Outline Network Externalities (망 외부성) Diffusion of network goods & services New Opportunities & Strategies What happens …

I. Network Externalities

1. Externalities in economics Externality exists when the production or consumption of a good directly affects businesses or consumers not involved in buying and selling it and when those spillover effects are not fully reflected in market prices. 재화를 사고파는 당사자 이외의 사람이나 비지니스에 직접 영향 그러한 파급효과가 시장 가격에 모두 반영되지 않는 현상

2. Typology of Externalities Hare/Software Paradigm (+) Development Effect Network Externalities Demand Supply Side Side Congestion (-) Pollution

3. Negative Externalities Negative externality is a cost imposed on a person by the actions of others, without the consent of the person who bears the cost (ex. air pollution in industrial areas). Negative externality is a cost imposed on a person by the actions of others, without the consent of the person who bears the cost (ex. air pollution in industrial areas). Usually, the supply curve includes the entire opportunity cost of production, so the optimal amount of output is determined by the intersection of the supply and demand curves. When a negative externality exists, there is an additional cost that must be added to the costs in the supply curve to find the optimal level of production. Including the cost of the externality, the optimal level of output is Q*. An optimal tax is one designed to make the producers in the market take account of the cost of the externality, so the optimal tax equals the cost of the externality.

4. Positive Externalities A positive externality is a benefit received by a person as a result of the actions of others. (a problem arises here as well) If the activities of some people produce benefits for others that the others do not have to pay for, then the true value of the activity will not be reflected in the market demand for the activity, and too little of the activity will take place. A positive externality produces benefits not captured in the demand curve for the good. When those benefits are added in, the optimal amount of output of the good is shown to be Q* rather than the amount Q that would be produced in a competitive market. A problem arises when too little of the activity that generates the positive externality is undertaken. One solution has been subsidies, which in effect represent negative taxes that correct for positive externalities. Rather than charging an individual a certain amount, the government pays the individual generating the externality a subsidy equal to the external benefit. However, as with taxation for negative externalities, the government can have trouble in calculating the appropriate subsidy. The market can work less than perfectly in allocating too few resources to a market that exhibits positive externalities.

5. Network Externalities The Utility that a subscriber drives from a communications service increases as others join the system (Rohlfs, 1974) The increasing utility that a user derives from consumption of a product as the number of other users who consume the same product increases (Katz & Shapiro, 1985) The benefit (i.e., network externalities) that a consumer derives from the use of a good often is an increasing function of the number of other consumers purchasing compatible items. In the case, there is a direct externality: the more subscribers there are on a given communications network, the greater are the services provided by that network (Katz & Shapiro, 1986)  Network externalities can influence consumers’ decision whether to adopt a new technology and producers’ decision whether to standardize their products.

6. Metcalfe’s law If there are n people in a market, and the value of the network to each of them is proportional to the number of other users, then the total value of the network (to all the users) is proportional to n x (n-1) = n2 – n. If # of users: 10 value: 102 – 10 = about 100 If # of users = 100 value of the network: 1002 – 100 = 10000 Increase in the # of user = 10x Increase in the value of network = 100x

7. Demand side economies of scale Network effect = network externalities = demand-side economies of scale The value of connecting to a network depends on the number of other people already connected to it. Key factor: positive feedback

8. Indirect Externalities: HW-SW pdm A system can be any combination of a durable good and associated goods or services that perform some desired function Virtual network: owners of compatible hardware and software systems constitute a virtual network.. Mechanism: Adoption of hardware by one purchaser confers external benefits on other users of the same software. Suppliers may therefore take advantage of scale economies and provide more varieties of software. The availability of more diverse and inexpensive software enhances the value of the existing users’ hardware. Expansion of the concept of network A system can be any combination of a durable good and associated goods or services that perform some desired function Virtual network: owners of compatible hardware and software systems constitute a virtual network. Indirect externalities can arise in these markets only when, as is typical, the components are purchased at different times. Mechanism: Adoption of hardware by one purchaser confers external benefits on other users of the same software. Suppliers may therefore take advantage of scale economies and provide more varieties of software. The availability of more diverse and inexpensive software enhances the value of the existing users’ hardware. Indirect benefits stemming from strong complementarities are said to cause positive feedback in much the same way as direct benefits of expand communication in physical networks.

9. Dark Side of Network Externalities Critical mass Multiple equilibria (next time…) Monopoly Lock-in into inferior good (Path dependency)

a. Critical Mass Critical mass is the milestone which leads to the promising future (virtuous cycle) The value of network depends on the number of subscribers. It means the importance of the size of installed base. What if the installed base was not large enough from the perspective of positive feedback?  getting shrink

b. Multiple Equilibria No one knows the future: equilibrium may not exist, or multiple equilibria may exist and the fundamental theorems of welfare economics may not apply. Networks may not reach optimal size, because purchasers do not take account of social benefits of their purchases, and markets may converge on an inferior standard that effectively excludes better products. No one knows the future: equilibrium may not exist, or multiple equilibria may exist and the fundamental theorems of welfare economics may not apply. The so-called “fundamental theorems of welfare economics” state that, under certain conditions, every competitive equilibrium is a Pareto optimum, and conversely, every Pareto optimum is a competitive equilibrium. Networks may not reach optimal size, because purchasers do not take account of social benefits of their purchases, and markets may converge on an inferior standard that effectively excludes better products.

b-1. Rohlfs model: Multiple equilibria For simplest model, let potential network value (V) of a network distributes uniformly from 0 to 1. Assume: (1) that only those who estimate value to them to be more than V subscribe and (2) that total number of population is 1. Then, subscription ratio r = (1- V) / 1. So V = 1 - r. ------------------- (1) On the other hand, a marginal subscriber expects that his benefit is equal to his subscription cost. Usually subscription cost is set as a constant (P). And the benefit which the marginal subscriber expects is calculated as (r x V). So P = r x V ------------------- (2)

b-2. Interpretation of Rohlf’s model: At equilibrium A, a small network is established, where a limited number of customers can enjoy benefits. At point B, however, the subscription ratio is higher than A, which qualifies point B as being the best equilibrium at the same subscription cost, P. Point A implies the existence of “critical mass”. One example in real world is introduction of videotex: French Minitel – which has already exceeded “critical mass” All other countries’ videotex networks, which don’t exceed critical mass point.  without passing critical mass, socially optimal size of network would not established. Same implication was given from the word, externalities.  market failure (I)

c. Monopoly If critical mass was passed, network externalities lead to monopoly though positive feedback. Ex) VCR War: VHS vs. Beta max QWERTY vs. Dvorak Windows vs. Mac Key problem: anti-trust implication monopoly in technology today leverages the market share in technology tomorrow. Ex) Microsoft operating system – browser market. MS browser – e-business tomorrow

d. Lock-in into inferior good With the advantage of huge installed base and first mover, positive feedback can make markets converge on an inferior good tomorrow. Ex) QWERTY Key board display VHS Dos  Windows Path Dependent 우주왕복선 추진 로켓의 폭: 4피트 8.5인치 기차 선로  영국의 석탄 운반용 마차 선로  말 두마리가 끄는 전차 폭에 맞춰진 로마 도로 폭

II. Competition Strategies

1. Key point to compete Expansion of installed base Business in the presence of network externalities is a war to expand installed base 망 외부성이 있는 재화를 거래하는 시장에서 가장 중요한 것은 가입자 기반의 확대

2. Strategic choices Compatibility vs. Incompatibility Openness vs. Control 호환성 허용 여부 개방 vs 폐쇄

3. Strategic Tools Give away: browser war (first Netscape, next MS IE) Ignorance of piracy (illegal of DOS) Alliance Subsidy Manipulation of expectation (Tactical tools: Pre-announcement & Forming reputation) If consumers expect your product to become popular, a bandwagon will form, the virtuous cycle will begin, and consumers’ expectation will prove correct – vice versa. Success and failure are driven as much by consumer expectations and luck as by the underlying value of the product.  marketing strategy designed to influence consumer expectations is critical in network market.

e. Example of Expectation-Work 6/91

III. 스마트 폰 시대의 경쟁 전략

1. Overview: 모바일 시대 시공간 제약의 극복 이동성  Battery (HP Slate, MS Currier 등) OS & H/W (CPU: Intel  ARM) 새로운 경쟁 환경이 조성되며 네트워크 재화가 본격적으로 경쟁하는 국면으로 진입 중 망외부성 이론을 적용하며 산업 전반을 관전할 수 있는 좋은 기회 28

2. 경쟁의 축 이동 SNS Wintel Google Apple Open Managed Opne Web App Facebook Twitter Wintel Google Apple Open Managed Opne Web App Search App Pages/ Graph Ad Sense iAd Prompted line 생태계 구성원의 확대 Android OS X Google TV iTV I.O. Conference WWDC f8 (Facebook) 29

3. Key points to watch Open vs. Closed (or Managed Open) Compatibility choice (w/ Flash) Give away (MS 전략  Google Android) Path dependency (Google line up vs iTune …) Manipulation of expectation (Thought on Flash) Hardware – Software Paradigm: 개발자 회의 Pre-announcement 30

4. Net Ext. & API What is API? API Application Programming Interface 응용 프로그램 프로그래밍 인터페이스 응용 프로그램에서 사용할 수 있도록 운영 체제나 프로그래밍 언어가 제공하는 기능을 제어할 수 있도록 만든 인터페이스

4. Net Ext. & API Why so fast growth?

4. Net Ext. & API Why so fast growth? Facebook Prepares To Announce 500 Million Users

4. Net Ext. & API Universe of Facebook Open Graph 50,000 Web Sites A Week

4. Net Ext. & API Universe of Twitter

4. Net Ext. & API Open vs. Control, again Sharing of Installed-base: How about Cyworld? Open vs. Control, again Sharing of Installed-base: Open or Collaboration?

5. Expectation Works? Screen shot of iPhone

여섯 번째 이유가 사실 제일 중요합니다. 써드파티의 소프트웨어 레이어가 플랫폼과 개발자 사이에서 어떤 결과를 가져오는지에 대해, 우리는 뼈아픈 교훈을 알고 있습니다. 써드파티 소프트웨어 레이어는 플랫폼의 개선과 발전을 막고 표준에 못미치는 앱을 낳게 됩니다. 만약 개발자들이 써드파티 개발 라이브러리와 툴에 의존하게 되면, 플랫폼에 개선이 생길 경우, 써드파티가 그 신기능을 받아들여야만이 개발자들도 그 기능을 활용할 수 있습니다. 그렇다면 우리는 앞으로 우리의 개선사항을 개발자들에게 제공하기 위해서, 써드파티의 결정에 자비를 구해야 합니다. 만약 그 써드파티가 크로스 플랫폼 개발툴을 제공한다면 상황은 더 나빠집니다. 지원하는 다른 플랫폼 모두에 신기능이 들어갈 때까지 한 플랫폼의 기능개선을 채택하지 않을 것이기 때문이죠. 따라서 최소공배수에 속하는 기능만 개발자들이 접근 가능해집니다. 다시 말씀드려서, 우리 경쟁 플랫폼에 아직 없다고 해서, 우리의 혁신과 개선을 개발자들이 사용하지 못하는 결과가 나온다는 것입니다. 그런 상황을 우리는 받아들일 수 없습니다. 플래시는 크로스 플랫폼 개발툴입니다. 즉, 최고의 아이폰과 아이포드, 아이패드 앱이 어도비의 목표가 아닙니다. 어도비의 목표는 크로스 플랫폼 앱의 작성을 돕는 것입니다. 그래서 어도비는 애플의 개선사항을 정말 느리게 채택해왔습니다. 가령 맥오에스텐은 이제 출하한지 거의 10년이 되어 갑니다만, 어도비는 이제서야 완전히 Cocoa를 채택했습니다. 2주일 전에 나온 CS5에서 말이죠. 맥오에스텐을 완전히 채택한 주요 개발사로 보면 어도비가 마지막입니다. 한글 번역된 글 읽기: http://imackorea.com/?p=426

5. Expectation Works? Adobe’s action! 우리가 사랑하지 않는 것은 무엇을 만들지, 그것을 어떻게 만들지, 나아가 웹에서 무엇을 경험할 지를 선택하는 자유를 앗아가는 바로 그 누군가 입니다.

5. Expectation Works? Apple’s response

5. Expectation Works? Does it work? Does it work?

6. 변화 전망 Platform의 병존 (과거의 반복) 모바일광고의 dominance & 다양한 BM 병존 Google’s Mass vs. Apple’s High end market Webtropolis vs. Suburban 모바일광고의 dominance & 다양한 BM 병존 Google’s Search-based LBS vs. App vs. SNS Local Portal’s Shrinkage New device comes with new BMs Apple’s iTunes, App Store, iBooks Google’s TV & iTV Managed Opne 전략, (당분간) 유효 Apple’s Success w/ Tri-vergence Failure of Linux 42

6. 변화 전망 New eco-system의 부상과 삼성의 고전 SNS의 부상과 Old media strike back Value Chain: C – P – N – T 상의 변화 P와 C가 없는 삼성의 고전이 상당기간 지속될 것 SNS의 부상과 Old media strike back Twitter의 트래픽의 증가와 시청률의 증가 미래 예측 방식의 변화 (전화 여론조사 vs. Google) Web 시대의 도래와 cloud computing 본격화 Native App의 퇴조와 Web 기반 App 유통 본격화 Sync를 통한 cloud computing 서비스 현실화 43