, k.C., and Traver, g.C. E-Commerce: Business, Technology, Society, 7th ed. upper Saddle river, Nj: prentice hall, 2011.

23. McDonald, r.L., and Siegel, D.r. Investment and the valuation of firms when there is an option to shut down. International Economic Review, 26, 2 (june 1985), 331–349.

24. Mishra, a.N., and agarwal, r. Technological frames, organizational capabilities, and IT use: an empirical investigation of electronic procurement. Information Systems Research, 21, 2 (june 2010), 249–270.

25. Mouawad, j., and Miller, C.C. The airfair maze. New York Times (February 11, 2011), B1.

26. pasztor, a. Changing course. Wall Street Journal (May 21, 2001), r22.27. pindyck, r.S. adjustment costs, uncertainty, and the behavior of the firm. American

Economic Review, 72, 3 (june 1982), 415–427.28. purchasing. university of California, Irvine, 2011 (available at http://snap.uci.edu/

viewXmlFile.jsp?resourceID=138/).29. riggins, F.j.; kriebel, C.h.; and Mukhopadhyay, T. The growth of interorganizational

systems in the presence of network externalities. Management Science, 40, 8 (august 1994), 984–998.

30. roberts, k., and Weitzman, M.L. Funding criteria for research, development, and explora-tion projects. Econometrica, 49, 5 (September 1981), 1261–1288.

31. Tallon, p.p.; kauffman, r.j.; Lucas, h.C.; Whinston, a.B.; and Zhu, k. using real options analysis for evaluating uncertain investments in information technology: Insights from the ICIS 2001 debate. Communications of the Association for Information Systems, 9, 1 (September 2002), 136–167.

32. Van alstyne, M.W. a proposal for valuing information and instrumental goods. In p. De and j.I. Degross (eds.), Proceedings of the 20th International Conference on Information Systems. atlanta: association for Information Systems, 1999, pp. 328–345.

33. Walden, E., and Subramani, M. Economic returns to firms from business-to-business elec-tronic commerce initiatives: an empirical examination. In S. ang, h. krcmar, W.j. Orlikowski, p. Weill, and j.I. Degross (eds.), Proceedings of the 21st International Conference on Informa-tion Systems. atlanta: association for Information Systems, 2000, pp. 229–241.

34. Wang, E., and Seidmann, a. Electronic data interchange: Competitive externalities and strategic implementation policies. Management Science, 41, 3 (March 1995) 401–418.

35. Williamson, O.E. Transaction Cost Economics. In r. Schmalensee and r. Willig (eds.), Handbook of Industrial Organization. amsterdam: Elsevier, 1989, pp. 135–182.

36. Xia, M., and Xia, N. The complementary effects of e-markets on existing supplier–buyer relationships in a supply chain. Journal of Management Information Systems, 25, 3 (Winter 2008–9), 9–64.

37. yoo, B.; Choudhary, V.; and Mukhopadhyay, T. Electronic B2B marketplaces with differ-ent ownership structures. Management Science, 53, 6 (june 2007), 952–961.


appendix a: Detailed Mathematical Explanation about Exogenous N Case (Small Firm) and Figure 6

for siMpliCiTy of analysis, the costs of implementation and maintenance are assumed to be equal (f

pr1 = f

pb1 = f

1, f

pr2 = f

pb2 = f

2), as in Equation (a1). The width of the region

of the “pick” strategy from N * for one side (ex ante) is as follows:

εε

λ− = = = =

2 11 1 1 2 2 2

f

n f f f f f fpr pb pr pbwhen , .

(a1)

Thus, the two limits implementing both channels at the first stage are as follows:

N Nf

nl = − −

* ε

ε

λ

2 1

(a2)

N Nf

nh = + −

* .ε

ε

λ

2 1

(a3)

at the point where the profits from the public and private channels are the same (N *), the benefit of using both channels is the greatest:

min ,* *ns

ns

12 2

−( )⋅ ⋅

θ θ

(a4)

θ

λ λλ

θ**

*

*, .=

− +( )+

∂∂

=N e

N s

N N s

1

2

1

(a5)

The profit function of the “both” strategy depends on the ex post cooperation level. The profit function is different when the cooperation level is higher than N *(Π

bothUp | N)

and when the cooperation level is lower than N *(ΠbothDn | N

):

Π Π

Π

bothUp N pick N pr

pb N pr pr

ns

f

f ns

f

= + −( )⋅ −

= −{ } + −( )⋅ −

12

12

2

1

θ

θ

*

*22

if N N v v vpb pr> = =*,

(a6)

Π Π

Π

bothDn N best N pr

pr N pr pb

ns

f

f ns

f

= + −( )⋅ −

= −{ } + ⋅ −

12

2

2

1 2

θ

θ

*

*

if NN N v v vpb pr< = =*, .

(a7)

Thus, the width of the “both” strategy area from N* for one side (Πpick | N

= Πboth | N

) can be calculated as

1

24 2 1 1 1 2 2 2n

ns f e f f f f f fpr pb pr pbλλ−( ) +( ) = = = =when , .

(a8)


There is no area for the “pick” strategy if the “both” strategy area is greater than the size of the ex ante area for implementing both channels and the width of uncertainty, as follows:

1

24

22

1

nns f e

f

nN Nh hλ

λ εε

λε−( ) +( ) >

+ + ′ −( ).

(a9)

appendix B: Detailed proofs of proposition 5

The opTiMal iT leVels of The Three sTraTegies are as follows:

∂∂

= → =+Πpr

prprb n

λλ

α0

2*

(B1)

∂∂

= → =+Πpb

pbpbb nN

λλ

α0

2*

(B2)

∂∂

= →

=

+ +− −( )

+

+ +( )

Πpickpick

pr pbb b nN N

nN N

λλ

εε

ε

0

2 2

*

* *

−− −( )

12

2

N N*

.

εε

α

(B3)

Then the meaning of the “pick” strategy chosen is that the uncertainty level (e) is large enough (bigger than e_) and is between the lower bound and upper bound of the range of N.

∂∂

>λ

εpick*

0

(B4)

if where pick private public pick privateε ε λ λ λ λ ε λ> > ( ) =, , ,* * * * ** *, .λ public

(B5)

Copyright of Journal of Management Information Systems is the property of M.E. Sharpe Inc. and its content

may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express

written permission. However, users may print, download, or email articles for individual use.

A Study of Sourcing Channels for Electronic Business Transactions. - Byungjoon

Documents

Transcript of A Study of Sourcing Channels for Electronic Business Transactions. - Byungjoon