Tilburg University
Essays on marketing strategy in technology-intensive markets
Stremersch, S.
Publication date:
2001
Document Version
Publisher's PDF, also known as Version of record
Link to publication in Tilburg University Research Portal
Citation for published version (APA):
Stremersch, S. (2001). Essays on marketing strategy in technology-intensive markets. CentER, Center for
Economic Research.
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal Take down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
-. -.
Essays
on
Marketing
Strategy in
Technology-Intensive
Markets
,//
Ditwerk terug tebezorgen uiterlijk op:
1 3 IUN! 2002
101.07-·oz
Bibliotheek - Katholieke Universiteit Brabant
Postbus 90153 5000 LE Tilburg
Essays
on
Marketing
Strategy in
Technology-Intensive
Markets
Proefschrift
ter verkrijging van de graad van doctor aan de Katholieke Universiteit Brabant, op gezag van
derectormagnificus.prof.dr. F.A. van der Duyn Schouten, in het openbaar te verdedigen ten
overstaan van een door
hetcollege voor
promoties aangewezencommissie in de aula van de Universiteit opvrijdag 7 december 2001 om
11.15 uurdoor
StefanStremersch
Geboren op29 maart 1972 teSint-Niklaas. Belgie.
, / / ..=,Mk
Preface
It is customarytodirect a word
of
gratitude to people thathaveshown enthusiasm and support over thecourseof
writing
a dissertation. in my case, this acknowledgement is not merely thecontinuation of
such custom, butrather it is
a sincere expression of the warm feelings, professionally and personally, 1 havedeveloped forthepeople 1 wish to mention at this crossroads of myacademiccareer. Firstandforemost, I wishto thank my advisors, Ruud Frambach (FreeUniversityAmsterdam)andGerry Tellis (Universityof
Southern California), for theirenthusiasm, support, commitment and high standards. As for Ruud, 1 look back on valuable intense debates on mydissertation and research in general. 1 have shared an openmind with him
for almost four years and always found our interactions veryfruitful and
rewarding. As for Gerry, Iam grateful thathetrained me in his research approach and have tremendous respect forhis productivity and enthusiasm. On a personal note, Gerry. Cheryl and the kids often welcomed meintheir home and made me feel part ofthefamily. This hasbeen animportant support formeduring the year and a half 1 visited with USC.
I am also grateful to the people 1 have worked with in the pastyears. Allen Weiss
(University
of
Southern California) strongly inspired my thinking on technology-intensive markets. 1will
always remember the long andfruitful talks we had
on research and philosophyof
science,during my visit at USC. 1 also look back andforward ontheexcellent co-operation with Benedict Dellaert (Tilburg University), Shantanu Dutta(University of
Southern California), Philip Hans Franses (Erasmus University Rotterdam), Stefan Wuyts (Erasmus University Rotterdam) and Eden Yin (Cambridge University). The interest and commitmentthey showed in ourjoint
projects as wellastheir intelligencereally amazed me at times.A
very special and warm wordof
gratitude goes to Harry Commandeur (Erasmus University Rotterdam) who has shown amazing supportwithout any sign of
self-interest.When providing me with
his wisdom and sympathy over the course of the past years, he always had mybest interest in mind. Harry, 1 do hope that at some point 1canreturn some of the favors you granted me! I alsothankJacquesDe Rijcke (Universityof
Ghent) who was the firsttoawakenmyacademic interest.I thank Philip Hans Franses, Hans Pennings, Jan-Benedict Steenkamp and Michel Wedel for
joining
the committee. 1 thank StefanWuyts and Eden Yin
forbeing my
"paranimfen". I amgrateful tothehigh-levelacademic setting thatthemarketingdepartment
addition. I would wantto thanktheIntercollegiateCenterforManagementScience (ICM) for generously providing financial supportduring the years of my PhD education. I alsowould liketo acknowledge the many otherinstitutesand companies that provided financial support to my research, more in particular The Marketing Science
Institute, The Institute for the
Studyof
Business Markets, the CentERfor
Economic Research (Tilburg University), The Goldschmeding Centerof
Economicsfor
Increasing Returns (NyenrodeUniversity) and
AlcatelTelecom. I also wishtothank Rajesh Chandy(University
of
Minnesota),PeterGolder (NewYork
University), DebbieMacInnis (Universityof
Southern California),R. Venkatesh (Universityof
Pittsburgh) and EdenYin
(Cambridge University), amongothers, for
providing me withvaluable commentsonearlierdrafts
of
chapters 2-5.In addition tothe academiccommunity, many managersprovided support in one way or another tomy research. I owe many thanks to DerrickGosselin (Alcatel Telecom) for his support tothis dissertation. Despite a busy schedule, he always was
willing
to answer some of my questions andhe posedastronginterest inmycareeringeneral. 1 alsoowegratitude to Henk Speijer(PhilipsConsumerElectronics) and Derek Baine (Paul Kagan Assiociates) forsharing their insights and for
helping me
with gathering data on consumer electronics (Chapter 5).Finally, only a fool thinks he can separate work fromhis private life. My stay of 18 months in LosAngeles hastaught me a lot ontheimportance
of
quality family time. Els, to you go mygreatest feelingsof
gratitude. We both know that 1 would not be what Iam today without you nurturingandcomforting me. Tomy parents, I owegratitude forthe interest andsupport they showed. They gave me a strong work ethic and the
feeling that you can
accomplish
anything if you go for it.
Many thingswill
change in thefuture, but this will
always remain the same.Contents
Chapter 1: Introduction and Outline 1
1.1 MarketingStrategy inTechnology-Intensive
Markets 1
1.2Motivation,Objectives andOutline
of
Dissertation 4I.2.1 Motivation 4
1.2.2 Objectives 5 1.2.2.1 SpecificObjectives
of
Chapter 2 51.2.2.2SpecificObjectives
of
Chapter 3 6I.2.2.3SpecificObjectives
of
Chapter 4 61.2.2.4SpecificObjectives
of
Chapter 5 61.2.3 Outline 7
Chapter
2: StrategicBundling of
Products andPrices: ANewSynthesisfur Marketing
1011.1 Abstract 10
II.2 Introduction 11
11.3 A Primer onBundlingStrategies 13
II.3.1 Definitions 13
II.3.1.1 ConfusioninLiterature 13
II.3.1.2
Bundling 14
II.3.1.3 BundlingFocus: Product versus PriceBundling 14
II.3.1.4 Bundling Form:Pureversus
Mixed. 15
II.3.2 ClassificationandLegality
of
BundlingStrategies 1611.3.2.1 Classification
of
BundlingStrategies 16
Il.3.2.2 Legality
of
Bundling 18Il.3.2.3 Legality
of
Various BundlingStrategies 2011.4 Optimality
of
BundlingStrategies 21II.4.1 Simulation 22
II.4.2Consumers'ConditionalReservation
Prices 25
II.4.2.1 Heterogeneity
of
Conditional ReservationPrices 2511.4.2.2Price
Bundling 27
Il.4.2.3Product
Bundling 30
II.4.3Objectives
ofthe Firm 33
II.4.4 Competition 36
Il.4.5 Costs 38
II.4.6Consumers' Perception
of
Bundles 4011.5 Conclusion andDirections
for
FurtherResearch 43
II.5,1 Implications for Marketing Management 43
11.5.2 Limitations and FurtherResearch 45
Chapter
3:Buying Modular
Systemsin Technology-IntensiveMarkets 47
III.2 Introduction 48
Ill.3
ConceptualFramework 51
III.3.1
Know-how 52
III.3.1.1 Transferability
oftacit know-how 52
III.3.1.2Presence
of
know-how 54111.3.2Technological
environment 56
III.3.2.1 Technological
uncertainty 56
III.3.2.2Interactions betweenTechnological Uncertaintyand Know-How 58
III.3.3 Control
variables 59
III.3.3.1 Transaction-specific
assets 59
III.3.3.2Supplierconcentration inthesystemintegration market 60
III.3.3.3 Importance ofthesystem 60
III.3.3.4Expertiseon
suppliers 61
III.3.3.5Componentsupplier
specialization 61
III.3.4Other Interaction
effects 61
111.4 Method 62
III.4.1 Research
context 62
Ill.4.2
Researchdesign 62
III.4.2.1 Benefits
ofa
conjoint study overaretrospectivestudy 64111.4.2.2Benefits ofarating task overaranking/choice task 65
III.4.3 Measuresfornon-manipulatedvariables 66
111.4.4 Pre-testing anddata
collection 67
Ill.4.4.1
Pre-testing 67
Ill.4.4.2 Data
Collection 67
III.5 Analysis and results 69
III.5.1 Model
specification 69
III.5.2
Results 71
111.6 Discussion 76
111.6.1 Theoretical
Implications 76
III.6.2 Implications formarketing
management 78
III.6.3 Limitationsanddirections
for
futureresearch 79
Chapter
4: Buying BehaviorinTechnology-IntensiveMarkets: The Role of
Embeddedness 81
IV. 1 Abstract 81
IV.2 Introduction 82
IV.3 Conceptual
Background 84
IV.3.1 Relational
Embeddedness 86
IV.3.2Structural
Embeddedness 87
IV.3.3 Direct
access 88
IV.3.4 Know-How 90
IV.3.5 Pace of Technological Change 90
IV.3.6 Other variables 92
IV.3.6.1
Valence 92
IV.3.6.2 Number
of Third Ties 93
IV.4 Methodology 93
IV.4.1 Research
Design 93
IV.4.2 Sampling andData
Collection 94
IV.4.3 Development
of
Questionnaire 95
IV.4.5 Test
of
Hypotheses 98IV.5Discussion 100
IV.5.1 Theoretical Implications 100
IV.5.2 Managerial Implications 102
1V.5.3 LimitationsandFuture ResearchDirections 103
IV.6APPENDIX:ScaleMeasures 104
IV.6.1 DirectAccess to Component Manufacturers 104
IV.6.2 Know-How onComputerNetwork 104
IV.6.3 Pace
of
Technological Change 104Chapter 5: Does it Really Take Two To Tango? Modeling
Indirect Network
Effects inHardware
andSoftwareGrowth
105V. 1 Abstract 105
V.2 Introduction 106
V.3 The Role
of
indirect Network Effects in NewProductGrowth 109V.3.1 The Effect
of
SoftwareAvailability
onHardwareSales 111V.3.2 Determinants
of
SoftwareAvailability
113V.3.3 WhichComes First, the
Takeoff
in Hardware Sales or theTakeoff
inSoftwareAvailability? 115
V.4 Modeling Indirect Network Effects 117
V.4.1 Multivariate Regime-Switching Model for Indirect NetworkEffects 117
V.4.1.1 Model Structure 118
V.4.1.2 Interpretation
of
StructuralParameters 120V.4.1.3Model Advantages 121 V.5 Data 121 V.5.1 DataCollection 121 V.5.2 Sampling 122 V.5.3 Measures 123 V.6 Model Results 124 V.6.1 Estimation 124 V.6.2ParameterEstimates 125 V.6.2.1 TakeoffParameters 125 V.6.2.2 Structural Parameters 126 V.6.3 Model Evaluation 127 V.6.3.1 Fit 127 V.6.3.2 Forecasting 129 ProductMarket 130 V.7Discussion 130 V.7.1 Summary
of
Findings 131 V.7.1.1 Takeoff 131V.7.1.2 IndirectNetworkEffects and SoftwareProvision 131
V.7.1.3 Modeling indirect Network Effects 131
V.7.2 Implications 132
V.7.3 Limitations andFurtherResearch 133
Chapter
6: ConclusionsandDirections for
Future
Research 135VI. 1 Introduction 135
VI.2.1 Bundling(Chapter 2) 136 VI.2.2 Buying ModularSystemsin Technology-Intensive Markets(Chapter 3) 137 Vl.2.3 Buying BehaviorinTechnology-Intensive Markets: The Role of
Embeddedness (Chapter 4) 138
VI.2.4 NewHardwareLooking
for
Software: Modeling Indirect NetworkEffects(Chapter 5) 139
VI.3 DirectionsforFuture Research 141
References 144
Chapter 1
Introduction
and
Outline
In this section, 1
briefly
review the field this dissertation aims tocontribute to. This review is notmeant to be acomprehensiveoverview ofthe literature. Rather it is apersonal perspective on the literature, relevant to marketing, on Technology-Intensive(TI)
markets and its gaps. Next, 1will
developthemotivation, objectiveandoutline of
thedissertation.I.1 Marketing Strategy in Technology-Intensive Markets
Marketing scholars onlyrecently started to explore the strategicmarketing challenges firms in technology-intensive markets face. At the time 1 started this dissertation in 1998, there probably were not even a dozen
of
articles in the major marketing journals, which focused on hightechtopics. But by the turn ofthemillenium -
inwhichperiod Nasdaq first rose from 1000 to 5000points in3 years and then camedownagain to 2000 points in 1 year -hightech marketscan count onarapidly growingattention from marketingscholars, whichcan benoted from the increasing numberof journal articlesand conference presentations on
high tech topics. The
uprising and fall
of
Nasdaq definitely created a buzz on high tech. Managers at tech companies wereliving the
new American Dream, which before was exclusivelyreserved to pop stars, moviecelebrities andTV
personalities. Marketing scholars aregrowinglybecomingaware ofthe specific challenges high tech companiesface,about 20yearsaftereconomists did and only after many high tech companies added many
billions of
dollars totheir market valuation.
Technology is scientific knowledge applied to useful purposes (Quinn, Baruch. and
Zien 1997). Technology is at the core of many contemporary markets and is a powerful
engine
of
today's andof
tomorrow's economic progress (John, Weiss and Dutta 1999). Although they span adiverse rangeof
industries, such as homeentertainment (TV, VCR andstereo system), telecommunication (transmission, switches, routers). computing (PCs,
computer networks, software) and medical systems (optical systems, laser), technology-intensive
(TI)
markets share similarities that make them into unique environments with unique challenges toourunderstanding (Glazer 1991).2 Introduction and Outline
problems because
of
their knowledge-intensity and technologicalturbulence. Not only is
knowledgeaprimeassettowards strategic advantage inthesemarkets (Glazer 1991), but also technologies are changing continuously making the associated knowledge bases obsolete (BourgeoisandEisenhardt 1988).
What are the theoretical problems that knowledge intensity and technological turbulence bear? We recognize two main theoretical domains. First, technology changes monolithic products into systems that typically consist
of
'technologicallydivisible'
components,joinedtogether by a set
of
interfaces thatallowsthecomponents to work (Katz and Shapiro 1994, Wilson. Weiss and John 1990). Second, in view of the importance and rapid obsolescenceof
knowledge, managing knowledge and informationflows is key to
companies inTI markets, forboth vendors and buyers (John, Weiss and Dutta 1999).
The recognitionthattechnologyturnsmonolithic productsintosystemshasnourished
most of
the research in economics onTl
markets. Three interrelated literature streams flow fromthis insight.A first stream
of
literature focuses on standardization ofthe interfaces between the different components ofa system and competition between different standards (e.g. Besenand Farrell 1994; Economides 1989: Einhorn 1992; Farrell, Monroe and Saloner 1998; Farrell and Saloner 1985,1986,1992; Matutes and Regibeau 1988). Standardization is an explicit or
implicit
agreement todo certain keythings in auniform way (Farrelland Saloner 1992).Thisliterature provides insights in issues such as: When isstandardization optimal for companies? What is the effectof
standardization on consumer surplus and social welfare?What is the influence
of
standardization on the industry structure and competition? This literatureislargely ofananalyticalnature.A second stream
of
research discussesnetwork
effects, whichrelates to the
increasing utility of the system with increasing adoption of system components (e.g. Brynjolfsson and Kemerer 1996: Church and Gandal 1992, 1993; Economides and Salop
1992, Gandal 1994; Gupta, Jain and Sawhney 1999: Katz and Shapiro 1985, 1986, 1992,
1994: Majumdar and Venkataraman 1998; Saloner and Shepard 1995). In case
of direct
network
effects,theexpectedutility
of
hardwareincreases directly withthenumberof
otherusers there are. Examples ofsuch products are the telephone. fax and E-mail (Shapiro and
Varian 1999). In case
of
indirect network
effects, the expectedutility of
the hardware increasesasavailabilityof
software increases (Economides 1996).Examples ofsuchproducts aretheCompact Disk Player, RadioandTelevision. Thisliterature providesinsights in issuesChapter 1 3
effects affect the choice betweenincompatible technologies? What isthe impact
of
network effects on standardization? Howdo firmscompete in networkmarkets? Cannetwork effects lead to socially undesirable outcomes (e.g. theadoption of
an inferior standard)? Thisliterature is of
an analytical nature, with rare empirical testing mostly through casualobservation.
Finally, a third,
although by far
the least developed, literature stream on systemsstudies the bundling
of
interdependent products (Bakos and Brynjolfsson 1999, 2000. Matutesand Regibeau 1992; Wilson, Weiss and John 1990). Bundling is the sale of two or more products inapackage (StremerschandTellis 2002). This literature provides insights inissues such as: Whenis which bundlingstrategyoptimal? What istheimpact
of
competitionontheoptimality
of
bundling? Alsothisliteraturestreamismainly of
ananalytical nature. A secondtheoreticaldomainrefers tothefocal role of
themanagementof
knowledge and information inTl
markets. Scholarshavegeneratedconsiderable research in this domain, which consistentlyrelatesto organizationaldecision-making inTI
markets(e.g.Acllrol 1991,Bourgeois and Eisenhardt 1988, Caponand Glazer 1987; Dutta and Weiss 1997; Eisenhardt
1989; Glazer 1991; Glazer and Weiss 1993; Hansen 1999; Heide and Weiss 1995; John,
Weiss and Dutta 1999; Judge and
Miller
1991; Tushman and Anderson 1986, Weiss and Heide 1993). This literature provides insights in issues such as: What are optimal decision procedures inTl
markets? What is the role ofthe marketingfunction in TI markets? What kindof
interfirm partnerships between companies are optimal in technology-intensive markets? What are the determinantsof
search. consideration and choice behaviorforbuyers in TI markets?Mostpapers in this literature streamareeitherconceptual orempirical.In sum,aspresented inFigure 1.1, I distinguish twomaintheoreticaldomainsrelevant tomarketing, namely hightechsystems andknowledge andinformationmanagement.Within
thefirstdomain, we recognize threeliteraturestreams thatarerelevant to marketing, namely on standardization, onnetworkeffects and onbundling.
Within
the seconddomain, we only recognize one consistent research stream that is relevant to marketing,namely on
4 Introduction and Outline
Research on Marketing Strategic
( Issues in Technology-Intensive
\ »» Markets
2
THEORETICAL DOMAINS
Knowledge and
High
Tech
InformationSystems Management
4
RESEARCH
STREAMS
-,A Standardization Bundling v Organizational V Decision-Making NetworkEffectsFigure 1.1 Research on Marketing Strategic Issues in Technology-Intensive Markets
I.2 Motivation, Objectives and Outline of Dissertation
I.2.1 Motivation
My
overallmotivationwasdriven bythegrowingrelevanceof
technologyto today's society.TI
markets contributed an estimatedfull
percentagepoint to thegrowth of the US
industry in 1998. And its share in the USeconomycontinues togrow.sinceTI markets grow at twice the rate of the US economy overall (John, Weiss and Dutta 1999), continuously increasing the impact technology has on today's economy. Also technology is growingly
Chapter 1 5
It is
thisgrowing presence in andrelevance totoday's society and therecognition that high tech isdifferentthatinitially
motivatedthis research more than three years ago.Surprisingly, incontrast with thishigh relevance thereisscarceattention forthetopic in marketing literature. Economists started to debate the particulars of high tech systems, standardization and standard
battles in the 60s and
70s. Economics researchers such as Michael Katz, Joseph Farrell, Stan Liebowitz, and Carl Shapiro, to name a few,built
their fame on technology topics,while thefirst
high-quality article on technology and marketing appeared inJoitrnal of
Marketingonly in 1987. The mix of high relevance and scarce
attention proved to be atrueopportunity. Overthe course of myPhDeducation, I gradually have seen an increasing numberof
peers, as well as research institutes andjournals, move intoandencouraging research in this area.I.2.2 Objectives
The different chapters clearly delineate the shortcomings of the literature, their objectives and contribution. Nonetheless, it is important to portray the overall
objective of
this dissertation and how the specific objectives ofthe different chapters are related to the centralobjective.
This dissertation
posits that the
most importantshortcoming of
the marketing literature on technology is that papers in marketing do not use the focalinsight that in
technologymarkets,monolithicproducts changeintosystems.This insighthasdriven most of the economics literature on technology, bus is almostcompletely neglected in themarketing literature. Therefore,thegeneralobjective ofthis dissertation is "to show how transactions of'
high tech systems are different from transactions of monolithic products and what the
implications of these differences are towards strategic marketing of such systems". As shown before. this "systems" approach has been the dominant research stream on technology in economics. To introduce this ideainmarketing literature is a verysubstantialcontribution to the technology in marketingresearch area. In addition,we develop more specific objectives forthedifferentchaptersalong our generalobjective.
1.2.2.1 Specific Objectives of' Chapter 2
6 Introduction and Outline
substantial differences. The specific objectives of chapter 2 are to: (1)clearly delineate price from product bundling: (2) discuss legality differences between the two', and (3) develop propositions on their optimality.
I.2.2.2 Specific Objectives of Chapter 3
Chapter 2 leads us to recognize the
focal role
of
integration
to distinguish the bundlingof
systems from mere price bundling. This insight leads toa theory onbuying of
modularsystemsasdeveloped in Chapter 3. in thischapter, we recognize that buyers have to
makedecisions towards the system components and towards the system's integration, when they buy a high tech system. Specific objectives of this chapter are to: (1) clearly discern the different bitying aspects of high tech systems. namely outsourcing of system integration and purchase concentration of the system components: 0) develop a consistent theory, grounded in technology literature. that explains buyers' preferences towards these two decision dimensions.
1.2.2.3 Specific Objectives of Chapter 4
In Chapter 3, we studyoutsourcing
of
systemintegration. What we donot study is thechoice of
aspecific system integratoror vendor. Since we posited in Chapter 3 that system integration isa processof
generating and exploiting know-how,economicsociologyteaches us that the attractivenessof
system vendorswill
depend to a large extent on thestructure of relationships (embeddedness) between system vendor and component manufacturers.Consequently, the specific objectives of Chapter 4 are to:
(1) argue atidshow that the
structure of relationships between system vendor and component manufacturers determines asystem vendor's attractiveness to btiyers. (2) examine how factors that are focal to TI
markets, such as the know-how buyers possess and the pace of technological change they perceive, injluence the relationship postulated under (1).
I.2.2.4 Specific Objectives of Chapter 5
Chapter 1 1
1.2.3 Outline
We explore four distinct theoretical issues in chapters 2 to 5 (see Figure I.2).
A
firstissue we explore extends upon the literature on the bundling
of
systems (Matutes andRegibeau1992; Wilson,Weiss and John 1990). Since systems are a set
of
componentsjoined by an interface, theoptimalityof
bundlingbecomesanimportantconcern. Chapter2providesan overview of
the bundling domain, and consequently discusses legality and economic optimalityof
strategies within this domain. More in particular, while the distinction this chapter makes between price andproduct bundling isquitegeneral, it is especiallyvaluableto Tl
markets.Price bundling is the sale of two ormore separateproducts asapackage at a discount without any integration ofthe products.Product bundling is
the integration and sale of two ormore separateproducts at anyprice.TI
companies often canchoosebetween unbundling,pricebundlingandproductbundling oracombinationof
thesestrategies.Conceptual
I
ECONOMICS
/4'ho does sys,I>\
Whenisbundling
(integration' Extent of'I
ofsystems
,omponent purchasej /
optimal?
Jiconcentration 5/
1/lit.-Ill-****.-dv
High
Tech
Systems'
..7
/ 'fo which extent'ar
driveWhich factorsabuyer's// iii..1 system components: )
\ interdependent? / preferences for a
/-
system vendor?ECONOMIC SOCIOLOGY
1
NetworkEffects Organizational Theory
ArchivalData Experiment/Survey Data
TimeSeries Limiteddependent variables
8 Introduction and Outline
Chapter3 elaborates on thisnotion
of
integrationbetween system components that is so typical to product bundling. It also studies purchase concentration towards the system components to create a theoreticalframework on
the organizational buyingof
modular systems in TI markets. Buyers thatconsider buying modularsystems need to consider two distinct decisions. The first decisionrelates to who does
theintegration of
the system components into afully
integratedsystem. A firm
may prefer tohave another organization integrate the components ofthe system or it may preferto integrate the components itself. The second decision relates to from whom the buyer purchases the components that will ultimately comprise the system. The buyer can purchase all system components from the same vendor or from multiple vendors. In this study, we identifydrivers of
both decisions. Interestingly, we findthatfactors thatare focal to TI markets, such astransferandpresenceof
knowledgetogetherwith technologicalturbulence determinebuyers'preferences.Chapter 4 continues upon Chapter 3 in that itconsiderswhichfactorsdrive abuyer's
preferences forasystemvendor in
Tl
markets. Thefocal argument inthis chapter is that TI markets are embedded and thus that buyers' preferences for a specific systemvendor will
depend upontheembeddedness of thevendorin systems
of
socialrelations, both in time and structure. Of particular interest in this chapter is the finding that the relations ofa system vendorwith
component manufacturerswill
significantly affectabuyer'spreferences towards that vendor. We also show that abuyer's preferences fortheembeddedness ofavendor are dependent upon the knowledgethey possess and thetechnologicalchange they perceive.Underlying chapter 2
through 4 is
the understanding that system components areinterdependent. in Chapter 5 we further
build upon
this insight. More in particular, we examine towhat extent system componentsareindeed interdependent. Thetheoryof
indirect network effects argues that theavailability
of
complements -
in
general referred to assoftware- increases the
utility of
theprimaryproduct -
in general referred toashardware. The objective ofthis chapter is tounderstand the role ofsuch indirect network effects in the
growth of
new hardware.Chapter 1 9
underdevelopedareas
of
research. Also, in dissertation research itconfronts the PhD student with differenttheoretical andmodeling traditions, whichenricheshistraining.Chapter2 contributes to the research stream onbundling, while Chapter5 continues the ongoing researchstream onnetworkeffects. Both Chapters 2 and 5 thusstrictlybelong to thetheoretical domain of hightech systems. Chapters 3 and 4, in contrast, continue the work onorganizational-decision making,related toknowledgeandinformation management, in Tl markets, although theystrongly link this research stream to the researchstream on high tech systems.
Regarding the theoretical backgrounds ofthe different chapters, we note that they specifically build upon economics and economic sociology. Chapter 2 alrnost exclusively builds upon economic
theory, more
in particular microeconomics. Chapter 4 almost exclusively builds uponeconomic sociology, in that itexaminestheeffectof
social networkson buying behavior. Chapters 3 and5 combineinsights fromboth economics andeconomic sociology. Chapter3 combines insights fromeconomic sociological literatureon knowledge
transferandlearningwith insights from institutionaleconomics. Chapter5combines insights from network economics with insights fromeconomic sociology regardingcollective action and diffusion ofaninnovation throughasocial system.
The different chapters are alsodifferent in the methods they employ. Chapter 2 is a conceptual chapter, in which theoretical deduction, logical reasoning, and numerical illustration is used to develop an innovativetheoretical frameworkand specifypropositions. Chapters 3,4 and 5 are empirical studies. In Chapters 3 and 4. we use quasi-experimental data from a conjoint design, gathered through survey research, in a limited dependent variables model. Chapter 5 usesarchival data inanon-linear timeseriesmodel.
Chapter 2
Strategic Bundling of
Products and Prices: A New
Synthesis
for Marketingi
II.1 Abstract
Bundling is pervasive in today's markets. However, the bundling literature reveals inconsistencies in the use
of
terms and ambiguity about basic principles underlying the phenomenon. The literature also lacks an encompassingclassification of
the various strategies, clear rules to evaluate thelegality of
each strategy, and a unifying framework to indicate when each is optimal. Based on areview of
the marketing, economics, and lawliteratures this chapter develops a new synthesis ofthe field
of
bundling, which provides three important benefits. First, it clearly and consistently defines bundling terms andidentifies two
key dimensions that enable a comprehensive classificationof
bundling strategies. Second,itformulates clear rules to evaluate thelegality of eachof
thesestrategies. Third,it
proposesaframeworkof
twelve propositionsthat suggestwhich bundlingstrategy is optimal in various contexts. The synthesis provides managers with a framework with which to understand and choose bundling strategies. lt also provides researchers with promising avenuesforfutureresearch.'
This chapterisbasedupon Stremersch, Stefan and GerardJ.Tellis (2002). "Strategic Bundling of
Products and Prices: A New Synthesis forMarketing," Journal of Marketing, 66(January)
(forthcoming).Rajesh Chandy,HarryCommandeur, Ruud Frambach, PeterGolder,R.Venkatesh,
EdenYin,theJMeditor andfour anonymous./Mreviewersprovidedmanyhelpfulcomments on
Chapter 11 11
II.2 Introduction
Bundling is the sale of two or moreseparate products in apackage. This strategy is pervasive in markets today in one form oranother. In the last decade, bundling has received growing attention in themarketingliterature. However, the publishedstudiesarefuzzyabout
some very basic terms and principles, do not discuss the legality
of
bundling, and do not provide a comprehensive framework on the economic optimalityof
bundling. As a result, marketing researchers may not appreciate thefull
meaningof
bundling and thevariety of
strategies encompassed by the term. Marketing managers may not appreciate the hazards
involved inthis strategy and
fully
exploittheadvantagesof
bundling in various markets. Examplesof
bundles that come to mind readily are opera season tickets (tickets to various events sold as abundle), luggage sets(various luggage items sold as abundle) and Internet service(bundle ofwebaccess,web hosting, email, personalized content, andInternetsearch). Lessstraightforwardexamples includemultimediaPCs,fixedprice menus, executive
MBA programs and premium brokerage accounts. The multimedia PC is a bundle of the traditional PC plus speakers, aCD-ROM,andothermultimediagadgets. A fixed price menu is a bundle ofachoice
of
appetizer, entree, anddessert. Theexecutive MBA isa bundle ofselectedbusinesseducation modules that managers couldotherwiseget separately at various
conferences and educational organizations. A premium brokerage account provides stock trades, stock research, margin trading, retirement planning, and free check
writing in one
account.The above examples show the pen'asiveness and strategic importance
of
bundling. Firms need to resort to bundling cautiously because of the legal pitfalls involved. For example, the landmark antitrust case againstMicrosoft is, at the core, a
case against its bundlingof
Windows andExplorer. Indeed the US Congress has a longhistoryof
legislatingon bundling issues and the US Department
of
Justiceextensively monitors its use by firms. Recentlythe Department has prosecuted substantially morecases. For example, the number o fantitrust cases ithandled between 1996 and 1999 is approximatelydouble the number of cases thatithandled between 1890 and 1996.12 Strategic Bundling of Products and Prices: A New Synthesis for Marketing
Weinberg 1996; Ben-Akiva andGershenfeld 1998; Hanson and Martin 1990; Mulhern and Leone 1991; Venkatesh and Mahajan 1993). Economists have focused mainly on the optimality
of
bundling for monopolists (AdamsandYellen 1976;Burstein 1960; Carbajo, De Meza and Seidman 1990; Long 1984; McAfee,McMillan
and Whinston 1989; Pierce and Winter 1996; Schmalensee 1982 and 1984; Stigler 1963, Whinston 1990), equilibrium theoriesof
bundling (Chen 1997; Kanemoto 1991; Matutes and Regibeau 1992), and thewelfare implications
of
bundling (Dansby and Conrad 1984; Martin 1999; Salinger 1995; Whinston 1990).We identify the
following
three shortcomings in the literature. First, thedomain of
bundling is
ill
defined and terms thatreferto distinct phenomena are used interchangeably. Second, there is noclear, comprehensive, andcoherentdiscussion ofthelegalityof
bundling. Third,there is no integrative frameworkthat explains the optimalityof
bundling conditionalon various factors. On the contrary, the literature contains ambiguity about some key conditionsforoptimality, whiletheoryon othersisincompleteorabsent.
This chapterprovides a new synthesis ofthe field
of
bundling based on a critical review and extension ofthe marketing, economics, and law literatures. In particular, this synthesis makes three importantcontributions to
the literature. First, itclearly and
consistently defines bundling terms and principles. Itidentifies two
key underlying dimensionsof
bundling that enable a comprehensive classificationof
bundling strategies. Second,it
formulates clear rules to evaluatethelegality of each oftheabove strategies. Such rules mustcomplement any discussionof
economic optimalityto ensure that economically optimal strategies are optimal in practice, aftertaking into account legal proscriptions andrisks. Third, it proposesa framework of12propositionsthat prescribe theoptimal bundling strategy in various contexts. The framework is a logical one that uses uniform terms and assumptions. The propositions incorporate all the important factors that influence bundling optimality. These propositions synthesize a body
of
knowledge that is
at least partly supportedby either verballogic,ormathematical proof,orempiricalevidence.Chapter H 13
II.3 A Primer on Bundling Strategies
Thissectionfirstdefines terms usedin bundling. ltthen classifiestheentiredomain of bundlingstrategies andclearlydemarcatestheir legality.
Il.3.1 Definitions
This subsection firstexplains the current confusion in thebundling literature. It then proposes cleardefinitions ofkey terms thatareparsimoniousandrooted in thelaw literature.
II. 3.1.1 Confusion in Literature
The confusion in
the literature arises from inconsistent useof
terms, ambiguous distinctionsbetweenimportantconstructs, andan unclear domainof
application. We explaineach
of
theseproblemswithrelevant examples.First, bundling does nothave consistent, universallyaccepteddefinitions. Adams and Yellen (p. 475,1976)definebundlingas"sellinggoods inpackages." Guiltinan (p. 74,1987) defines bundling as "the practice
of
marketing two or
more products and/or services in a singlepackage foraspecialprice."YadavandMonroe(1993)define it as "the selling of two or more products and/or services at a single price." Without consistent definitions, the legalityof
bundlingbecomes fuzzy and itspracticalimplicationsbecome imprecise.Second, the distinction between a product and a bundle is not clear. For instance, Salinger (1995) treats a pair
of
shoes as abundle of a left and
a right shoe. Telser (1979) considers a car asabundleof
differentparts, such as the engine,wheels, etc. As such, every productwould be
a bundleof
parts, and the
termwould lose
its strategic and legal importance.Third, the domain
of
bundling strategies is not clear. Mulhern and Leone (1991)introduce the concept
of
implicit
price bundling as "the pricingstrategy whereby theprice of
a product isbased on the multitudeof
price effects that arepresentacross products without providing consumers withan explicitjoint
price" (p. 66). By this termtheauthors imply that retailers who decreaseprice in
one category must consider potential sales increases ordecreases in other categories. However, this extension ofthe meaning
of
bundling runs therisk
of
increasing ambiguity about the concept and its domain without enhancing understanding of the core concepts andprinciplesof
bundling.14 Strategic Bundling of Products and Prices: A New Synthesis for Marketing
II. 3.1.2 Bundling
Bundling is the sale of two or
more separate products2 in one package. The term separate hasenormous implications for understanding the legality and optimality of the
phenomenon. So it merits precise definition. We define separate products as products for whichseparate marketsexistbecause at least somebuyers buy or want to buythe products separately. For example, combinedofferings
of
bankingand insurance products arebundles because at least some consumers buy insurance and banking separately.A
travel package including airandgroundtravel is abundle consistingof
procedurallyseparate services. Note that products canbeseparate atonelevel in the channel, whilebeing mereparts ofaproduct atanother level.Whileaprocessor and a hard diskareparts in a PC foranend-user, they areseparate products for aPC manufacturer. Thischapter focuseson bundling fromanend-user
perspective and does not dealwith bundling inachannel context. Il. 3.1.3 Bundling Focus: Product versus Price Bundling
At
present, researchers use the terms product bundling and price bundlinginterchangeably without clearly distinguishingbetween the two strategies. This study is the first inmarketingto
clarify
thisdistinction, articulate theramifications ofeach strategy, and relate the two toeachother.We defineprice bundling as the sale of two ormoreseparateproducts inapackage
at a
discount, without any integration of the products.Because the products are not
integrated, thereservationprice fortheprice bundle is,bydefinition, equal to the sum of the conditional reservationprices ofthe separate products. 3In other words, price bundlingitself
doesnotcreateaddedvalue to consumers and thusadiscount has tobeofferedtomotivate at
least some consumers to buythebundle. Think of a set
of
luggage items,asix-pack of beer, a combo meal,asoftware suite, oraseasonticket for
theopera.We define product bundling as
the integration and sale of two or more separate products or services at any price. Thisintegration generallyprovidesat leastsome consumers withadded value, such ascompactness (integrated stereosystems), seamless interaction (PC systems), non-duplicating coverage (one-stop insurance), reduced risk (mutual fund), inter-connectivity (telecom systems), enhanced performance (personalized dieting and exercise- The term product in thisdefinition and the rest of thetextrefers to both goods and services.
1 Thereservationprice ofaproduct isthemaximumpriceaconsumeriswilling to pay for the
product. The conditional reservation price is the reservation priceofaproduct,conditional on the
Chapter 11 15
program) orconvenience from an integrated
bill
(telecom calling plans). The greater valueraises consumers' reservation
prices for
the product bundlecompared to the sum of the
conditionalreservationprices oftheseparateproducts.
A product bundle can thus be thought of as having an integral architecture
(Ulrich
and Eppinger 1995). It implementsthedifferentfunctions ofthebundledproducts inasingle productbundle.The multimedia PC hasanintegral architecture, in that itintegrates functions such as connection (e.g. modem), data storage and retrieval (e.g. CD-ROM), which were
separatephysicalchunksbeforetheadvent ofthemultimedia PC.
The distinction between price and product bundling is important because it entails different strategic choices forcompanies
with
different consequences. While pricebundling is a pricingandpromotional tool. product bundling is more strategic in thatitcreates addedvalue. Managers can thususeprice bundlingeasily, atshortnotice, and fora short duration, while productbundling is more ofalong-term differentiationstrategy. In the case
of
physicalgoods,product bundlingrequires a new design, research tooptimizethe design,andretooling to manufacture the product bundle. In the case
of
services, product bundling requires re-definitionof
services,optimization of
the interfaces among the services, and redesign ofservice delivery processes. Managers frequently approach product
bundling from a (new)
product development perspective,
involving the R&D
and manufacturingdepartments. Price bundlingdecisionsareoften thesoleprerogative ofthemarketingdepartment.For example, consider the strategic
options of Dell
that markets to consumers whowant to buy
a portable computer system consisting ofabasic laptop, a modem, and a CDburner. First, it can sell theseproducts asseparate items, such thatthe price of each item is independent
of
consumers' purchase ofthe other item. In this case, consumers could easilyforego purchasing a modem or CD
burner or
they could purchase it from a competitor. Second, Dell can selltheproducts asaprice bundle. For example,itcould, without physically changing any ofthe products, give a discount to consumers if they buy all three products together. This offer would probably motivate at least some consumers to buy all three products from Dell. Third, Dell can sellthe three items as aproduct bundle. To meetthe latter classification, Dellmust design someintegration ofthethreeseparateproducts.For example, it couldcreateanenhancedlaptop. Not only couldthistriggersomeconsumers tobuying all products from Dell,but throughthevalue added they might even do so atapremiumprice.II. 3.1.4 Bundling Form: Pure versus Mixed.
16 Strategic Bundling of Products and Prices: A New Synthesis for Marketing
not thebundle. Typically,becausethisstrategy is abasestrategy formostfirms,the strategy
is called unbundling only when
contrasting it with
a bundling strategy. Pure bundling is astrategy in which a
firm
sells onlythe bundle and not (all of) the products separately. Pure bundlingissometimes also calledtying ineconomicsandlegal literature.4A
tying product isa separate product that is bundled
with
other separate products. Tie-ins are secondaryproducts thatarebundled withtheprimaryproduct. Mixed bundling isastrategyin which a firm sells both the bundle and all of the separate products in the bundle separately.Table II. 1
presentsatabular view
of
theseterms.Term
Definition
ExamplesBundling
Bundling is the sale of two or moreseparate Opera season tickets,products in one package. multimedia PC
Price
Bundling Price bundling is the sale of two or moreseparate Luggage sets, variety packproducts asapackage atadiscount, withoutany of cereals
integration ofthe products.
Product
Bundling
Product bundling is the integration and sale of two Multimedia PC, soundor moreseparateproducts at any price. system
Pure
Bundling
Pure bundling is a strategy in which a firmsells IBM's bundling of onlythebundle and not (all of)the products tabulatingmachines andseparately. cards
Mixed Bundling
Mixed bundling is a strategy in which a firmsells Telecom bundles boththebundle and (all of)the productsseparately.
Table Il. 1 Bundling Terms
II.3.2 Classification
andLegality of
Bundling
StrategiesII.3.2.1 Classification of Bundling Strategies
To classifyandrelate various bundlingstrategies, we identify two keydimensions of bundling: 1)thefocus
of
bundling, whetheronpriceorproduct, and 2) the formof
bundling, whether pure or mixed. These dimensions encompass a rich setof
bundling strategies that have substantially differentcharacteristics andimplications. Byusingthese twodimensions, focus and form, TableII.2classifies thedomainof
bundlingstrategies. The focusof
bundling is alongthe horizontal axis, i.e.,oneither priceor product. Theformof
bundlingis along the vertical axis, i.e., none, pure, or mixed. The table considersageneral case with two products4 Many economics scholarswillapproachtyingmorenarrowly, aspurebundling
of
products infixedproportions, e.g.abundle of a car and car insuranceisalwaysthecombinationof one car with one
Chapter II 17
X and Y. Combinations of X and
Y
represent the termsofthe sale. Thus (X,Y)
represents the sale ofa price bundle, (XEBY) represents the sale ofaproduct bundle, and X andY
without parenthesesrepresent the saleof
separateproducts.Focus : Form - _ - _ Price Product
0
Unbtindling x
Ie e
Pure (X.Y) (X W Y) Bundlinge e
Mixed
(X.Y) (X * Y)Bundling x
X Y Y-Table 11.2: A Classijication of Bundling Strategies
When the products are sold separately, the strategy is unbundling, and remains the same for the price and product columns (cell 1). Sears sells Kenmore home appliances unbundled. Cell 2represents a case ofpure pricebundling. In this case, a
firm
bundles the two products foronefixedprice,without integratingtheproducts orofferingthem separately.A
classic example would include restaurants thatoffer only
a fixedprice menu with
appetizer, entrde,anddessert. Cell3 represents a case o
f
mixedpricebundling, inwhich case thefirm
sells theseparateproducts (unchanged) in aprice bundle, and also sells the productsseparately at their regular price.
An
example would be Samsonite's strategyof
selling differentsizesof
suitcases separately as wellascomplete sets atadiscounted price.Cell 4 represents a case ofpure product bundling. 1n this case the
firm
physically integrates the products and sells only this integrated product bundle. An example is Apple Computer's strategyof
sellingits computers andsoftware asone package. Cell 5represents18 Strategic Bundling of Products and Prices: A New Synthesis for Marketing
H.3.2.2 Legality of Bundling
The above classification helps us to sort out the legality
of
various strategies (see Table Il.2). This is valuable in view of the limitedattention devoted to a cleardelineation of the legalrulesonbundling inthemarketing,economics, andlawliteratures. in particular, the marketing literature avoids all discussion ofthe legalityof
bundling although it covers the legalityof
truth inadvertisingor pricediscrimination substantially (alsoseeWerner 1991 and 1993). Yet bundling is pervasive in marketing and as important as price discrimination or advertising. Moreover, understanding the legalityof
bundling is crucial to developing successful priceandproductbundlingstrategies.Legal or economic analysts have not made an effort to abstractclearrules from past
cases, despite a body of caselaw. Based on a review of the law literature and case law, we
synthesizetheproscriptionscontained in the relevant Federal laws in two clearrules, the per se rule and the rule
of
reason. This simple distinctionhelps to explain much ofthe apparentconflict
in court rulings on various cases over the last century. The spirit underlying both rules is that the bundling strategy ofafirm
should not hurt buyers bylimiting
competition. The per se rule is the morestringent of the two rules.We describe the per se rule in terms of four conditions, as follows: Bundling is illegal per se when it concerns (1) pure bundling (2) of separate products (3) by a firm with market power and (4) a substantial amount of commerce is at stake.
We
have already clarified the meaning ofpure bundlingand separateproducts earlier. We hereexplain marketpower and substantiality.Market
power meansthebundling firm
can"forcea consumer to do something that he would not do in a competitive market" with regard to the tying product (Soobert 1995,footnote 87). Although a monopoly is a clear indication
of
dominant market power, acompany's power does not have to be complete over all buyers in the
market(Fortner
Enterprises v. United States Steel Corp 1969).
Substantiality
means thattheamountof
commerce that isatstake should be high. If this amount is not high,thepractice islegal. How high is high? The U.S. SupremeCourt hasnoted that as little as $60,800 would be considered
substantial (United States v. Loew's 1962).This small number means that this condition is easily met in most markets.Note that a firm may trytocircumvent the law by adoptingamixed bundlingstrategy
in which
it prices individual products so high thatconsumers buy only the bundle. In thiscase, mixed bundling is de.facto pure bundling and
will
receive the same legal treatmentChapter 11 19
We describe the rule of reason in terms of six conditions, as fullows: Bundling is illegal linder the rule of reason. when it concerns (1 ) pure bundling (2) qf separate products (3) by a jirm with market power, (4 j involving a substantial amount of commerce, which (5) poses a threat that the bundling jirm acquires additional market power over at least one of the products that is bundled with the tying product. and (6) no plausible constimer benefits of.fset the potential damage to competition. So, under the rule of reason. each of the four conditions mentioned under the per se rule are
still
necessary, but notjointly
sufficient, for bundling to be illegal. While assessingthe legalityof
bundling under the per se rule can be relatively easyandobjective, doingsounder the ruleof
reason isgenerally more difficult due to thesetwoadditional conditions, whichwe explainbelow.Under the rule
of
reason, the mere existenceof
marketpower overthetyingproduct is not sufficientfor bundling tobeillegal.In addition, there should beasubstantialthreat of the bundlingfirm
acquiring additional market power over at least one ofthe products that are bundled with the tying product. For example, Sandoz Pharmaceuticals bundled Clorazil, a drug for schizophrenia, with CPMS, Clorazil Patient Management System, a system thatmonitored the patient on sideeffects of thedrug(Hurwitz 1991). AlthoughSandozpossessed
market power through the patented Clorazil drug, it could not acquire additional market
power in
the market for monitoring systems because thespecific use
to monitor schizophrenics' reaction toClorazil was avery small part ofthetotal market for monitoring systems. Sothisstrategy was notillegal.The sixth condition for bundling to be illegal, under the rule
of
reason, is that it
producesno benefits to buyers thatmay offset the potential damage to competition(Meese 1999). Ifsuchbenefits arepresent. bundling canstill be legal, even though all five previous conditions are met. Typical offsetting benefits are substantial reductions in costs or major increasesinvalue bybundlingtheproducts. Bythislogic,pureproductbundles may be legal,
if
theyprovideaddedvalue and are not merelyaboltingtogetherof
products. For example, inUS v. Jerrold Electronics (1961),
thecourt used this factor to find an otherwise illegal
bundlingstrategylawful.Jerrold Electronics,anearly producer
of
cabletelevisionequipment, soldcommunity televisionantennasonly bundled withaservice contract. The equipment was very sensitive and customers had no expertise inusing it,thuswarrantingabundlingstrategyto assure quality. Recently the DC Circuit (appeals court), in the
context of
the Microsoft20 Strategic Bundling of Products and Prices: A New Synthesis for Marketing
from second-guessing manufacturers' design decisions
(United States v. Microsoft Corp.,
D.C. Circuit, 1998)Both
of
theserules, the per se rule and the ruleof
reason, have been used over thecourse
of
different legal cases and therefore different applications seem todisplay a
"conflicting set
of
rules" (Dansby and Conrad 1984, p. 377). The IBM case(1936) showedrigid legal scrutiny
of
bundling practices, in that the Supreme Court applied the per se rule avant-la-lettre.Thefirst explicit application of the per se rule occurred in 1947(International Salt Co. v. United States). in this case, International Salt Co. leased patented salt dispensing machines onthecondition thatthelessee purchased salt forthemachines fromthe company. Withoutanalyzing evidenceof
substantial anti-competitive effects nor holding into account evidence presented bythecompany that bundlingwas necessarytomaintain quality control, thecourtfound the company perseguilty
of
illegalbundling (Soobert 1995)in 1969, the Supreme Court relaxed the harsh per se rule and moved towards the rule
of
reason(Fortner Enterprises v. United States Steel Corp 1969). While it is uncertain which of these two rules acourt will use in a
specific bundling case, use of the ruleof
reason is growing more common. Therefore, the ruleof
reason is the most suitable benchmark for judging onthelegalityof
variousbundlingstrategies.H.3.2.3 Legality of Various Bundling Strategies
We apply
the previous discussion onlegality to
our classificationof
bundling strategies (see Table II.2). We discuss only the case inwhich, a firm
with market power bundles separate products, because these are necessary conditions forillegality. All
unbundling (cell 1)
and mixed bundling strategies(cells 3 and 5)
are legal. Pure pricebundling (cell 2) is always illegal, both under the per se rule and under the rule
of
reason.Pureproduct bundling (cell 4) is legal under the rule
of
reason, ifthe benefits to consumers offset potential damage tocompetition. Note however, thatit
wouldbeillegal under the per se rule. Bolting products together is also illegal. Merely bolting products together does not constitute genuine integration and thus cannotbebeneficialto consumers.Although we have specified clear rules for legality and applied them to various bundling strategies, ambiguity may still occur in the factual evidence
of
specific cases. A good example is the landmark antitrustcase United States v. Microsoft Corp. lfone takes the position that Microsoft does not possess a monopoly position (asMIT
economist Richard Schmalensee did), or that the bundlingof
Windows and Explorerprovides consumer value (as MicrosoftCEO SteveBallmer did), thenthe bundlingof
ExplorerandWindows islegal.Chapter 11 11
Explorer with Windows by
Microsoft does not
add value forconsumers (as the US
government did), then
Microsoft's bundling of Explorer
and Windows is illegal. Thus the critical issue inthe Microsoft case is the factualassessmentof
marketpowerand consumer benefits.First, does Microsoft possess market power? The presence
of
marketpower is a
difficult fact
to establish objectively. Especially, thedefinition of
the relevantmarket is the subjectof
intenseeconomic and legal debate. In the Microsoftcase, Judge Jackson definedtherelevant market narrowly as "the worldwide licensing of Intel-compatible PC operating systems" (Jackson 2000). In this market, Microsoft Windows indeed has market power because of its dominantmarket share (95%). One canhoweverdebatethis choice. Would it
be more relevant to also include other desktop operating systems such as Apple orLinux? Howaboutnetworkoperating systems, such as UNIX orNovell?
Second, does the bundling
of
Explorerand Windows provide consumers withadded benefits, or were the twosoftware packagesjustbolted together to tempercompetition? As the DC Circuit Courtindicated, it is
verydifficult for
an outsider to second-guess a company's design choices,especially in hightechmarkets. Besides, in this case,what wouldbe the optimal degree
of
integration between the two software packages? Who would determine that?Thus, while our identificationandformulation
of
clear rules reduces the ambiguity incase law, ambiguity still remains in the empirical evidence to which the rules apply. This problem canbeclarifiedby separating out two stages in the legal process: findings of fact and conclusions oflaw. Establishingfindings of fact isanempiricalissue,which couldbeclear in
some casesandhighly controversialin others, such astheMicrosoftcase. However,based on
those findings, the
conclusions of
law should become muchclearer with
the rules we formulated.II.4 Optimality of Bundling Strategies
This section discusses the
optimality of
the various bundling strategies. It explains underwhatfactorswhich strategy becomes dominant. Relative to work done intheliterature this section makes thefollowing
contributions. First, the economics literature has focused12 Strategic Bundling of Products and Prices: A New Synthesis for Marketing
consumers' conditional reservation prices, objectives of the firm, competition, costs and consumers' perception
of
bundles. Second, the literature is ambiguousabout the
heterogeneity
of
reservation prices. In particular, most authors focus entirelyon asymmetryof
reservation prices. However, thedistributionof
reservation pricesinvolvesasymmetry and variation, eachof
whichcan affect theoptimumstrategy.Ourdiscussionclearlyexplains therole of
each. Third, the literature has largely ignored the important distinction between product and price bundling. Product bundling is an important alternate focusfor
bundlingstrategies, especially in hightech markets. We formulate propositionsthatcover the area of priceandproductbundling.
0 f
thepropositionswe advance, most have neverbeendiscussed.Of
thosepreviously discussed,atleastone(proposition 2)haspreviouslybeenimpreciselystated and a few are at leastpartially supported in the literature (propositions 1,8 and 12). We discussand classifyall
of
these propositions in the interestof
completeness. Whereverthe literature contains apartial or
fullproof for
aproposition, we cite
it. Although our propositions arefirmly
grounded in marketing or economic literature, we also develop a simulation that illustrates themechanisms underlying most of ourpropositions. The use
of
simulation to do this kind of sensitivityanalyses is a bituncommon in marketing, although it has been used successfullybefore (Rajendran and Tellis 1994;
Tellis
and Zufryden 1995). Before we go to the propositions,weexplainthesimulationindetail.11.4.1 Simulation
We illustrate the
logic of some of
our propositionswith
numerical examples (see Table 11.3 and II.4). These examples give theoptimal prices forasupplierbased onvarious distributionsof
consumers' reservation prices or costs of the supplier. To generate theseexamples, we developed aprogram that runs onMicrosoft
Excel:
To determinetheoptimal prices, we useasubroutine calledEvolver'
fromPalisade (see www.palisade.com). This is a powerful optimization routine based on an innovative genetic algorithm.It
replaces the subroutine, Solver, inMicrosoft Excel,which does not work well whenaspreadsheet has "if-then-else"statements such asour program does.Figure 1 provides a
flow
chart ofour program. It consists ofthefollowing
fivestepsor components.
A) The user
must first specify the segments, their sizes, and thedistribution of
reservation pricesbysegmentandproduct(Figure l A).Chapter 11 13
C) Based
on those prices, the program contains formulas to determine consumer surplus forthe various offerings (i.e., product and price combinations) (Figure1C).
D) Next an array computes unit sales for each offering (Figure 1 D). This array
contains formulas that incorporate the
following
rules.i)
Salesoccur for a
particular
offering if and only
if
consumer surplus is positive and amaximum among alternatives available. ii)If
consumers are exactly indifferent between buying and notbuying, they buy
aproduct. iii) If
consumers are exactly indifferent between buying a bundle and the separate products in the bundle, they buythebundle.24 Strategic Bundling of Products and Prices. A New Synthesis for Marketing
A:Segment Sizesand Reservations Prices (Input by user)
SegmentSizes Reservation PricesforProduct
Segments Combinations X Y Bundle A User to jill B User tofill 11 B: Prices to
Offer
(Cells to be changed by Excel-Evolver)
PricesforProductCombinations
X Y Bundle
Excel-Evolver to determine
11
C:Consumer Surplus (Calculated by Excel with use offormula)
Consumer Surplus
Segments X Y Bundle
A Formula:
B Reservation Price - Price Offered
11
D: Sales
of Product to
Each Segment(Calculated by Excel with use offormula)
Sales toEachSegment
Segments X Y Bundle
A Formula: If (1) TRUE = bundle sales
If ( 1) FALSE => separate product sales for products with surplus 2
B 0
With (1)=Surplusfrombundlepositive and 2Esurpluses from
separateproducts
11
E: Revenues (Objective Function)
Revenues
Excel-Evolver to maximize product of (Sales, Segment Sizes and Prices
Offered)
Figure II.1: Flow Chart Of Excel Optimization Program for determining optimal pricing of bundles (displayed here to generate a 2-segment, 2-product example)