Research " Ownership Structure and Business Model of Electronic Business-to-Business Marketplaces " pptx

Trang 1

Dissertation Ownership Structure and Business Model of Electronic Business-to-Business Marketplaces By Byungjoon Yoo Committee

Chair: Professor Tridas Mukhopadhyay Professor Rahul Telang

Professor Young Chan Choe

Reader

Professor Vidyanand Choudhary

Graduate School of Industrial Administration

Carnegie Mellon University

Trang 2

UMI Number: 3104365 ® UMI UMI Microform 3104365 Copyright 2003 by ProQuest Information and Learning Company

All rights reserved This microform edition is protected against unauthorized copying under Title 17, United States Code

ProQuest Information and Learning Company 300 North Zeeb Road

Trang 3

Carnegie Mellon

Graduate School of Industrial Administration

DISSERTATION

Submitted in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY INDUSTRIAL ADMINISTRATION

(INFORMATION SYSTEMS)

Titled

“Ownership Structure and Business Model of Electronic Business-to-Business Marketplaces” Presented by Byungjoon Yoo Accepted by S087 05

Chair: Prof /Tridas Mukhopadhyay Date

Approved by The Dean

⁄⁄ 2á.— _ 28/23

Trang 4

Acknowledgements

I thank my advisor, Tridas, who has been a mentor to support and encourage me to exploit my limit throughout my Ph.D study Thank you, VC, who has been a great supporter and also a good friend I would like to thank Professor Choe and Rahul who are my dissertation committee members I want to thank Professor Lee who was my advisor when I was at the

master program of University of Arizona He introduced me to this great research area I thank Professor Kesop Yoon and all professors in Seoul National University who taught me

‘Management’ in my undergraduate years Thank you, Sandy and []-Horn, for advice and encouragements during my Ph.D study, too Thank Professor Park, Rahul and all colleagues I have had in Tucson and Pittsburgh If it had not been for your encouragements, my life must have been a life in a desert without oases Thank Carnegie Mellon University for giving me this great research opportunity that was given to me among hundreds of applicants four years ago Thank my wife, Yeonjoo, for supporting me and enduring difficulties with me in

this long road Thank God for sending all these people to me For the rest of my life, I want

to share all the precious supports and encouragements I got from all of you with all I will meet from now on

Finally, I dedicate this honor to my parents who have supported me and have waited so long with their prayers for me to get my Ph.D degree

Trang 5

Table of Contents Abstract

1 Introduction

2 Ownership Structure Part I: A Model of Neutral B2B Intermediaries Research objectives and questions

Theoretical foundations of the study Basic marketplace model

Trang 6

Abstracts

Many experts have predicted that the main thrust of electronic commerce in the future will be business-to-business (B2B) commerce According to eMarketer, huge amount of B2B transactions are already occurring In 2004, B2B electronic commerce is expected to be 12 times larger than

business-to-consumer (B2C) electronic commerce Because the electronic B2B marketplace is at the

core of B2B transactions, it is one of the most popular and important research topics in electronic commerce Because of the high expectations for B2B, thousands of electronic B2B marketplaces have been created However, only a few of them are active and many are dying, leading to a lot of

speculation about which electronic B2B marketplaces will be successful and survive

This research deals with two interrelated aspects of the B2B marketplace First, I have been

investigating which ownership structure of the B2B marketplace will work well in terms of com-

petition and social surplus I am analyzing B2B marketplaces with different ownership structures:

neutral marketplaces and biased marketplaces Biased marketplaces are owned by the suppliers

or buyers in the industries which these marketplaces serve, while neutral marketplaces are owned by independent third parties I have found that biased marketplaces provide greater surplus to both buyers and suppliers compared to neutral marketplaces Also, neutral marketplaces are more likely to be successful in industries where both buyers and suppliers are fragmented Second, I am trying to identify which business model of B2B marketplace, private or public, will be used in

different industries What factors affect the choice of a business model (such as joining a public

marketplace or creating a private marketplace)? I have identified factors in each industry and each firm which affect this choice of business model, and I seek to examine the impact of these factors My research especially focuses on the impact of uncertainty on a firm’s choice of B2B marketplaces Following the paradigm of literature on the real option, I see uncertainty as an opportunity in the

uncertain future of electronic B2B commerce Thus, the value of IT investment in implementing

Trang 7

1 Introduction

The rapid growth in electronic commerce has led to unrealistically high expectations The bursting of the dot com bubble and the bankruptcy of many well known e-tailers (such as pets.com and egghead.com) have created skepticism about the profitability of Internet ventures Media attention has largely focused on business-to-consumer firms, but far greater volumes of transactions have been conducted in business-to-business markets According to

eMarketer [27], the level of business-to-business (B2B) revenues via electronic commerce is six times greater than that of business-to-consumer (B2C) revenues (2001) and will be far

greater in the near future In this dissertation, I examine electronic B2B marketplaces that provide the infrastructure for B2B transactions, focusing on two questions related to these B2B marketplaces: the ownership structure and the business model First, I seek to examine the impact of the ownership structure of B2B marketplaces Then I explore what business model of B2B marketplaces will prosper and will be chosen by firms

Electronic B2B marketplaces have different ownership structures I refer to marketplaces that are owned by independent third parties, such as Citadon and Chematch, as neutral marketplaces, while marketplaces that are owned by either suppliers or buyers, such as

Covisint and Quadrem, as biased marketplaces Each type has its strengths and weaknesses

Trang 8

relative to neutral marketplaces

However, there is considerable concern about the fairness of biased marketplaces as ob- served in the cases of marketplaces for airline tickets and automobile parts Buyer/supplier owned marketplaces raise the possibility that the owners of such marketplaces could devise trading rules or use proprietary market information to gain unfair advantage Covisint, the automobile parts marketplace developed by major automobile manufacturers, was investi- gated by the Federal Trade Commission concerning the possibility of a price cartel and other antitrust-related issues The owners of biased marketplaces claim that such markets will be fair and that they only seek to reduce their transaction costs and acquire the asset value of

the marketplaces resulting from their own transactions [68] Similar phenomena can also be

observed in B2C marketplaces; for example, Orbitz was recently created by the major air- lines to compete with existing neutral marketplaces such as Expedia and Travelocity Such competition between neutral and biased marketplaces is likely to occur in many industries

Trang 9

private marketplaces because they are concerned about maintaining their own competitive

advantage and securing their private information [25] This is a complex decision because

they must consider a number of factors such as the security of their private information and the degree of uncertainty involved in making and joining public marketplaces

I focus on two questions related to B2B marketplaces: ownership structures and business models First, I seek to examine the impact of the ownership structure of B2B marketplaces on their profitability, and on the level of participation of buyers and suppliers Second, I explore what business model will work in the industry areas where the B2B marketplaces exist, dependent on different conditions

There are several questions about the ownership structure of B2B marketplaces What is the role of ownership structure in determining the competitiveness of these marketplaces? Which marketplace will be better in terms of the surplus generated to each party, the in- termediaries, the suppliers and the buyers? How are the benefits likely to be shared among buyers and suppliers in biased marketplaces? What is the impact of fragmentation among buyers and suppliers? Many practitioners speculate on these questions, but there is little academic research to provide answers to them In this research proposal, I analyze and compare neutral and biased marketplaces in terms of the level of participation in electronic marketplaces, price, and surplus generated when markets exhibit network effects I also examine the social welfare and competitiveness of different, marketplaces

Trang 10

Research on B2B Marketplaces B2B Marketplace B2B Marketplace Level (Paper 3: Private vs Public) | outa Love _ | (Paper 1: ls ŸÑ 2 Figure 1: Structure of Dissertation the business model of B2B marketplaces? Also, how much should firms invest in information technology?

Even though practitioners have made many predictions, there is not much theoretical explanation for these predictions The contributions of my research are in the following areas My research explains current phenomena from a game theoretic perspective and provides a theoretical framework to help understand the current status and future status of B2B marketplaces Then by observing the past and current states of B2B marketplaces in the real world, I discuss the implication of my findings with real business cases From my analytical research and discussions on real-world cases, insights into the success and failure of B2B marketplaces in each industry will be developed

Trang 12

2 Ownership Structure Part I: A Model of Neutral B2B Intermediaries

2.1 Research Objectives and Questions

The rapid growth of electronic commerce led to many expectations, some of which were too unrealistic to hold true The struggles of famous dot-coms such as Amazon.com and Buy.com in business-to-consumer markets have cast doubt on the future of electronic com- merce While much attention was focused on business-to-consumer firms at the outset, far greater volumes of transactions were conducted in business-to-business markets According

to Forrester Research, the amount of business-to-business (B2B) sales is several times that of

business-to-consumer (B2C) sales and will be even greater in the near future However, in the context of designing B2B marketplaces on the Internet, there are several unanswered ques- tions How to attract buyers and suppliers to B2B marketplaces? How can intermediaries extract profits from the suppliers and the buyers of the marketplaces? Many practitioners have speculated about these questions, but there is little academic research to answer this

question

Trang 13

EDI The same problem exists for intermediaries who make electronic B2B marketplaces From an intermediary’s point of view, the question is, how to successfully create an electronic marketplace by acquiring enough customers: suppliers and buyers

Attracting suppliers and buyers to a marketplace is a difficult job as explained in the

previous paragraph But, in the real world of electronic B2B markets, some B2B market intermediaries have successfully created marketplaces One common factor among them is that, along with transaction services, they have provided some individual services such as

information portal services for the industry (Citadon), consulting services (FreeMarkets) and

other services After they acquire a customer base with individual services which attract customers regardless of the level of network externalities, they begin providing full-scale transaction services which are the most important source of the intermediaries’ revenue from their marketplaces

An important question facing intermediaries is how should they provide and price their services? How will network externalities affect the strategy of the intermediary? In this research, we study the strategies of intermediaries including their pricing of services when the impact of network effect is significant Usually network externalities affect a consumers’ valuation of a product depending on the number of compatible products being used by other consumers The network effect in a marketplace is different in that the value of a marketplace to a buyer depends on the number of suppliers and vice versa We discuss the nature of the network effect in marketplaces in detail in subsequent sections

Trang 14

2.2 Theoretical Foundations of the Study

There are two main benefits of electronic B2B marketplaces The first one is the speed and efficiency of transactions enabled by information technology Using advanced information technology, suppliers and buyers can reduce transaction costs The second benefit accrues

from the larger number of participants By bringing together a huge number of buyers and

sellers, electronic marketplaces increase choices Because of the ease of searching for suppliers

in electronic marketplaces, buyers have a greater chance of finding cheaper prices or better

transaction conditions Also suppliers can easily find proper buyers when they want to sell

their products to them This function is called the ‘aggregation’ function [35]

The benefits from having a larger number of participants are called ‘positive network externalities’ When the value of a product depends on the number of users, we consider the product to exhibit As an example, a telephone is only valuable if there are other people with compatible telephones that a user wishes to call In earlier research, Katz and Shapiro

[36] discussed the strategy for products with network effects They showed how network ex-

ternalities can affect the decisions of companies, especially those relating to the compatibility

of their products to industry standards Brynjolfsson and Kemerer [10] empirically showed

Trang 15

1 Supplier 2 B2B Marketplace (Intermediary) Supplier 3 Supplier 4

Figure 2: Two-sided network externalities of the marketplace

better offers On the other hand, the value of the marketplace decreases for each player when there are more competitors on her side We call this ‘negative network externalities’

and it is determined by the participation of the players’ own side Wang and Seidmann [63]

showed that the participation of more suppliers can generate positive externalities for the buyer and negative externalities for other suppliers in an electronic data interchange (EDI) In this model, we only assume negative network externalities for the supplier side and don’t expect product capacity restrictions and negative network externalities for the buyer side Bhargava et al [14] analyzed the decision of an intermediary when there are aggregation

benefits for buyers But they only considered the aggregation benefit for buyers, not the aggregation benefit for suppliers, nor the negative aspects of network externality

While there are benefits from joining electronic marketplaces, some switching costs also exist When suppliers and buyers join these marketplaces, they have to incur some expenses

to adapt their current distribution channels In previous Electronic Data Interchanges (EDI)

research, many studies examined on how to attract suppliers and buyers to EDI networks

from traditional channels Riggins et al [53] showed how to attract suppliers to an EDI

network run by a buyer in the presence of network effects Even though, in past studies, researchers mentioned that network externalities in the marketplace depends on the level of participation of other parties, such as buyers to suppliers and suppliers to buyers, little

Trang 16

research has examined cases where the level of participation of other parties affects the value of the marketplace In this research, we will see how the interaction between the levels of network externalities of two parties, buyers and suppliers, affect the optimal market share and pricing strategy of an intermediary when switching costs are also considered

2.3 Basic Marketplace Model

We analyze pricing decisions of an independent intermediary who owns an electronic marketplace in an industry We model two types of players in the marketplace: suppliers and buyers Suppliers are companies that want to sell their products to buyers in the industry which the marketplace serves Buyers are companies that want to buy resources from the marketplace We denote suppliers and buyers as ‘s’ and ‘b.’ In our model, we assume that the intermediary owns the marketplace and are independent from any supplier

or buyer Also, its decision won’t be affected by the interests of either buyers or suppliers

So, the intermediary’s only concern is to maximize its profit from the marketplace As an example in real world, Citadon, Inc is an independent intermediary which provides a marketplace for suppliers and buyers in the construction industry The marketplace provided by the intermediary offers some individual services for suppliers and buyers such as technical support and managerial consulting services, v, and v, which are not dependent on the number of suppliers or buyers Also, the marketplace provides aggregation benefits, e, and €,, which are dependent on the number of suppliers and buyers, n, and n, The negative network externalities for suppliers are represented by e, and these are proportional to the number of suppliers, n, The aggregation benefits to suppliers are proportional to the number of buyers and, likewise, the aggregation benefits to buyers are proportional to the number of suppliers as in Eq 1 where r,, 7, and r„ are constants that represent the intensity of the

network effects e,, e, and e, respectively We normalize n, and ny so that n, is the fraction

of suppliers in the marketplace relative to the total number of suppliers in the industry and

Trang 17

similarly for np

€,(Mp) — Tạ ' Tùp, €p(m;) = Tp Np, Cn (Ms) =Tn* Ts (1)

We assume that suppliers and buyers are currently using traditional distribution channels but are considering switching to the electronic marketplace In the electronic marketplace, as seen in Fig 2, the intermediary provides a marketplace where suppliers meet buyers and vice versa Suppliers and buyers who are in the traditional marketplace are heterogeneous

in terms of the switching costs of transferring their transactions from the traditional to the

electronic marketplace We assume that the supplier and buyer types (z,, x») are uniformly distributed in (0, 1) where 0 represents no switching costs and 1 represents high switching costs For example, for buyers and suppliers who have already adopted information tech- nology, it would be easier to switch to the electronic marketplace while other buyers and suppliers with a lesser degree of information technology adoption may encounter greater difficulty of switching The difficulty of switching also depends on the general nature of the industry served by the marketplace For example, if an industry has not developed stan- dardized ways of product specification, it would be more difficult for buyers and suppliers

in that industry to transfer to electronic marketplaces We model the switching costs as

S,*X, and s,- 2, where s,, s, represent the general industry-level difficulty of switching and Xs, Zp represent the heterogeneity among individual suppliers and buyers in terms of their ability to adapt to the electronic marketplace These switching costs include both the costs of purchasing and installing equipment to connect the business to the electronic market- place and the costs involved in adapting the traditional business’ processes for the electronic marketplace

Buyers and suppliers pay fees, p, and pp», to the intermediary for using its marketplaces The profit functions of suppliers and buyers are as follows in Eq 2 and Eq 3

Us = Us tTsMp — TrNs — SsLg — Ds (2)

Up = UtTeNs — $52o — Do (3)

Trang 18

Electronic Market Traditional Market Suppliers [ E-Market Trad-Market | 0 / 1 Easy to X, Difficult to transfer Transfer Buyers | E-Market Trad-Market | , 0 x, 1

Figure 3: Market share of electronic markets in both markets

Suppliers and buyers will join the marketplace if and only if the profit from joining the

marketplace is positive as in Eq 4 Let 2, and x, be the buyer and supplier that is

indifferent between switching and not switching to the electronic marketplace Thus we

obtain the indifference equations stated in Eq 5 and Eq 6 where 2’, and 2’, are the supplier

and buyer indifferent between joining and not joining the marketplace

Us > Ú,uạ >0 (4)

Us = Uy TTsTiẹ — Tn†ìa — S5X, — Pps = 0 (5)

Up = U+TeNs — $0, — Pp = 0 (6)

As shown in Fig 3, all buyers (suppliers) of type 2 (,) less than 2}, (x4) will switch to the electronic marketplace Therefore, the market share levels (n, n,) of the electronic

marketplace are given by:

Ns =X, Ny = ÿy (7)

The intermediary intends to maximize its profit by charging fees to suppliers and buyers for using its marketplace Then the intermediary’s decision problem P1:

P1: Find p,,p, in order to maximize total profit function of the intermediary (Eq 8) subject to the constraints on the market shares (Eq 9 and Eq 10)

max 7 = ÿÐ; 'Tt¿ Ps;Pb + Po: (8)

Trang 19

s.t 0 lA Ss A — (9)

0 < m<1 (10)

What will be the optimal price levels and market share levels? The propositions for those

are in the following subsections

2.3.1 Interior Solution

This is the case when the market share levels (n, and n,) that maximize the profit of the intermediary, Eq 8, is not bounded by the constraints of market share level in Eq 9 and Eq 10

0 < n<1 (11) 0< n<l (12)

For the interior solution to hold, the relationships among parameters as in Eq 13, Eq 14 and Eq 15 should hold to satisfy conditions Eq 11 and Eq 12

4s,(8,+Tn)—(rs +7)? > 0 (13)

49,(s, +r„) — (rạ +)?” > (Te+17s)¥o + 2500s (14) 49g,(8; + ra) — (rạ +Ts))” > (re +r;}0; +2(rạ + 5a)9y (15) Proposition 1 (Interior Solution) The optimal price leuels (p}, p) and market share leuels

(nz,nj) which maximize the profit of the intermediary are

* — (rz—Ty)(ss+ra)Uy+(Js,(sa+rn)—ru(ra+ru))0 nx _ (n—Ts)sp0s+(3su(ra-Ess)—rs(rs+re))uy

Ps 4sp(ss-+rn)—(ra-+rp)2 » DP, 4s,(ss+rm)—(ra-try)2 Ệ

n* = —(totrelot2spvs _ an n= (rotrs)Ust2(rnt+Ss)up

8 —— 4sp(sa+rn)—(ra+rp)2 4s,(sa+rn)—(ra+rp)2 `

Proof From solving the profit functions of buyer and supplier at the indifferent points as in Eq 16 and Eq 17, we can get the price levels (p,, pp) in terms of the market share levels

(ns, np) as in Eq 18 and Eq 19

Us = 0s +T¿f¿ — Tan — 8s7„ — pạ = Ö (16)

Trang 20

U = WtToNs — SoNpy — Dạ = (17)

> Ps = Us tTsNb — TaNs — SsN5 (18)

—>_ Đụ = Uụ tT THNs — SoN4 (19)

Then these are substituted into the profit function of the intermediary as in Eq 8 We get

the first order conditions of the profit function (7) for each market share level (n,, ns)

an, dx = 0 (20) 20

dr

dn TỦ

Using Eq 11 and Eq 12, if the market share constraints are not binding, the optimal market share levels are as follows in Eq 21

ne = rs(Up — Po) + $6(Us — Ds) — Tn(Us — Po) (21) S;Öp — TsTp + TpTa Ty(Ð; — Ds) + 85(Up — Po) Ss3p — TsTp + TwTu nM = Then the optimal price to maximize the profit of the intermediary is as follows (r¿ — Ts)(8; + Ta)0y + (28(8; + Tạ) — Te(r; + rp))0; Aso(ss +n) — (rs +70)? + (Tạ —T)8g0; + (25;(rn + 5;) — TsẲr; + Ty))0o Py = 45,(8,+Tn) — (rs +10)? Ps (22)

Trang 21

Figure 4: Market Share Levels with different levels of the difficulty of switching: The changes of optimal market shares with different level of the difficulty of switching for suppliers (s,)

from 17 to 1.5 (s» = 1, vs = U = 2, Ty = 4,75 = 2, rn = 05) Here, both the optimal market shares are reduced when there is higher difficulty in switching

From the optimal solutions of price levels and market shares of suppliers and buyers’

types (p*, py, n*,nz) in Eq 22 and Eq 23, we can see that both optimal market shares (n*,nz) are influenced by the parameters that define the two types ( 1s, Tn, Us; Ss) Tb» Ub; Sb)-

As an example, if there is a reduction in switching costs for suppliers (s,), thus causing an

increase in their market share (n,), the optimal market share of buyers also increases because of the indirect effect from the increased number of suppliers through network e,(n,) This

can be seen in Fig 4 In the construction industry, if Internet technologies become wide- spread among suppliers, it will be relatively easier for suppliers to participate in electronic marketplaces and more suppliers will join the electronic marketplaces This in turn will attract more buyers due to the larger aggregation benefits resulting from an increase in the number of suppliers Even though there is no change in the industry parameters for buyers, the market share of buyers in the electronic marketplaces is increased In this model, we want to consider and analyze these mutual effects between suppliers and buyers which have not been accounted for in prior literature

All the effects of the changes of the parameters on the optimal market share levels (n*, n})

are straightforward as expected The market share of one type is always affected by changes

Trang 22

in the conditions of the other type through indirect effects For example, the optimal market

share level of supplier type increases with increase in the strength of the network externalities for either type (= > > 0, ou > 0) The negative network externalities have a negative impact

only on both the market share level of suppliers and buyers (32 ec 0, a b< 0) When more individual services are provided for either type, the market share level of supplier type is higher (Ss i> 0, 24 Ou > 0) Also, as expected, the optimal market share levels are lower when there is greater difficulty in switching for either type at the industry level (2 2< 0,2 Os, *< 0)

2.3.2 Boundary Solution

Here is an interesting result about the boundary solutions which occurs when the market share levels (n,,n») are either zero or one In other words, these results apply to the cases when market share level constraints in Eq 9 and Eq 10 are binding There are eight

possible cases where either constraint is tight (1 (n, =0, nm, = 0), 2 (n, =0, n, = 1), 3 (n; =0, 0 < n <1), 4 (n, =1,0< m <1), 5 (n, = 1, m = 0), 6 (n, = 1, m = 1), 7 (0 < n; < 1, nạ = 0), 8 (0 < n¿ < 1, nạ = 1))

Proposition 2 Whenever the value of individual services (v,,v») is positive, the market

share levels (n,, ny) are always positive

This is a very intuitive result Whenever it is possible for the intermediary to provide benefits at no marginal cost, it is optimal for the intermediary to set prices so that some buyers and suppliers can use its services So five of the eight cases are not feasible ( 1

(n; =0, nụ = 0), 2 (n; =0, m = 1), 3 (ny =0, 0 < m < 1), 4 (n, = 1, m = 0), 5

(0 < n¿ < 1, nạ = 0)) The detailed proof is in Appendix A

There are three remaining cases with boundary solutions where either n, or n, is equal to one

Trang 23

Case 1: Both markets are fully covered (n} = 1, n¿ = 1)

This is the case where both constraints, Eq 9 and Eq 10, are binding (n, = 1,n, = 1) Then from the indifference points in Eq 16 and Eq 17, we get the optimal price levels

Py = Us tls — Tn — Ss

Pp = UWtTs— So

Case 2: Buyers’ market is fully covered (nj = 1, 0 < nÿ < 1)

In this case, the constraint for the market share of suppliers Eq 9 is binding Then we

get price levels in terms of market share levels (n,, ny)

Us tT.° 1 — Tats — Sstbs

Ps

Po = Vat TeNs — 8p

With these prices (ps, Pb) plugged into the profit function of the intermediary, we can get

optimal market share of suppliers (n*) from the first order condition ({* = 0) Then the

optimal price levels and the market share level of suppliers are as follows 2 * Ty + Đa — Tp — Tạ Ts — Tn + Tos — 2885 + 285Up + THVs Pe = ng HC 2s nt _ rats

Case 3: Suppliers’ market is fully covered (n* = 1, 0 < nj <1)

Following the steps in Case 2 (nj = 1, 0 < nz < 1), the optimal price levels and the market share level of buyers is as follows y= r.(To + 1s + Uy) — 25y(Tr + 35 — Us) p= Tp — Tạ + Up $ 255 nb 2 * * Tp + T;s + Đọ Ne = 1, Tp = Is —

We can see the optimal market share levels and price levels of buyers and suppliers in Fig 5 and Fig 6 In both figures, the optimal price and market share levels are kinked at two

Trang 24

Market Share

Figure 5: Market Share Levels at the boundary case: The changes of optimal market share

level with different level of the strength of network externalities on buyers (r,) from 0 to 2.5

(ss = 3 = 1, vs = U = 2,75 = 2, Tn = 15) Here, the market of buyers is fully covered

at first when there is more strength of network externalities on buyers Then the market of suppliers is fully covered

Price

1.5

-1

Figure 6: Price Levels at the boundary case: The changes of optimal price level with different

levels of the strength of network externalities on buyers (r;) from 0 to 2.5 (s, = s = 1,

Us = Up = 2, 5 = 2, Ty = 15) Here, the slopes of optimal price levels are kinked at the point where either market is fully covered

Trang 25

rp levels The first kink is the point where the market of buyers is fully covered, and Case 2 (ng = 1, 0 < n* < 1) applies The second kink is the point where the market of suppliers is

fully covered, and Case 1 (n* = 1, nj = 1) applies

We summarize the boundary solutions of the market share and price levels as follows Solutions\Cases | n, = 1,n, = 1 n = 1,0 <n <1 [n,=1,0< n <1 rs(rotrstvy) —tr+— 2s ?} Us + Ts — 8; — Tạ Ta+Us a Tr b — 2su(rn-+Sa =s) 285 re —TETn tos —258d8s 2s To—Tstv, Pe Tp + Up — 8 ° =f -_28s9b-+LrpUs 25a Tụ+rs+us—Tn Ns 1 os 1 Totrstvp Thy 1 1 2s;

Table 1: Boundary Solutions of the Optimal Market Share and Price Levels

2.4 Impact of Market Conditions on Optimal Strategies

When there is an interior solution with condition Eq 11 and Eq 12, we can see the relationship between the parameters and optimal price levels

Impact of Network externalities on Price Levels

Optimal price levels are different when the strength of network externalities is different Intuitively, when the strength of network externalities is greater, the value of the marketplace and the price are expected to be higher But, in some cases, the optimal price levels are lower with a greater strength of network externalities The following propositions state the impact of positive and negative network externalities of both supplier and buyer types on optimal price levels

Proposition 3 When the strength of the network externalities for suppliers, r,, is greater, the price charged to suppliers, p*, is higher while the price charged to buyers, px, is lower

Trang 26

if the difficulty of switching and the value of services are the same to buyers and suppliers

(Ss = 8, Us = Up) Likewise, the same result applies for buyers

Proof Because there are many mixed effects of several parameters, the direction of the change of optimal price to the change in the relative strengths of network externalities (r;, 5) is ambiguous However, when the difficulty of switching and the value of individual services are the same to buyers and suppliers, the changes in p<, pj are unambiguous The conditions are stated in Eq 24

Up = Us =U, Sụ = 85 = S$ > 551% (24)

The first derivatives of network parameters (r,,7,) on optimal prices of suppliers and buyers, ps and p; in Eq 22 are:

Op* (8sgs, — 3rÿ — 2rgrs)T„y + (4sgs; + rỆ — 3rệ — 2rgTs)3 „Uy Or, {455(8s + Tn) — (rs + 70)? }? (4rss,s; — Tỷ — 2rậr; — rụr2)0; + r„T20y + Ar2syUy + ÁTnTsSpU; {4s,(s, +) — (rs +7)?}? >0 Ops — > 0 Or, Op — —(Ñss&, — 2ryr, — 3r2)mnUy — (4sy8, + Tỷ — 2rwr, — 3r2)5suy Ôn _ {4s;(s; + r„) — (r; + r;)2}? —(ÁTySy8; — Tận; — 2ryrỆ — TŸ)U; — TỆTnUy — T2 SuUy — ÁThTnSpUs 51a <0 {4S6(5; + T„) — (r2 + 7¡)?} Op, — < 0 Or, a

When the effect: of network externalities for a type is greater, the value of the marketplace to that type is greater Furthermore, the price charged to that type will be increased (32 > 0, oe > 0) Then we may expect that the price for the other type also will be higher because of the indirect effect due to two-sided network externalities But, counterintuitively, the optimal price for a type is lower when the strength of network externalities for the other

type is greater (Be <0, ore <0) The reason for this result is as follows

Trang 27

The increase in r, has two opposing effects on p*: (1) The increase in r, causes an increase

in n, since the buyers have an increased valuation for the marketplace due to higher r, This effect increases the suppliers valuation of the marketplace which drives the intermediary to increase its profits from suppliers by increasing p, (2) An increase in r, increases the importance of n, since r,-n, is part of the profit function of buyers This result encourages the intermediary to reduce p, to attract more suppliers The strength of r, relative to r, determines which of these two effects is greater Specifically, when r, is greater than r,, the second effect dominates and the intermediary reduces p, when there is an increase in r, We derive these results assuming that the conditions in Eq 24 hold

In real life, the price for suppliers to use electronic B2B marketplaces will be lower in the reverse-auction or buyer-favored marketplaces compared to that at the forward-auction or supplier-favored marketplaces The reason why buyers favor reverse auction and suppliers favors the forward auction is that reverse auction mechanism intensifies the price compe- tition among suppliers and forward auction mechanism enables greater competition among buyers For example, Citadon, Inc runs a buyer-favored B2B marketplace in the construc- tion industry using reverse-auctions The price to buyers for using this marketplace is likely to be higher than that charged by other marketplaces that are not buyer favored Similarly, the suppliers in such a marketplace are likely to pay a lower price compared to marketplaces that are supplier favored

We derive some counterintuitive results about the optimal price levels with different levels

of negative network externalities

Proposition 4 This proposition is about the impact of changes in the strength of the negative externalities of suppliers’ side (r,) on the optimal price levels of the marketplace

(1) When the strength of network externalities for buyers is greater than that for suppliers (ru > r;) then the price for suppliers (buyers) is higher (lower) if the strength of the negative

externalities due to competition among suppliers (r,,) is greater

Trang 28

(2) When the strength of network externalities for buyers is smaller than that for suppliers (ry <r.) then the price for suppliers (buyers) is lower (higher) if the strength of the negative externalities among suppliers (r,) is greater

From the first derivatives of the parameter of negative network externalities (r,) on the

optimal prices (p%, ps) as in Eq 22, we can see these somewhat counterintuitive results We may expect that, because of the greater negative network externalities among suppliers and the resulting negative effect on the valuations of suppliers and buyers, the prices for suppliers and buyers will always be lower However, when ry > r, (the participation of suppliers is more valuable to buyers than that of buyers to suppliers), the price for suppliers is higher and the price for buyers is lower when r, increases The reason is as follows

There are two opposing effects: (1) Consider the intermediary’s optimal prices when the negative network effect is small As discussed previously and as shown in Fig 7, when rT, > Ts, the intermediary lowers p, and raises p, since the participation of suppliers is more valuable to buyers than that of buyers to suppliers However, when the negative network effect, increases, there is a reduction in the benefits from having a large number of suppliers This causes the intermediary to reduce some of the imbalance between prices charged to suppliers and buyers, leading to an increase in p, and a reduction in p, (2) The increase in r, causes a reduction in suppliers’ valuation of the marketplace which encourages the

intermediary to reduce the price charged to suppliers (p,) When ry > r,, the first effect

dominates and the firm increases p,

The value of the marketplace is lower when there is a stronger negative network external- ities effect But the optimal price already include the price lowered for suppliers to induce more suppliers to the marketplace because their participation is more valuable than that

of buyers Thus when the negative externalities are higher, the benefit of joining is lower

Then the amount of price which already lowered returns to the normal monopolistic price

levels and this effect is greater than the reduced price resulting from the reduced value of

Trang 29

Price 0.2, 0.175 | 0.15 Ƒ * ota fF ~- ~~ EBL ———= 0.1Ƒ 0.051} 0.05 Ƒ Ps 0.025 } In 0.2 0.4 0.6 0.8 1

Figure 7: Price Levels with different strength level of negative network externality: The changes of optimal prices with different strength level of negative network externality (r,) from 0 to 1 (s, = s = 1, vs = uy = 2, ry = 6,7, = 2) Here, the difference between prices

is reduced when there is more strength of negative network externality

marketplace when the value of suppliers’ joining is more valuable to the marketplace than

that of buyers (7, > 1s)

For example, in the example of Citadon, which is a reverse-auction based, buyer-favored

marketplace (r, > 1), if there are greater negative effects to suppliers from having more of

their competitors (increase in r,), the intermediary should charge more for suppliers This result seems counterintuitive because the value of the marketplace is reduced to suppliers

with this negative effect The reason is as follows Due to the greater negative effect, the

effect of positive network externalities is lower So, for the intermediary, the incentive to lower the price for suppliers to attract them is reduced when the positive effect of network externalities is reduced When the marketplace is a reverse-auction based, buyer-favored marketplace (r, > r,), the optimal price for suppliers is higher because the “price-reducing” effect from the reduced value of the marketplace to suppliers is smaller than the ‘price- increasing’ effect from the reduced incentive for the intermediary to reduce the price for suppliers

Trang 30

Impact of Individual Service Levels on Price Levels

Optimal price levels are different with changes in the level of individual services (v,, vp)

Intuitively, when the value of individual service levels is higher, the value of the marketplace and the price of it are expected to be higher But, in some cases, the price levels are lower with higher individual service levels The impact of individual service levels of both supplier and buyer types on optimal price levels is stated in the following proposition:

Proposition 5 The impact of individual service levels on the optimal price levels of the marketplace is:

(1) When the value of the individual services for suppliers, v, (buyers, vp), is greater, the price charged to the suppliers, p* (the buyers, p;), is higher

(2)a When the value of the individual services for buyers, v, (suppliers, v,), is greater, the price charged to suppliers, p: (buyers, p;), is lower if the strength of network externalities for buyers (suppliers) is greater than that for suppliers (buyers), rp > 1s (Ts > Tb)

(2)b When the value of the individual services for buyers, uv, (suppliers, v,), is greater, the price charged to suppliers, p< (buyers, p;), is higher if the strength of network externalities

for the buyers (suppliers) is smaller than that for suppliers (buyers), ry < Ts (Ts < To) From the first derivatives of the service levels (v,,v») on the optimal prices (p*, pf) as in

Eq 22, this proposition can be easily proven When the value of the individual services for a type is greater, the value of the marketplace to that type is greater Also, the price charged to that type will be increased (ee > 0, ot > 0) Then we may expect that the price for the other type also will be higher because the greater value of one type increases the value of the other type by indirect effect But, counterintuitively, the optimal price for a type is lower when the value of the individual services of the other type is greater if the strength of network externalities of the type is greater than that for the other type The reason for this

result is as follows

Trang 31

The increase in œ has two opposing efects on ø;: (1) The increase in 0; causes an

increase in nz since the buyers have an increased valuation for the marketplace due to higher up This increases the suppliers valuation of the marketplace which drives the intermediary to increase its profits from suppliers by increasing p, (2) An increase in vu, forces the intermediary to reduce p, to increase n, to maximize its profit by manipulating the size of n, and n, Specifically, when r, is greater than r,, the second effect dominates and the intermediary reduces p,

For example, when more individual services are provided for suppliers, more suppliers join the marketplace and the intermediary can charge higher price to suppliers But the intermediary should carefully consider the pricing decision for the price charged to buyers In the reverse-auction based, buyer-favored market (r, > 7,), the participation of suppliers is more important than the participation of buyers When more individual services for buyers are provided, the value of the electronic marketplace is greater for suppliers because of the indirect effect from having more buyers in the marketplace But by reducing the price for suppliers and attracting more suppliers to the marketplace, the profit for the intermediary

can be increased

Impact of Switching Costs on Price Levels

Optimal price levels are different with the different levels of difficulty of switching (s,, s,)

Intuitively, when switching to electronic marketplace is more difficult, the value of the mar- ketplace and the price of it are expected to be lower But, in some cases, the price levels are

higher with higher difficulty of switching The following is the proposition about the impact

of switching difficulty of both suppliers and buyers’ sides on optimal price levels

Proposition 6 The impact of switching costs on the optimal price levels of the marketplace

are:

(1) When the strength of network externalities for buyers is greater than that for suppliers

(r, > 1), the price for suppliers (buyers) is higher (lower) if the difficulty of switching for

Trang 32

Price 0.4 r 0.35 | 0.3 | 0.25 Ƒ * 0.2r \ Pp 0.15 | ¬ 0.1 | eee es TT” 0.05 | Pg

Figure 8: Price Levels with different levels of the difficulty of switching: The changes of optimal prices with different level of the difficulty of switching for suppliers (s,) from 17 to

1.5 (89 = 1, vs = Uy = 2, Ty = 4,75 = 2, Tn = 05) Here, the difference between prices is

reduced when there is higher difficulty in switching

suppliers or buyers, 8, and Sp, is greater

(2) When the strength of network externalities for buyers is smaller than that for suppliers

(ry < rs), the price for suppliers (buyers) is lower (higher) if the difficulty of switching for

suppliers or buyers, s; and sy, 1s greater

From the first derivatives of the switching cost parameters (s,, s,) on the optimal prices

(p*, pg) as in Eq 22, this proposition can be easily proven We may expect that, because of the higher difficulty of switching, the electronic marketplace becomes less attractive and the price for suppliers and buyers would be lower if it is more difficult for suppliers to switch to the electronic marketplace However, when r, > r, (the participation of suppliers is more valuable to buyers than that of buyers to suppliers), the price for suppliers is higher and the

price for buyers is lower when s, or s, increases The reason is as follows

There are two opposing effects: (1) Consider the intermediary’s optimal prices when the difficulty of switching to the electronic marketplace is small As discussed previously and as shown in Fig 8, when r, > r,, the intermediary lowers p, and raises p, since the participation of suppliers is more valuable to buyers than that of buyers to suppliers However, when the

Trang 33

difficulty of switching increases, there is a reduction in the benefits from having a large number of suppliers This causes the intermediary to reduce some of the imbalance between prices charged to suppliers and buyers, leading to an increase in p, and a reduction in py (2) The increase in s, or s, causes a reduction in suppliers’ valuation of the marketplace which encourages the intermediary to reduce the price charged to suppliers (p,) When r, > 7s, the first effect dominates and the firm increases p,

We summarize the propositions made in section 2.4 Parameter (Increase) pe |p, | Assumptions Ts + - 83 = 8b, Tọ - + Us = Up Tn - + when r, > Tp - when r, < Tp Us + + when 7s < rp - when 7; > r; Up + + when rs < 1p - when r, > r7; Ss + - when r, < Tp - + when r; > Tp Sb - + when r, > 1 + - when r, < Tp Table 2: Summary of Propositions of Section 2.4 2.5 Conclusion

Our theoretical analyses have shown that the existence of network externalities and the mutual effect of the buyer’s market on the supplier’s market and vice versa affect the optimal pricing strategy and optimal market share level of B2B intermediaries From our model, we derive the pricing strategy that an independent intermediary can use to maximize its profit

Trang 34

How much should the intermediary charge? Based on our assumptions, the market shares

as well as the prices charged from the two types: suppliers and buyers in the electronic marketplace are dependent on the parameters that define the two types In other words, an impact on the conditions of one type also affects the market share and price levels of the other type because of the existence of two-sided network effects Also, when the relative strength of network externalities of a certain type is greater, that type pays a higher price to the B2B intermediary Thus the type, either buyer or supplier, to whom the network externalities or aggregation benefit is more important compared to the other type, would pay more to the B2B marketplace and exploit the competition among the other type When the individual service for a type is more valuable, the price charged to that type is higher But the impact of the different levels of individual service on the price charged to the other type depends on the relative strength of the network effect For example, in reverse-auction based, buyer-favored marketplace, we can expect that the price charged to suppliers will be higher when more individual services are provided for buyers leading to greater participation of the buyers But, counterintuitively, since the value of participation of suppliers to the electronic marketplace is more valuable than the participation of buyers in a buyer favored marketplace, therefore the intermediary should lower the price charged from suppliers to attract more suppliers Our results including this example suggest that intermediaries should have a broader vision in designing their marketplaces to increase their profits in electronic marketplaces

In this research, we assume a monopolistic and independent intermediary with a single period model We intend to extend this analysis to a dynamic model with multiple periods We also hope to analyze a duopoly model which incorporates competition between interme- diaries in electronic markets in our future research Due to some characteristics of the B2B marketplace such as network effects, it is possible that, in certain segments, only one mar- ketplace will survive In other segments, the emergence of a dominant marketplace is likely In financial markets, for instance, a single exchange tends to receive most of the transactions

Trang 35

and profits In such cases our monopoly model is a close approximation, and we think the propositions developed in this model are applicable to such real-world situations Also, we hope to analyze different ownership structures of B2B marketplaces which involve suppliers and buyers who own the marketplace From these extensions, we expect to propose models of B2B marketplaces that are owned by suppliers or buyers and to compare the results on the price levels and market share levels to the results of the current model of an independent intermediary

Trang 36

3 Ownership Structure Part II: Neutral vs Biased

marketplaces!

3.1 Introduction

The recent rapid growth in electronic commerce led to unrealistically high expectations The bursting of the dot com bubble and the bankruptcy of many well known e-tailers such as pets.com and egghead.com has created skepticism about the profitability of Internet ven- tures Media attention was focussed largely on business-to-consumer firms, while far greater volumes of transactions were conducted in business-to-business markets According to eMar-

keter [27], the level of business-to-business (B2B) electronic revenues is six times greater than

that of business-to-consumer (B2C) electronic revenues in 2001, and is expected to be 12 times larger than that of B2C electronic revenues in 2004 In this paper, we examine B2B marketplaces that provide the infrastructure for B2B transactions In particular, we focus on the impact of the ownership structure of such marketplaces on their profitability, and the level of participation from buyers and suppliers

There are different types of electronic B2B marketplaces We refer to marketplaces such as Citadon and Chematch which are owned by independent third parties as neutral market- places, while we call marketplaces such as Covisint and Quadrem which are owned by either suppliers or buyers as biased marketplaces Each type has its strengths and weaknesses Independent intermediaries who own neutral marketplaces claim that they have expertise in information technology and credibility in terms of neutrality But they also have several limitations, an important one being the lack of commitment of buyers and suppliers Since they are independent from both suppliers and buyers, there is no guarantee that buyers and suppliers will adopt such a marketplace There is a high risk of failure when a neutral

1Barly results from this work were published as a two page abstract in the ICIS 2001 proceedings and won the best completed research paper award

Trang 37

marketplace is unable to attract sufficient numbers of suppliers and buyers

Biased marketplaces claim to be in a better position to offer successful marketplaces since they are guaranteed to receive a large volume of their owners’ transactions helping them achieve sufficient liquidity They also claim to possess industry-specific expertise which pro- vides them an advantage relative to neutral marketplaces However, as observed in the case of marketplaces for airline tickets and automobile parts, there is considerable concern about the fairness of these biased marketplaces Buyer/supplier owned marketplaces raise the pos- sibility that the owners of such marketplaces could devise trading rules or use proprietary market information to gain unfair advantage Covisint, the automobile parts marketplace developed by major automobile manufacturers, was investigated by the Federal Trade Com- mission concerning the possibility of a price cartel and other antitrust-related issues The owners of biased marketplaces claim that such markets will be fair and that it would allow

them to reduce their transaction costs [41] and to acquire the asset value of the marketplaces resulting from their own transactions [68]

There are numerous neutral and biased marketplaces with the strengths and weaknesses discussed above Similar phenomena can also be observed in B2C marketplaces For ex- ample, Orbitz was recently created by the major airlines to compete with existing neutral marketplaces such as Expedia and Travelocity Such competition between neutral and biased marketplaces is likely to occur in many industries

However, there are several unanswered questions in the context of designing B2B mar- ketplaces on the Internet What is the role of ownership structure in determining the com- petitiveness of these marketplaces? Which marketplace will be better in terms of the surplus generated to each party: intermediaries, suppliers and buyers? How are the benefits likely to be shared among buyers and suppliers in biased marketplaces? What is the impact of frag- mentation among buyers and suppliers? Many practitioners speculate on these questions, but there is little academic research to provide answers to such questions In this paper, we

Trang 38

analyze and compare neutral and biased marketplaces in terms of fractions of buyers and suppliers participating in electronic marketplaces, price, and surplus generated when markets exhibit network effects We also examine the social welfare of different marketplaces

The article is organized as follows: Section 3.2 briefly reviews previous literature related to B2B marketplaces Section 3.3 discusses the neutral marketplace model and Section 3.4 discusses the biased marketplace model Section 3.5 compares the results from the two models Section 3.6 discusses the managerial implications of our results Finally, Section 3.7 presents our conclusions

3.2 Literature Review

The benefits of electronic B2B marketplaces can be categorized into two The first is

the speed and efficiency of transactions enabled by information technology Using advanced

information technology, suppliers and buyers can reduce transaction costs The second ben- efit accrues from the larger number of participants By bringing together a large number of buyers and sellers, electronic marketplaces increase liquidity Due to the ease of searching

for suppliers in electronic marketplaces, buyers have a greater chance of obtaining cheaper

prices and better quality products Suppliers find it easier to identify potential buyers in their efforts to sell products This aggregation benefit [35] that an intermediary provides to buyers and sellers is akin to a network effect with some distinguishing features that we elaborate later in this section This network effect is an amalgam of price effects, product variety, availability etc Bakos [5] categorizes the functions of marketplaces in three cate- gories: matching, aggregation functions and providing an institutional infrastructure such as legal and regulatory framework He expects that electronic marketplaces will leverage information technology to perform these functions with increased effectiveness and reduced transaction costs, resulting in more efficient ‘friction-free’ markets

When the value of a product depends on the number of users, the product exhibits

Trang 39

network effects Katz and Shapiro [36] show that network externalities affect firms’ decisions

about the compatibility of their products In this paper, the marketplace is assumed to exhibit network externalities with some characteristics that distinguish it from the network externalities discussed in previous studies [13, 14, 36] We refer to the network externalities modeled in previous studies as one-sided network externalities to differentiate it from the network externalities of the marketplace modeled in this paper Unlike the one-sided network externality of products which results from the size of the customer base of the products, network externalities in the marketplace exhibit an important difference: for each buyer

(supplier), the value of the marketplace is dependent on the number of suppliers (buyers)

The existence of a two-sided network effect implies that the buyers’ environmental at- tributes have an impact on the suppliers and vice-versa For example if the security environ- ment makes it expensive for buyers to conduct their transactions online, then buyers would demand lower prices to compensate for such costs Some marginal buyers and suppliers will choose not to participate in such a marketplace This reduces the volume of transactions on the marketplace further reducing the number of buyers and suppliers that participate In this research, we compare biased intermediaries to neutral intermediaries when the marketplace exhibits two-sided network effects We present generalized results for the case of non-linear

network effects A linear network effect implies that the marginal value of an additional

buyer or supplier is always the same - no matter the size of the network For example, this would imply that the increase in network effects is the same whether the supplier (buyer)

being added is the second supplier (buyer) or the twentieth supplier (buyer) However, it is

intuitively clear that the network effects increase at a diminishing rate The addition of a buyer or supplier in a marketplace is likely to have greater impact when there are few buyers and suppliers in the marketplace For example a buyer is likely to reap significant benefits if the number of suppliers for cold-rolled steel increases from one to two, and comparatively smaller increases in benefits if the number of suppliers increases from ten to eleven In this

Trang 40

(Intermediary) Supplier 3 Supplier +

Figure 9: Two-sided network externality of the marketplace

paper, we focus on the comparison of neutral marketplaces and biased marketplaces with generalized concave network effects As seen in Fig 9, buyers in the marketplace care about the number of suppliers and suppliers care about the number of buyers In addition, the value of the marketplace decreases for each supplier when there are more suppliers because of competition among suppliers We call this ‘negative network externalities.’ Wang and Sei- dmann [63] show that the participation of more suppliers can generate positive externalities for the buyer and negative externalities for other suppliers in an electronic data interchange (EDI) network In this paper, we model negative network externalities only for the supplier side since we assume that there are no capacity constraints on the supply side

While there are benefits from joining electronic marketplaces, some switching costs may also exist When suppliers and buyers join these marketplaces, they have to invest in new technologies to adapt their current distribution channels In previous EDI research, many studies examined how EDI networks could attract suppliers and buyers from traditional

channels Riggins et al [53] show how a buyer can attract suppliers to its EDI network

in the presence of network effects, and Barua and Lee [8] show how to subsidize suppliers to attract them when an EDI system is introduced In previous studies, researchers have noted that network effects in the marketplace depend on the number of participants But few researchers have disaggregated participants into buyers and suppliers and examined the network effect of each type of participant on the value of the marketplace In this research,