Lade Inhalt...

Radical Product Innovation in Network Economies

Lessons for M-Commerce

Diplomarbeit 2001 112 Seiten

BWL - Offline-Marketing und Online-Marketing




List of Figures

List of Tables

List of Abbreviations

List of Symbols


1. Introduction
1.1 The Business Challenge
1.2 Scope and Outline of this Thesis
1.3 Establishing an M-Payment System: The Story of AG

2. The Environment: Impediments for Product Introductions
2.1 Chapter Overview
2.2 Radical Innovations and Network Economies
Introducing a Radical Innovation
An Introduction to Networks
Types of Networks
Value of Networks
2.3 ‘New Economics’
Increasing Returns
Commitment and Lock-In
New Rules?
2.4. Market Environment for Payment Systems
Success Factors for Payment Systems
Cashless Payment Processes
Potential Innovators in the Payment Arena
Summary: Payment Systems and the Theories of ‘New Economics’

3. Remedies: Tackling Network Externalities
3.1 Chapter Overview
3.2 The Radical Innovator
The Vertical Silo
The Atomizer
The Independent Team
3.3. Market Success Factors
Open System Strategies versus Proprietary Network Control
Focusing on Niche Markets
Expectations Management
Favorable Resource Allocation to Encourage Adoption
Increasing Returns Revisited: Feedback Systems
3.4 Establishing a Mobile Payment Solution

4. Frameworks for Successful Product Launches
4.1 Chapter Overview
4.2 A Market Based Approach
Control the Substituted Product
Link to an Existing Network
Link to an Existing Product in the Same Strategic Market
Use Market Shaping Power
Ally with Market Shapers
Spread the Product ‘Virally’
Sell the Idea

5. Conclusion
5.1 Proposals for Further Research
5.2 The Introduction of the Paybox System
5.3 Implications


Referenced Web Pages


For their cooperation and support, I would like to thank

- Klaus Felsch from DiamondCluster International, as well as all his colleagues from the Düsseldorf office.
- Eckhard Ortwein, CTO of AG, and all Payboxers at the Raunheim Headquarter.
- as well as my co-students Jördis Harkányi, Sebastian Just, Nina Knolle, Stefanie Leenen and Maren Otto at HHL.

List of Figures

Figure 1: The Hypercube of Innovation (Afuah, Barahm, 1995)

Figure 2: A Simple Star Network (Economides, 1996)

Figure 3: The Information Super Highway (Economides, 1996)

Figure 4: Broadcasts, Networks, and Communities (Reed, 2001)

Figure 5: Value of Networks (Reed, 2001)

Figure 6: Technological Lock-In (Arthur,1988)

Figure 7: Value of Networks with respect to Customer resources

(Reed, 2001)

Figure 8: Credit Card Transactions (Adopted from Economidas, 1995)

Figure 9: A Market Based View on Radical Product Introduction

in Network Economies

Figure 10: Integrating a radical innovation into a network market

List of Tables

Table 1: Value of Additional Telegraph Connections (Krugman, no publication date)

Table 2: Capabilities of potential m-payment entrants (De Lussanet, 2001)

List of Abbreviations

illustration not visible in this excerpt

List of Symbols

illustration not visible in this excerpt


The focus of this thesis is set on radical product introductions in network markets. It appears that these have to overcome significant impediments to reach critical mass in the form of network externalities. A literature overview of economic analysis in this field, including network types, value of networks, path dependencies and lock-ins, as well as a detailed critique of these frameworks, is provided. We then discuss several possible avenues to overcome these challenges, some of which relate to the optimal boundaries of the firm, as well as some which attempt to tackle market externalities. From these, a market based framework will be developed as a guideline for radical product introductions into network economies. We analyze business cases in light of this framework in a selection of historic examples. In addition, as we find that contradictory arguments abound in chapter three, a second approach is proposed tentatively as basis for further research.

Besides it’s theoretic approach this thesis also provides real world examples from different geographic areas, such as the U.S., Europe and Japan. A special emphasis will be put on the market of payment solutions now emerging in Europe, and, in particular, on the German start-up AG.

1. Introduction

1.1 The Business Challenge

Some things appear as if they were never bound to happen. Let us consider a marketing text released for Askey’s Video phone in 2000: “The world will no longer be far apart. With integrate voice transmission and real-time imaging technologies, Askey’s new stand alone video phone allows people to meet in person even when they are far apart.”[1] Somehow, consumers were not as enthusiastic; market adoption of the video phone, despite all its without doubt great features, was dismal. In fact, this was not the first attempt to market a new, innovative communication device like that, as the basic idea had been around since 1964, when AT&T demonstrated a picture phone at the New York World Fair.[2] What went wrong?

The answer appears intuitive: the primary use of the video phone is to call other people and see them on the screen; the product derives it’s value proposition not from inherent features, but from external factors like the number of consumers that already have such a phone installed and can be contacted.[3] It is what economists like Economides[4] call a network good, and its value for consumers is derived from the size of the network it can draw on.

When organizations attempt to introduce these goods into their markets, they face what is colloquially referred to as the ‘chicken-and-egg’ problem: if a sufficiently large base had already adopted the product, others would join in too, ascertaining the success of the product. However, as long as the installed base is small, the value proposition is weak, so purchasers are reluctant to sign up for the offer. This makes it unlikely that the product will ever achieve the large network size in the first place.[5]

With the advent of data applications over wireless phones, decision makers in the telecommunications industry are bound to face this paradox situation, that their predecessors marketing the video phone were obviously unable to solve, again and again. Many of the much-anticipated so called m-Commerce products, like mobile auctioning services, wireless portal sites or m-Payment systems, draw a large part of their usefulness on the willingness of third parties to participate, be that as buyers and sellers on auction sites, content providers and content consumers for portals, or acceptance points and payers for mobile payments.

European telecommunication industry leaders as recently as summer 2000 appeared confident that they would have the abilities to successfully launch these services. Had not the use of the fixed line sibling of wireless data and m-Commerce, the Internet and e-commerce, sky-rocketed over the last number of years? Now, however, enthusiasm appears to have cooled off, and initial experiences show mixed results at best: while NTT DoComo’s I-mode service blossomed in Japan, the European standard WAP so far failed to catch on with consumers.[6]

1.2 Scope and Outline of this Thesis

The aim of this thesis is to analyze the chicken-and-egg problem and it’s effects on new product introductions. Although examples will draw from a wide range of business and non-business contexts, the exemplary focus will lie on m-Commerce applications, and in that area, m-Payments in particular. As one current example of an m-Payment system trying to achieve critical mass, we will consider the German start-up AG[7] in detail.

The analysis will be set out in the following way:

- The first chapter will end with a short description of AG, including its service offering and business model.
- The second part will start out by defining the characteristics of the product and the markets associated with the chicken-and-egg problem. To this effect, we will describe radical product innovations and network goods. We will then critically summarize and analyze a set of inter-related theories attempting to describe the environment which is often referred to as ‘New Economics:’[8] path dependencies, increasing returns, and lock-in effects. As these ideas have seen some heated discussion in the last years,[9] a critical appraisal will follow. We will try to apply these concepts to the market for m-Payment methods at the end of part two, to provide a basis for our discussion of AG.
- In the third chapter, we will address concepts and frameworks that attempt to resolve these issues. Here, the boundaries of the firm for radical innovations, as well as a range of market development strategies will be discussed and critically acclaimed. However, we will not consider the internal capabilities necessary for a successful product launch, nor internal organizations which may prove beneficial, although both are, of course, of critical importance for a company’s success. We will then turn to and describe the firm’s strategy to overcome the network effects.
- We will develop a preliminary framework to categorize different challenges and the associated solutions in chapter four. The main differentiator between various categories will be the market power the innovator holds in the respective target market.
- In the final chapter, a second framework will be suggested as part of further research proposals. Our main impetus here will be to deal with the conflicting managerial implications from the arguments presented in chapter three. In addition, the current strategy of Paybox will be evaluated before some general conclusions will be drawn.[10]

1.3 Establishing an M-Payment System: The Story of AG

“So far, every attempt to create a new currency for the Internet Economy has failed miserably, the digital-cash business has been plagued by a terrible chicken-and-egg problem: Consumers won’t buy into new payment systems that aren’t in common use, and without mass consumer participation, merchants will be indifferent at best."

Megan Barnett

The Industry Standard[11] AG claims to introduce the worlds’ first ‘mass market mobile payment system.’[12] The company was founded in July 1999 and launched its payment system in May 2000 in Germany. In March 2001, the company was awarded the Innovation Prize 2000/2001 at CeBIT, the worlds’ largest IT fair. At that point, the company claimed 4,500 acceptance points and 250,000 consumers in Germany alone and was in the process of launching the system in Austria, Sweden, Spain and the UK, as well as several other countries in Europe and beyond.

The company’s main financial backer is Deutsche Bank AG, which owns a 50% controlling stake of the equity and also handles the actual money transactions in Germany as well as most other countries. The mobile phone reseller Debitel was won as the strategic distribution partner in Germany, and holds an equity stake in the German subsidiary.

The system is based on payment authentication via PIN that is transmitted over the touch-tone technology used in large call center applications (called Interactive Voice Response, or IVR). The consumer value proposition includes enhanced security over credit card payments especially for Internet purchases, as physical access to the mobile phone is required to complete a transaction. The service offerings currently include:

- Internet to Paybox, which allows purchases at Internet shops and contributes about 50% of the transactions made,
- Mobile to Paybox, which subsumes payments to mobile merchants like cab drivers and pizza delivery services, and
- Paybox to Paybox, a peer-to-peer payment service between Paybox users.

In addition, Paybox allows money transfers over it’s web site utilizing the Paybox security features.

As an example on how these services work from a user’s perspective, consider a consumer purchasing a book at an online merchant. Instead of choosing ‘payment per credit card’ at checkout, he decides to ‘pay with a smile,’ as the official company slogan proposes. He is asked to enter his mobile phone number (or a pre-configured alias) on the Internet merchant’s web site. Seconds later, his phone rings and an automatic voice asks the customer to confirm his transaction by entering a PIN on the phones’ keypad. The PIN is checked, and the merchants web site displays a confirmation of payment to the consumer.

The original revenue model was based on a number of important pillars. Similar to a credit card company, all users would pay a service subscription fee, which was set at 5 € for consumers. In addition, merchants would pay a disagio on transactions made over Paybox at a standard rate or 3%. Consumers would pay 0.25 € per 25 € of transaction value on peer-to-peer transfers. In addition, all calls required would use cost-splitting service call numbers common in Germany.

However, in order to drive consumer adoption, the company decided only months after launch to offer the system free of charge for consumers, dropping the first year of subscription fees for consumers, as well as peer-to-peer transaction charges and offering the service on free-call phone numbers.

Transactions are fulfilled via direct debit payments carried out by Deutsche Bank. Paybox is not allowed to offer a legal payment guarantee to merchants, as it is not a bank, but promises that ‘payments can be expected.’

2. The Environment: Impediments for Product Introductions

2.1 Chapter Overview

We will commence our analysis by determining which criteria must be met by the innovation and it’s market for the chicken-and-egg problematic to ensue. To this effect, we will first turn to radical innovations, before discussing the economics of networks. Our argument will begin with a proposition on the nature of the relationship of the two:

the chicken-and-egg problem is caused by an innovation that changes the product architecture from a consumer / complementor perspective in a market segment characterized by strong network effects.

We will define innovation types as well as the characteristics of network goods, and, more specifically, network externalities, in the following. To further analyze the consequences of this problem, we will then turn to the ‘New Economics’ of positive feedbacks, path dependency and lock-ins. We will close the chapter with a critical evaluation of these terms and the concepts they describe.

2.2 Radical Innovations and Network Economies

Introducing a Radical Innovation

Before we define the nature of these networks and the effect associated with them, let us dwell some more on the innovation itself and its’ inherent characteristics. These are usually due to radical shifts in underlying technologies. In general, the radical change hypothesis can be summarized as “abrupt technological change [that] punctuates the environment”[13] and thereby creates “unmistakable challenges for established firms.”[14] These shocks can lead to the emergence of new classes of products (e.g. fax machines, personal computers), to product substitution (e.g. compact discs vs. conventional records) or to fundamental product improvements (e.g. turbojets vs. jets).[15] Processes can also be innovated in a radical way,[16] but these examples will not be part of the discussions in this thesis, as they do not add value to the issues at hand.

To lead the analysis one step further, we introduce the notion of the ‘Hypercube of Innovation’ as discussed by Afuah and Bahram.[17] The framework is based on previous research of Clark and Henderson[18] that differentiates four distinct cases of innovations visualized by a simple two by two matrix.[19] The authors distinguish between a change in the core concept of the product and a reconfiguration of actual components, and thereby create the following classification:

- Incremental Innovation: a minor change on both axis, refining and reinforcing the original design. These changes are common and can be made on a routine basis.
- Modular Innovation: a change in core concepts of the product, but not in the overall architecture, i.e. the configuration of components.
- Architectural Innovation: a change in the overall configuration, but not in the core concepts of the product.
- Radical Innovation: a change along both axis of the matrix.
Afuah’s and Bahram’s hypercube also includes a third dimension, which subsumes the direct external environment of the firm launching the innovation. They identify suppliers, innovators, customers and complementary innovators as critical participants in the new innovations value chain,[20] or, more precisely, value net.[21]

Abbildung in dieser Leseprobe nicht enthalten

Figure 1: The Hypercube of Innovation (Afuah, Barahm, 1995).

A somewhat different approach on classifications of innovations that has been hotly debated is Christensen’s perception of the disruptive technology, as opposed to sustaining technologies.[22] His methodology differs from the above in that he places a strong emphasis of the product’s value proposition for it’s customers. In his mind, a disruptive technology is defined by the facts that:

- It offers a new value proposition to the market.
- It – at first – does not compete with older technologies for the same product / market space as it is unable to offer the required performance.
- As time goes by, sustaining innovations improve the disruptive technology, it becomes acceptable for the majority of the customers of the older technology and – although it may still not excel versus the old technology – subsequently replaces it.[23]

For the basis of our following analysis, we find that only a few of the cases differentiated above can produce the chicken-and-egg problem that is the focus of our study.

First, it appears from the Clark / Henderson framework, that only innovations that change the linkage between core concepts and components – i.e., architectural innovations and radical innovations – actually qualify as being likely to create the chicken-and-egg problem. A modular innovation – for example, the introduction of Short Messaging Service (SMS) in Europe – does not produce the studied result, as the general architecture remains unchanged with an add-on ‘feature’ of SMS service. Incremental innovations, similarly, are usually without consequences.[24]

Adding the third dimension to the framework, the value net, as proposed by Afuah and Barahm, we can stipulate that the chicken-and-egg problem occurs only when this innovation affects customers or complementary innovators. For example, the introduction of several new rechargeable batteries was definitely a radical step for battery producers.[25] Consumers and complementors may have found the innovation a major improvement, but it was still basically an incremental innovation from their point of view, and no chicken-and-egg problematic could ensue.

The question whether an innovation is sustaining or disruptive in Christensen’s view does not allow us to conclude whether a chicken-and-egg problem may arise. However, it will help us in the later analysis of the perspective of the various players. While empirical studies suggest that it is just as likely that industry incumbents introduce radically new products into the market than new entrants[26] the latter appear to hold a clear edge in disruptive technologies. In fact, in his newest writings, Christensen’s definition for a disruptive technology is the ability to create a new business model.[27]

Although the technology shocks and anteceding radical innovations are often portrayed as being dangerous for the incumbent technologies, this is being offset by two effects that work in favor of the status quo: positive network externalities and lock-in effects. What exactly can be inferred from these terms and what their consequences may be will be discussed in detail in this chapter.

An Introduction to Networks

Networks play such an ubiquitous role in modern life that many are not even realized as such. Next to transportation (railroad), information (TV), and communication (phone) networks, one of the most important sets of networks are, for example, languages.[28] Lewin even points out that the social institution ‘the market’ itself is a network.[29] They can be described as being composed of links and nodes, of which multiple are usually required for the provision of any typical network service. Thus, from a microeconomic point of view, many network components are basically complementary to each other.[30] A simple star network, which may, for example, be a telephone or computer network, is depicted in Figure 2.

Abbildung in dieser Leseprobe nicht enthalten

Figure 2: A Simple Star Network (Economides, 1996).

A phone call placed by a customer at N1 to a customer at N2 will be composed of the links N1S as well as SN2, and will make use of the switch at node S. Despite the fact that the goods N1S and SN2 appear very similar and, in fact, may be technically identical, they are not economic substitutes, but complements. However, they may be components of substitute goods, for example N1SN3 and N2SN3.[31]

In the above case, where N1N2 and N2N1 are distinct and different goods, the network is classified as a two way network. When one of the two is unfeasible or does not make economic sense, so that N1N2 and N2N1 are indistinguishable from each other, the network is defined as one-way.[32] A typical example may be the network of TV Broadcasts and TV sets. Note that this classification is not a function of the topological structure of the network. Rather, the interpretation of the network to represent a specific service is important, as, in fact, Figure 2 may represent both the two-way example of a telephone network or the one-way example of the TV broadcaster.[33]

More complex networks may offer similar services through different channels, as can be seen in Figure 3, which is an exemplary network topology for the Information Superhighway.[34] Obviously, a number of close substitutes exist in this topology, as, for example, the final access channel to the consumer may be via cable TV, telephone line, wireless service or satellite link, among others.[35]

Abbildung in dieser Leseprobe nicht enthalten

Figure 3: The Information Super Highway (Economides, 1996).

This is the basic topology that is the foundation of the so called web strategies[36] and the economics of co-opetition,[37] in which all players in a network seek to increase the size of the network overall, as well as, simultaneously, their share of the network market itself.[38]

So far, when the complementary nature of network goods has been discussed, we have implied the notion of compatibility. This can be defined as the possibility that singular goods can be combined without cost to produce bundles of goods that are demanded by customers.[39] Considering the phone call example of Figure 1, the call N1SN2´, which was demanded by the customer, is possible because the links N1S and SN2 as well as the switch S can be combined without further cost. Thus, Economides concludes: “We have pointed out that links on a network are potentially complementary, but it is compatibility that makes complementarity actual.[40]

While some network goods have inherent properties that make them immediately combinable, for far more products, actual complementarity can be achieved only by standardization. Standards are defined by Lewin as “’a shared way of doing things, of interacting.’ Standards serve to co-ordinate individual activity by reducing costs of interacting.”[41] Thus, providers of networks or vertically integrated goods have the option to reduce the cost of interacting by adhering to common standards instead of making their products wholly or partially incompatible.[42]

Economides points out that the question of complementarity of network goods can also “be observed between different classes of goods in non–network industries.”[43] In fact, vertically-related value chains (value nets) are formally equivalent to one-way networks.[44] This feature will play a crucial role in our discussion on the chicken-and-egg problem, as the relationships along the value chain are of significant importance in many problems of this kind.

Fontana notes on the effects of innovations on the relationship of networks: “innovations may alter the existing complemetarity / substitutability relationship among system [network] components.”[45] Thus, an innovation will usually affect one link or node of the network, which may either create a new complementary service to the existing network (i.e., the Fax machine), substitute a link or node (fiber optic cable versus copper cable) or substitute a range of network nodes and links, thus creating competition on an industry level (e.g. the magnetic rail train Transrapid versus the conventional railway system).[46]

As an extended example, consider the PC industry of the 1980s, specifically the competing IBM and Apple designs. The PC industry is comprised of companies offering a variety of vertically integrated and related products, with a multi-tiered software and hardware supplier structure. The industry clearly displays network properties as a one-way network, as Hagel notes: “In the desktop computing business ... highly specialized participants acted both independently and interdependently to assemble a complex package of technology components and services."[47] Nodes of the network structure may include operating software companies (Microsoft), application software developers (Lotus), microprocessor manufacturers (Intel), printer manufacturers (Hewlett-Packard), service and consulting companies (Accenture), and others. Compatibility, which is paramount to enjoy the complementary nature of networks, is guaranteed by a set of standardized software and hardware interfaces to which all players adhere. However, producers may have the option to stand apart from this network. Hagel notes on Apple: “The Lisa and Macintosh were more self contained, with only limited expansion slots and interfaces for peripherals.”[48] Hagel believes that this was a central deficiency of the Apple products, which allowed IBM / Windows based systems to take an early lead.[49]

We will now address some further differentiating characteristics of networks.

Types of Networks

There are several other dimensions by which networks, of any of the types considered above, can differ. As some of these will play a crucial role in the analysis to follow, we will now go into some greater detail concerning those differences here. In all, we find that networks can be described along four dimensions:

1. Virtual versus Real Networks: Shapiro differentiates between real networks that manifest physically, like compatible modems or fax machines, and virtual networks, such as networks of Apple Macintosh or DVD users that share ideas in news rooms and notice boards on the Internet.[50] Quite often, we see that virtual networks build on real networks and further augment their strength. For example, the American Auto Association is a virtual network that strengthens the network effect of the real network of highways and their complements, cars.

Virtual networks are common in the Internet, where they are often referred to as virtual communities. Vitale and Weill[51] report a ‘Townhall’ on the car-mart site[52] “’where smart shoppers talk about cars, trucks and related consumer issues’.” At the book merchant’s[53] community site they find “’a gathering place where you can share your passions for everything’” and they relate the story of a bulletin board at fast food chain’s Taco Bell[54] web site, where fans can exchange opinions on “’stupid Taco Bell managers’.”

2. Horizontal versus Vertical Networks: In horizontal networks – which may be referred to as peer-to-peer or consumer to consumer businesses – the users are all of the same type, although not necessarily of the same value.[55] As an example, consider telephone customers, which are all offered the same basic service – making phone calls – although their average use of the phone may vary greatly.

Vertical networks, or vertically interrelated industries in Economides’ terminology, are linked to a specific value chain often based around a single platform. Different economic goods are combined to create a bundle of goods which is purchased by the customer.[56] Thus, within the network, cascading levels of participants exist which are the suppliers of downstream partners and customers of upstream businesses.

In the computer industry, a host of complementors would usually service customers on the basis of a single standardized architecture. The network’s value depends in a large part on the contribution of different complementors to the final product / service offering. As an example for network externalities in a platform industry, consider the chicken-and-egg problem of a computer operating system like BeOS. Consumers preferred Microsoft Windows over other operating systems because of the rich market of programs supported by the system. Software developers, because of the large installed base of the Windows operating system, would first – and often exclusively – program software for Windows, increasing its lead in consumer preference over competing systems. The effect is so dominant that it was referred to as the ‘applications barrier to entry’ in the Findings of Fact for the antitrust allegations against Microsoft Corporation in November 1999.[57]

Economides notes for this situation: “In a network where complementary as well as substitute links are owned by different firms, the questions of interconnection, compatibility, interoperability, and coordination of quality of services becomes of paramount importance.”[58] However, these questions will not be addressed in great length here.

3. Literal versus Notional Networks: the network may be based on contractual agreements or strategic alliances of some kind, like the telephone network, or on a more informal basis, like a network of University alumni.[59]

4. Proprietary versus Open Networks: some networks are controlled by singular companies, like the network of SAP R/3 installations, which is closely dependant on SAP.[60] Other networks rest on open standards – the fundamental technology for the Internet, the TCP / IP protocol, is an open standard which can be used free of charge. Economides and Flyer define a proprietary network as having “a coalition ... [with] veto power over entry of a new member,”[61] with open networks making up the balance. Proprietary networks tend to be problematic in a regulatory perspective, as these may create a monopolistic situation, as may be the case with Microsoft’s hold on the operating system market. Open networks fuel competition between different, compatible products thriving on the same network, and are not as easy to monopolize.[62]

In addition, two special cases are also of note:

- Intermediary Network: The network connects two distinct groups of customers and thus has the role of an intermediary.

An example may be auction sites, connecting buyers and sellers, or payment systems like Paybox, connecting payees and payers. Other intermediaries may offer dating services or content aggregation (Portals), as well as Compact Discs, where consumers and record labels are connected. From the examples listed, note that the linkage between intermediary networks may be either horizontal (dating service) or vertical (B2B marketplace) in nature. Aside from that, we will define an intermediary network as a combination of two horizontal networks.

- Internalized Network: The choice whether to participate in the network is made not by the actual network users, but by a single decision making institution. This setup is referred to as a welfare maximizing planner by Economides and Himmelberg.[63]

Consider, for example, a large company rolling out a group scheduling tool, like Lotus Notes, to its’ employees. The internalized network is differentiated by the fact that the purchasers of the system are not fragmented. In the case of the scheduling tool, the problem that appears solved may recur, however: the newly created network has to be adopted by the employees in order to actually harness the benefits.

We will, however, not discuss these differences in great length here, and instead focus on the fundamental network property of interest for the chicken-and-egg problem: positive externalities of networks,[64] more commonly known as the Laws of the value of networks.[65]

Value of Networks

Peculiar for networks are positive consumption and production externalities. As Economides relates: “A positive consumption externality (or network externality) signifies the fact that the value of a unit of the good increases with the number of units sold.”[66] This statement is a formalization of the intuitive observation on the lacking success of the video phone in the introduction chapter.

Reeds analysis on the value of networks differentiates broadcasts, networks, and communities (see Figure 4).[67] According to him, a networks’ worth can be assessed as the sum of the value of all potential transactions in the network.[68] In Economides terminology, this corresponds to the sum of the value of all economic bundles of the network.

Abbildung in dieser Leseprobe nicht enthalten

Figure 4: Broadcasts, Networks, and Communities (Reed, 2001).

Reed argues that a “potential transaction can be treated like a call option - its value grows with expected return and with standard deviation of returns.”[69] In addition, transactions have an inherently low downside risk equal to the “strike price” of purchasing that option.[70]

The three network valuation methods will be described with examples from different online sites:[71]

- The web’s first auction system was[72] Similar to real world auction sites like Sotheby’s and Christie’s, the web site did not offer the peer-to-peer functionality of today’s web auctioning services. Instead, it offered a fixed number of products for which a variable number of consumers were allowed to bid. With one central ‘auctioneer’ the system was similar to a broadcast. The structure is in line with Economides’ one-way network designs. Each additional user adds one additional potential transaction. With the number of users (nodes) depicted as N and the value per potential transaction as a, the total value of the onsale network can be considered as aN. This progression is referred to as ‘Sarnoff’s law of broadcasts.’

- A somewhat more advanced offer is’s classified section.[73] Users may act both as sellers and buyers, thus scaling both supply and demand with increased network size. With additional options being created by both suppliers and purchasers joining the system, the valuation of the network has more in common with a two-way network in Economides’ classification, as exemplified by telephone or telegraph networks. The total number of potential transactions, individually valued at b per transaction, is N buyers * N sellers, creating a network value of bN2.[74] This progression equals the widely known ‘Metcalfe’s law of networks.’[75]

- Reed stipulates that the auction system of[76] follows a still different pattern. Instead of pre-configuring set classified categories as in yahoo classifieds, each subset of users can create a separate auction, involving an individual set of transactions. Thus, the total number of possible transactions is one set of two-way transactions (including all members) per member. If the value per transaction is c, the total value of the network can be considered as being c2N. This progression is referred to as ‘Reed’s law of the value of communities.’

Reed admits that it is highly likely that the value per potential transaction is far higher for Sarnoff broadcasts than for Metcalfe networks, which are in turn more valuable than Reed networks, primarily because the probability of exercising the option is higher. Thus, a > b > c. However, due to the higher magnitude of growth in the number of transactions, Reed communities will – in the long term – dominate the other two networks, independent of the differences in valuation of a single potential transaction.

Abbildung in dieser Leseprobe nicht enthalten

Figure 5: Value of Networks (Reed, 2001).

The findings of Reed and Economides translate into what appears to be a system of increasing returns: the number of consumers joining accelerates with the number of consumers already signed up. Armstrong and Hagel agree when they relate: “organizers will need to understand the revenue growth pattern that is distinctive of certain types of ... businesses. This pattern involves a gradual buildup of revenues followed by a sharp acceleration...”[77]

To explore these ideas in a bit more detail, we will now follow on with the concepts of the ‘New Economics.’

2.3 ‘New Economics’

Increasing Returns

Armstrong and Hagel list three basic business characteristics that allow for increasing returns to exist.[78] The first case is often found in software and pharmaceutical industries: the company incurs a large up front investment to design the initial product, be it a new operating system or a medical drug. The cost of producing copies of this product to sell is minimal compared with the up front investment. Average cost per unit are thus bound to fall per unit produced, thus creating an increasing pay off per unit.[79] This effect is referred to as ‘amortization effect’ by Hagel and Singer.[80]

A second commonly cited manifestation is the so called experience curve. Because of incremental innovations in the production processes, the variable cost per unit falls per output, often by a steady percentage per unit produced. This effect, common in machinery manufacturing, also leads to a lower cost base over time.[81]

However, note that both these effects, although they are commonly associated with the ‘New Economics,’ focus on internal – ‘supply side’ – characteristics of the firm, and do not require network effects to manifest.

The economics of networks, however, create a ‘demand side’[82] increasing return to scale, as the attractiveness of the product increases for consumers the greater the network already is. This counter intuitive situation is also described as follows: “once a network is established, scarcity isn't the source of perceived value; instead, ubiquity is. In the physical world, the more fishers who come to a lake, the fewer fish each one will catch; the lower the benefit, and hence the value, for each one. In the cyberworld, on the other hand, the more people who participate in an online network, the greater the benefit – the larger the network, the greater the likelihood that you'll find the person, information, or resource you're seeking.”[83]

The reason for this effect is part of the same analogy:

“Fishers aren't adding fish to the lake; online users are. As each new web site goes online, it adds to the common resource pool not only the information and resources it contains but also the server(s) and bandwidth it brings to the network as a whole. As each new person registers her email address, that's one more person with whom you can communicate quickly and easily. Connections and information are no longer scarce resources in the Internet economy. The one resource that's becoming scarcer and scarcer (and therefore more valuable) is customers' time.“[84]

We have now established that existing networks carry huge benefits for the users involved, and that, in fact, the size of a network is the best guarantee for additional, future growth. By now discussing lock-in and commitment effects, we will try to analyze why existing networks make it so hard for new networks to establish themselves in the same product / market space, regardless of which product offers the superior value proposition, network effects notwithstanding.

Commitment and Lock-In

Commitment is generally defined as a barrier that makes a change in a once chosen path of action impossible, or at least economically undesirable.[85] Its’ general idea, ‘History Matters’ is one that conventional microeconomic theory usually does not address.[86] Indeed, the standard assumptions in microeconomics include fully liquid markets, full tradability of all goods, and minimal information, transaction, and switching cost, measured in time and money.[87] If these assumptions were correct, all market agents would act completely myopically, without regard of past or future developments. Market participants would, in any given situation, realize marginal benefits equal to their marginal cost of production. Obviously, these assumptions do not carry well over into reality.[88] Instead, continuous commitments of resources – time, money – are intrinsic to all economic activity.

Lock-in, as described by Arthur,[89] is a specialized form of these commitments, and may lead to a situation where an economy continues to embrace an inferior technology or standard although superior systems are available. According to him, two effects can be subsumed under the lock-in effect:

- Lock-in Due to Positive Feedback: This idea takes up on the learning curve argument presented by Armstrong and Hagel for increasing returns. Because of incremental learning curve improvements due to its adoption, the standard is already more efficient at delivering value than a competing, newer technology. Participants thus see no reason to switch. Had both technologies been already available at the start, or the same amount of learning already been invested into the second technology, all participants would have agreed that its’ value proposition is actually superior to that of the older technology.[90]

Note that this is a very similar situation to the one already described in Christensen’s ideas on disruptive technology. There, the currently inferior product targets niche markets better to be served by it’s differing value proposition. As soon as the system has become superior because of incremental innovation, it also captures the mass market. In this case, we can thus establish that the critical point for lock-in is, whether head-on competition is required because the market for the technology in question is completely homogenous, or whether niches can be identified and served cost-effectively.

- Lock-in Due to Network Effects: A second argument is that,[91] although the new technology may have better features than the current standard inherently, the latter already benefits from network externalities that cause its total value proposition to be superior to that of the challenger. The nature of these effects, again, have already been discussed previously.

Arthur’s trivial example on technology adoption assumes homogenous agents are choosing between two increasing returns technologies. The payoffs for the technologies are presented in Figure 6. Consider the choices possible for the first agent, a payoff of 10 for technology A or 4 for technology B. He will obviously choose A, as will those coming after him, locking the economy into this technology as an industry standard. Note, however, the additional advantages technology B would have brought all agents in the long run.[92]

Abbildung in dieser Leseprobe nicht enthalten

Figure 6: Technological Lock-In (Adopted from Arthur, 1988).

Nordan stipulated that network goods can be difficult to replace when they satisfy minimum requirements of consumers – meaning, that it is not vital for a product to offer the best technology solution, but rather simply a solution that satisfies customer requirements,[93] an argument to which Christensen and Bower agree.[94] As an example, he refers to the Nintendo Gameboy, and states: “Nintendo’s Gameboy dominates despite ancient technology.” Albeit the fact that the Gameboy employs “archaic” 8-bit technology and low-resolution displays, a number of high technology contenders, like Sega Game Gear, Atari Lynx, and others, failed to break its’ market hold.[95]

When considering the adoption of new technologies despite path dependencies, we can return to the issues of network ownership, as addressed previously. A network owned by a single institution, which we have coined an internalized network, can easily overcome the lock-in effects as presented in Figure 6, as it can centralize the decision making process of it’s agents and internalize any network externalities that exist.[96] For example, even if no fax machines existed in the world, a single large company may find sufficient value in equipping all regional offices with faxes to speed up communication that it may push ahead in establishing the technology as an internal network.

New Rules?

The frameworks discussed so far – Networks, Increasing Returns and Lock-In – of course constitute an interrelated body of economic principles that claim to be ‘New Rules’ for the ‘New Economy.’ Protagonists like Arthur, Hagel and Armstrong assert that their results overturn the established frameworks of Economics. Arthur identifies the following characteristics: “They tended to have similar properties. There was typically more than one long-run equilibrium outcome. The one arrived at was not predictable in advance; it tended to get locked in; it was not necessarily the most efficient; and its ‘selection’ tended to be subject to historical events. If the problem was symmetrical in formulation, the outcome was typically asymmetrical.”[97]

In conclusion, the ‘New Rules’ see industries influenced by strong networking effects that can be detrimental to competition. There appears to be a “natural tendency towards de facto standardization”[98] as positive feedback cycles and lock-in effects favor the product that has some kind of early leader advantage.[99] The industry dynamics have a tendency to “tipping”: after an initial lead due to minor differences in development at the beginning, one product attains prominence due to a snowball effect of network externalities. Dranove and Gandal stipulate on competing technologies: “One system ends up being the market standard with large amounts of compatible products; the other system has a very small market share, if any at all.”[100]

An archetypal example for this problem is the market failure of the Betamax VCR system. Betamax had – network effects notwithstanding – equivalent features to the competing VHS video system. However, by 1981, VHS held a 61% share between the two technologies. Gandal, Kende and Rob note: “When pre-recorded video cassettes became important in the early 1980s, rental stores preferred to carry VHS tapes because of their compatible installed-base advantage.”[101] This development caused the market to tip in the favor of VHS, which has been the undisputed technology standard since 1988.[102] The basic proposition for firms is thus that “an overriding concern … [when] introducing new system technologies is to ensure that the technology is widely adopted, with an eye towards creating a de facto standard.”[103]

These ideas have been presented here in one continuous description to show how much they are intermeshed and interrelated with each other. We will now examine some arguments in more detail, and discuss in what way the statements put forth are debatable and how much the conclusions drawn have to be modified because of this criticism.

First, let us address some of the issues discussed in the theories of the value of networks. Reed himself already acknowledges that the exponential functions of the values of networks do not directly translate into real monetary value. Instead, he argues that the three network types vie for a limited pool of resources of customers - time or money. He assumes that each consumer has a spending capability of a single unit of resources, and that the combined resources of all users in the network are spent on the three network types proportionate to their relative value. As the size of the network increases, the value of the Reed community dominates the other two networks by such a wide margin that the total value they capture actually falls despite an increase in total resources.[104]

Abbildung in dieser Leseprobe nicht enthalten

Figure 7: Value of networks with respect to customer resources (Reed, 2001).

In this framework, Reed already contradicts increasing returns advocates like Arthur and Hagel. Other examples argue against the increasing value of networks entirely, as they assume that different nodes have different values associated with them, and that more valuable nodes are connected first. Consider Krugmans example of a telegraph network. The value of the singular nodes are equal to the number of customers in each city around the node, and the size of the cities is variable. He postulates the “rank-size”– rule, meaning, the size of the n ths city as a proportion of the largest city is 1/ n. Table 1, based on these assumptions, clearly shows the decreasing returns to scale of the network.[105] In fact, studies in wireless phone usage in Europe show declining revenues per user (decreasing demand side returns to scale) although the networks value supposedly grows.[106] The reason is – as Krugman has postulated – that the more valuable customers have been connected first, and a “mix-driven decline”[107] in average revenues per user has set in.

Abbildung in dieser Leseprobe nicht enthalten

Table 1: Value of additional telegraph connections (Krugman, no publication date)

Critics also refute the idea of increasing returns in the supply side of economcis, as, for example, in the case of the software company with minimal reproduction costs of its software. Liebowitz and Margolis illustrate: “Assume … that there is one technical support specialist for each 25,000 users [of] … Windows 95. If the hiring of technical support personnel tended to bid up their wages, this diseconomy alone could overwhelm the decreasing average fixed cost. Suppose … that hiring an additional technical support specialist (for 25,000 additional users) increased the wages of technical support specialists by $ 22 a year, or an hourly wage increase of a penny. This small change in wages would be sufficient to make overall average costs increase, not decrease with output.”[108]

On the idea of technological lock-in, Liebowitz and Margolis also disagree with the theoretical argument that Arthur illustrates with the adoption of payoffs for the hypothetical technologies A and B. For them, Arthur’s model is deficient because it does not include foresight in the calculations of the agents. Agents would chose differently if they knew that demand was large and technology B was likely to become more efficient. Companies may use advertisements and promotions to communicate these facts. Finally, they state: “…the inefficiency that seems inescapable in the table [reproduced here as Figure 6] is a profit opportunity for someone who can figure out the means to move the outcome from A to B and appropriate the difference.”[109] The Betamax case does not convince them of the contrary. As the cassette, as opposed to the larger VHS tape, did not record an entire feature on a single tape, they state: “Apparently, being able to record an entire movie on one tape was more important to consumers than a smaller cassette.”[110]

Economides agrees with these arguments, and especially highlights the importance of the expectations of the agents. He argues: “Thus, the … statement, ‘the value of a unit of a good increases with the number of units sold,’ should be interpreted as ‘the value of a unit of the good increases with the expected number of units sold.’ Thus, the demand slopes downward but shifts upward with increases in the number of units expected to be sold.”[111] The final market outcome depends on the network size expected by users.[112] Shy adds: “The reliance on joint-consumer expectations generates multiple equilibria where in one equilibrium all consumers adopt the new technology, whereas in the other no one adopts it. Both equilibria are ‘rational’ from the consumers’ viewpoint as they reflect the best response to the decisions made by all other consumers in the market.”[113]

Lewin also explains the effects of the ‘New Economics’ in a more conventional framework when he addresses lock-in effects. Referring to Coasian transaction cost economics, he notes that: “absent transaction costs and transaction impeding wealth effects, apparent externalities and other disefficiencies would spontaneously be removed by the market process.”[114] That is to say, the lock-in effect disappears once a certain threshold value of disefficiency is passed: the information and co-ordination costs incurred in coercing all agents to switch cannot be lower than the efficiency gap between the implemented technology and any other technology. Otherwise, the incumbent is displaced, if need be by entrepreneurial action.[115]

Even the market position of Microsoft vis-à-vis prominent competitors, which many authors see as epitomal evidence for network effects, does not convince the doubters. After an empiric study on software markets,[116] Liebowitz and Margolis come to the conclusion: “The evidence indicates that the most important factor when replacing an incumbent firm is to produce a better product.”[117] Especially, they highlight “the amazing track record of Microsoft in terms of producing better products at low prices,”[118] and come to the conclusion: “Any claims that Microsoft harms consumers through monopolistic pricing is wholly at odds with the evidence we have uncovered.”[119]

Last, they state, if the theories on the ‘New Economy’ are correct, there might not be any cars, as cars are useless without gas stations, which are unprofitable without cars. There might not be any fax machines, because these are only of use if they are already common in the marketplace. Liebowitz and Margolis continue: “But something is amiss. We have cars and we have faxes. We found ways out of these traps. People are clever. They anticipate the future, they look for profit opportunities, they advertise, contract, warranty, and make other sorts of commitments.”[120]

We will now turn to the market situation that Paybox faces, and then return to the assertions that Liebowitz and Margolis made and try to highlight some “ways out of these traps.”

2.4. Market Environment for Payment Systems

Success Factors for Payment Systems

We will try to link the market characteristics described in the previous chapters with the actual occurrences in the mobile payment industry, the market space where operates. Mobile payments can be defined as:

The use of the mobile phone as a payment instrument to make purchases in either the real world or an online environment.”[121]

De Lussanet further defines mobile payment subcategories, micro payments (for sums lower than € 10) and macro payments (higher than € 10).[122] The usual expected product characteristics of payment systems include factors like convenience, security, ease of use, and other internal characteristics.[123] The emphasis among these factors may vary with the type of payment the system is designed for: for micro payments, more value is placed on convenience and speed of service characteristics of the payment methods, while macro payments demand higher transaction security.

In addition, however, payment systems are epitomal network goods: their value proposition is influenced by external factors like ubiquity of use and conversion to other money standards (i.e., cash).[124]

The proponents find that the needs of the market are easy to analyze. The success of a new payment system, in Entenmann’s view, is based on two key factors:

- Additional value for consumers over traditional systems,
- Ubiquitous acceptance by all market participants.[125]
On top of that, he identifies six important characteristics:
- The payment system should be open to all participants,
- It should be usable in all situations,
- It should be easy and convenient to use,
- It should meet the security perceptions of consumers,
- Transaction costs must be attractive and
- The system must be usable internationally.[126]

However, the expectations have not been fulfilled so far. Demaray et. al. note: “The paradox that exists is that contrary to the rapid development of technology and the promise of e-commerce, payment systems have failed to keep up. As a result, technological advancement could be slowed until payment systems develop.”[127] They speculate that despite the promises of the Internet – “Convenience, Reach, Richness”[128] – the adoption of emerging payment systems is slower than anticipated. The cited reasons: network externalities and path dependencies.[129] They add: “much of the slow adoption is the failure to change behaviour.”[130] De Lussanet agrees with these general views: “... [retailers] expect 10% of their transactions to be paid for via mobile phone in three years“ she states, “but would prefer three times more than that.”[131] However, she continues: “Despite retailers’ enthusiasm about mobile payment, consumers don’t want it, providers can’t offer it, and technology can’t support it.”[132]

We will now continue by describing a standard cashless payment method, the credit card transaction, in detail. We will then identify the players that may potentially enter the m-Payment market, and their specific strengths and weaknesses.

Cashless Payment Processes

An archetypal credit card transaction between a consumer and a merchant in Europe as depicted in Figure 8 would be processed as follows.[133] The consumer would present his card upon purchase, with which the merchant would create a credit card receipt (technically an IOU). The payer would sign the receipt, thereby confirming the amount and validating the transaction. The merchant would hand the information in with his bank, which identifies the card issuing bank of the consumer (referred to as the member bank), and informs it of the payment. The card issuing bank would collect all incoming receipts and present a monthly listing of payments to the consumer. It would then withdraw the required funds from the consumers bank account. Part of the money is retained by the card issuing bank as a commission, commonly referred to as disagio. The remainder would be transferred to the merchant bank, which would credit the money received to the merchants account.

Abbildung in dieser Leseprobe nicht enthalten

Figure 8: Credit Card Transactions (Adopted from Economidas, 1995).

In addition, the card issuing bank charged an annual card membership fee to the consumer, and the merchant bank charged the merchant for its’ services, also on a subscription basis.

As value propositions, consumers benefit through enhanced convenience, higher security and later payment dates. For the merchant, the card issuing bank guarantees payment, and subsumes any losses due to fraud. This is particularly an issue on the Internet, where neither physical possession of the card nor a signature is required to make a Purchase.

In advertisements and consumer information, credit card companies often highlighted the ubiquity of their service – a direct attempt to communicate the presence of positive network externalities to the consumer. In Germany, Eurocard ran an advertisement with the slogan “Germany’s most numerous Credit Card.”[134] In the States, Visa advertised with “…and they don’t take American Express.”[135]


[1] Askey Computer Corporation (2000), no page numbers.

[2] Carey (1998), p. 5.

[3] Poor image quality and high service costs are also cited. For a full review, see: Noll, Woods (1979), pp. 29 – 36. However, modern phones like the Askey have solved these technical issues.

[4] Economides (1996), pp. 673 – 699.

[5] Economides, Himmelberg (1995), p. 5.

[6] The Nielsen-Norman Group (Eds.) (2000), p. 11.

[7] For company information, see URL

[8] The term is attributed to Brian Arthur and others by Beinhocker (1997), p. 113.

[9] For a review, see: Lewin (2001), pp. 65 – 96.

[10] For more complete information on AG and its attempt to establish the Paybox payment system, the reader is referred to the forthcoming case “ AG: Paying with a Smile,” written in co-operation with Professor Tawfik Jelassi.

[11] Barnett (1999), no page numbers.

[12] Other alternatives for mobile payments will not be discussed here in detail as they have already been the focus of a number of previous diploma students, among them Markus Wolf of the Leipzig Graduate School of Management’s K7 class.

[13] Arnold (2001), p. 1.

[14] Clark, Henderson (1990), p.11.

[15] Anderson, Tushman (1990), p. 629.

[16] Arnold (2001), p. 2.

[17] Afuah, Bahram (1995), p. 53.

[18] Clark, Henderson (1990), p. 12.

[19] Afuah, Allan (1998), p. 18.

[20] Afuah, Bahram (1995), p. 53.

[21] Evans, Wurster (2000), p. 11. For a deeper analysis, see: Tapscott , Lowy, Ticoll (2000).

[22] Christensen (1997), p. 29.

[23] Bower, Christensen (1995), p. 50.

[24] Several examples on how certain innovations affect which parts of the value chain can be found in Afuah (1998), p. 20.

[25] Lankey, McMichael (1999), p. 11.

[26] Chandy, Tellis (2000), p. 32. In fact, incumbents have turned the race to their favor since the 1950’s.

[27] Christensen, Tedlow (2000), p. 42.

[28] Economides (1996), p. 674.

[29] Lewin (2001), p. 69.

[30] Economides (1996), p. 673.

[31] Economides (1996), p. 674.

[32] Economides, White (1994), p. 660.

[33] Economides (1996), p. 676.

[34] Economides (1996), p. 673.

[35] Economides (1996), p. 675.

[36] Hagel (1996), p. 73.

[37] Brandenburger, Nalebuff (1995), p. 57.

[38] Brandenburger, Nalebuff (1996), p. 21.

[39] Economides (1996), p. 677.

[40] Economides (1996), p. 677. Italics included in original source.

[41] Lewin (2001), p. 70.

[42] Economides (1996), p. 677.

[43] Economides (1996), p. 676.

[44] Economides, White, (1994), p. 664.

[45] Fontana (2000), p. 2.

[46] Fontana (2000), p. 2.

[47] Hagel (1996), p. 77.

[48] Hagel (1996), p. 74.

[49] Hagel (1996), p. 74.

[50] Shapiro (1999), p. 1.

[51] Vitale, Weill (2001), p. 204.

[52] See URL

[53] See URL

[54] See URL

[55] Black Dykema, Delhagen, Audito (1999), p. 2.

[56] Economides (1996), p. 677.

[57] United States District Court (1999), paragraph 30.

[58] Economides (1996), p. 678.

[59] Lewin (2001), p. 70.

[60] Marti, Saloner, Spence (2000), p. 1.

[61] Economides, Flyer (1997), p. 2.

[62] Lemley, McGowan (1998), p. 726. The authors probably assume a software product as their archetypal network good. For other networks, like train tracks, a number of important issues may arise, and a whole body of literature is devoted to ensuing competition in these markets.

[63] Economides, Himmelberg (1995), p. 10.

[64] Economides (1996), p. 679.

[65] Reed (2000), no page numbers.

[66] Economides (1996), p. 679.

[67] Reed (2001), no page numbers.

[68] DiamondCluster Professional Education (2001), no page number.

[69] Reed (2001), no page number.

[70] DiamondCluster Professional Education (2001), no page number.

[71] All three examples taken from: DiamondCluster Professional Education (2001).

[72] The service is now found at under URL

[73] See URL

[74] For completeness, it should be noted that the true value should be N2 – N, as customers are unlikely to network themselves.

[75] Rabaey, Jan (2001), p. 2.

[76] See URL

[77] Armstrong, Hagel (1997), p. 42.

[78] Armstrong, Hagel (1997), p. 43.

[79] Shy (2001), p. 5.

[80] Hagel, Singer (1999), p. 113.

[81] Ghemawat (1985), p. 3.

[82] This term is coined by Lewin (2001), p. 69.

[83] The Diamond Team (no publication date), no page number.

[84] The Diamond Team (no publication date), no page number.

[85] Pedell, Burkhard (2000), p. 1.

[86] Ghemawat (1991), p. 14.

[87] Dobson, Maddala, Miller (1995), p. 5.

[88] Ghemawat (1991), p. 30.

[89] Arthur (1990), p. 97.

[90] Arthur (1990), p. 97.

[91] It appears that there is some overlap between these two categories as an older technology is bound to have learning curve and network effects to its advantage.

[92] Arthur (1988), pp. 119-120. Unfortunately, although this appears to be Arthur’s prime example on lock-in, it does not match any of the two effects described above.

[93] Nordan, Godell (2000), p. 2.

[94] Bower, Christensen (1995), p. 46.

[95] Nordan (2000), p. 1.

[96] Lewin (2001), p. 75.

[97] Arthur (1999), p. 114.

[98] Katz, Shapiro (1994), p. 105.

[99] Arthur (1999), p. 115.

[100] Dranove, Gandal (2001), p. 4.

[101] Gandal, Kende, Rob (2000), p. 43.

[102] Gandal, Kende, Rob (2000), p. 43.

[103] Gandal, Kende, Rob (2000), p. 58.

[104] DiamondCluster Professional Education (2001), no page numbers.

[105] Krugman (no publication date), no page numbers.

[106] Bernstein Research (Eds.) (2001), p. 18.

[107] Bernstein Research (Eds.) (2001), p. 16.

[108] Liebowitz, Margolis (1990), pp. 81 – 82.

[109] Liebowitz, Margolis (1995), p. 3. It should be pointed out that this argument does not refute the ideas of the ‘New Economics’ per se – it only stipulates that the effects may be overcome. This may also hold true for a number of later arguments of Liebowitz and Margolis.

[110] Liebowitz, Margolis (1995), p. 4.

[111] Economides (1996), p. 679. Italics included in original source.

[112] Shy (2001), p. 3.

[113] Shy (2001), p. 3.

[114] Lewin (2001), p. 74.

[115] Lewin (2001), p. 75.

[116] Evaluating prices, quality and market share of several software products over time.

[117] Liebowitz, Margolis (1999), p. 34.

[118] Liebowitz, Margolis (1999), p. 34.

[119] Liebowitz, Margolis (1999), p. 35.

[120] Liebowitz, Margolis (1995), p. 5.

[121] De Lussanet (2001), p. 2. Italics included in original source.

[122] De Lussanet (2001), p. 7.

[123] Entenmann (2001), p. 9.

[124] Demaray, Frasure, Nelson, Pacchini, Perkins (2000), p. 456.

[125] Entenmann (2001), pp. 9 – 11.

[126] Entenmann (2001), pp. 11 – 12.

[127] Demaray, Frasure, Nelson, Pacchini, Perkins (2000), p. 456.

[128] Evans, Wurster (2000), p. 70.

[129] Demaray, Frasure, Nelson, Pacchini, Perkins (2000), p. 458.

[130] Demaray, Frasure, Nelson, Pacchini, Perkins (2000), p. 456.

[131] De Lussanet (2001), p. 2. Italics taken from original source.

[132] De Lussanet (2001), p. 6.

[133] The complete description is taken from Economides (1995), pp. 60-63.

[134] Hundt (2001), p. 1.

[135] Leuty (2001), p. 1.


ISBN (eBook)
ISBN (Buch)
960 KB
Institution / Hochschule
Handelshochschule Leipzig gGmbH – Betriebswirtschaft
innovation pfadabhägigkeit netzwerkökonomie externalität m-commerce



Titel: Radical Product Innovation in Network Economies