Presentation on theme: "INFORMATION TECHNOLOGY IN BUSINESS AND SOCIETY SESSION 23 – NETWORK EFFECTS & POSITIVE FEEDBACK SEAN J. TAYLOR."— Presentation transcript:
INFORMATION TECHNOLOGY IN BUSINESS AND SOCIETY SESSION 23 – NETWORK EFFECTS & POSITIVE FEEDBACK SEAN J. TAYLOR
ADMINISTRATIVIA G1: Great job! G1: Submit group feedback forms G2: Posted
NETWORK EFFECTS: LEARNING OBJECTIVES Understand the idea of positive feedback and describe the role it has played in some prior technology industries (railroad, electricity, telephony) Define network effects (demand-side economies of scale) and understand how they lead to positive feedback Describe the difference between supply-side and demand-side economies of scale Understand the typical sources of network effects in information technology industries Be able to recognize these sources for specific technology products or in specific business contexts Understand the trade-offs between performance and compatibility, and between openness and proprietary control of a technology
POSITIVE FEEDBACK: OVERVIEW Historical examples Railroad gauges, AC versus DC power, telephone networks What is positive feedback? when a firm becomes successful, its past and current success make it more likely to succeed in the future …success feeds on itself, the strong get stronger… When does this happen? More customers lower unit cost (supply-side economies of scale) More customers larger network more valuable product (demand- side economies of scale caused by network effects) Possible consequences of positive feedback Dominance of a single firm or technology Dominance of an inferior technology that got an early lead Critical Mass: below the critical mass, few are willing to buy (inertia); beyond the critical mass, the market takes off. Introducing a new product is difficult because of collective switching costs
SOURCES OF POSITIVE FEEDBACK Supply-side economies of scale (Traditional markets) More customers more units produced lower average cost per unit Marginal cost less than average cost Spreading fixed costs across more units Manufacturing efficiencies, learning by doing Demand-side economies of scale (Digital markets) More units consumed higher value per unit The value of the good comes from the network of consumers who use it (at least in part) Most commonly caused by network effects (Microsoft, Playstation, Facebook) Positive relationship between popularity and value Consumer expectations are key!
VIRTUOUS VS. VICIOUS CYCLE Expectations matter! Users want to join the network of winners! number of compatible users value to user virtuous vicious
DOCTOR CRAZIE http://www.youtube.com/watch?v=utHAlHDXKms
NETWORK EFFECTS Network Externalities Positive Negative the less the betterthe more the better Telephone service Computer software User-generated content Highway traffic Internet congestion Radio frequency interference
MARKETS WITH NETWORK EFFECTS A market exhibits network effects (also known as increasing returns to scale in consumption) when the value to a buyer of an extra unit is higher when more units are sold, everything else being equal A node can reach more nodes in a large network Large sales of components of type A induce larger availability of complementary components B 1,..., B n, thereby increasing the value of components of type A
Person-to-person communication feature Telephones, fax machines, email, Instant Messenger Value from trading volume, number of partners eBay, B2B exchanges Value from more nodes in a network BitTorrent, P2P Networks Value from user-generated content Web 2.0, Wikipedia, online communities, … Value from complementary assets Software -> System: Windows, PlayStation Medium -> Device: HD-DVD vs BlueRay player Training -> Complex software: Oracle, WebLogic NETWORK EFFECTS: SOURCES
NETWORK EFFECTS: TIPPING number of users value to each user More units consumed –> higher value per unit Tipping : Success feeds on itself and strong positive feedback can lead to a winner-take-all situation. (eg: Netscape vs. Mosaic, IE vs. Netscape, Wintel vs. Apple, Nintendo vs. Atari) Inferior products that move first may dominate Product introduction is difficult, entry strategy is crucial
THE MODEL Value of a product in a market with network effects is given by: Z t is the size of the network at time t, represents the value without network effects represents value from network effects.
NETWORK MARKETS: HISTORY MATTERS (I) A and B are incompatible but have the same price A is available at time 0. B will be available at time t, but customers do not know its availability until t. A and B have intrinsic values of a and b respectively Network value is c per user for both products Customer arrival rate is 1 per unit time
NETWORK MARKETS: HISTORY MATTERS (II) a b a+ct t 0 Value Time Q: Which product will a new customer at time t adopt? Why?
NETWORK MARKETS: HISTORY MATTERS (III) The superior product, B, is not adopted. For network products, both intrinsic performance and installed base matter. A has an inferior performance, but has an installed-base advantage by time t, with total value a+ct>b. This is precisely why the inefficient QWERTY keyboard hasnt been replaced.
What happens if B is compatible with A? a b b+ct t 0 Value Time Q: Whats the network size of B at time t? Why? a+ct NETWORK MARKETS: COMPATIBILITY MATTERS
AN ECONOMIC MODEL IN A PICTURE (ARTHUR, 1989)
Taking the evolutionary path: Offer migration path (see lock-in strategies) BW Color TV, MS Office upgrades, … (your examples?) Converters Need good design/engineering to minimize disruption Need to overcome possible legal problems (e.g., a new entrant may face patents for existing technologies) Performance (quality) Compatibility (value carried over from an existing network) Evolution: Lower performance, but backward compatibility provides easy migration path Revolution: Offers radically superior performance, but creates the need to build an installed base from scratch STRATEGY: COMPATIBILITY
A firm benefits from generating network effects if it: Is the only supplier of the product (control) Tries to get a very large user base rapidly (openness) However: Adoption is more rapid with open standards Profit margins are much higher with proprietary standards Total value added to industry Your share of industry value Control: Ensure high profit margins, face an uphill task of getting to critical mass Openness: Facilitate rapid adoption, but face difficulty in keeping margins high The place to be? TRADEOFF: OPENNESS VERSUS CONTROL
When no firm has enough power to dominate: With openness, company tries to maximize the network Standards: Allow anyone to join by following guidelines Important to influence standards early Alliances: Give access to allies, charge rest Total value added to industry Your share of industry value Control: Ensure high profit margins, face an uphill task of getting to critical mass Openness: Facilitate rapid adoption, but face difficulty in keeping margins high The place to be? STRATEGY: OPENNESS
Total value added to industry Your share of industry value Control: Ensure high profit margins, face an uphill task of getting to critical mass Openness: Facilitate rapid adoption, but face difficulty in keeping margins high The place to be? STRATEGY: CONTROL Possible only when: Technology is clearly superior Firm has enough power to control standards Standards still have to be sensible
FOUR GENERIC NETWORK STRATEGIES Controlled Migration Discontinuity Open Migration Performance Play Compatibility Performance ControlOpenness Licensing patents, etc. Provide converters etc. efficient manufacturing technological advantage