information rules: a strategic guide to the network economy lock-in effect & networks...

Information Rules:

A Strategic Guide to the Network Economy

Lock-In effect &

Networks externality

Seungkyoon Shin

Information Rules Spring 20002

Recognizing Lock-In

• Cost of switching

• Compare– Ford v. GM– Mac v. PC

Information Rules Spring 20003

What’s the Difference?

• Durable investments in complementary assets– Hardware– Software– Netware

• Switching cost and lock-in are ubiquitous in information systems

• Supplier wants to lock-in customer• Customer wants to avoid lock-in• Basic principle: Look ahead and reason back

Information Rules Spring 20004

Examples of lock-in

• Bell Atlantic and AT&T– 5ESS digital switch used proprietary operating

system– Large switching costs to change switches

• Computer Associates– Vender Level Locking– System Level Locking

Information Rules Spring 20005

Small Switching Costs Matter• Look at lock-in costs on a per customer basis• Phone number portability• Email addresses (Mail Forwarding issue)• Hotmail (advertising, portability)

– $400 mil for 9.5 mil subscribers

• ACM, CalTech– Provide forwarding service to approach possible donors

Information Rules Spring 20006

Valuing an Installed Base

• Customer C switches from A to "same position" w/ B– Total switching costs = customer costs + B's costs

• Example– Switching ISPs costs customer $50 new ISP $25

– New ISP make $100 on customer, switch

– New ISP makes $70 on customer, no switch

• In a competitive market, Profit=switching costs

Information Rules Spring 20007

Profits & Switching CostsIn General:

• Profits from a customer =

total switching costs + quality/cost advantages• In commodity market like telephony, profit per

customer = total switching costs per customer• Use of this rule of thumb

– How much to invest to get locked-in base– Evaluate a target acquisition (e.g., Hotmail)– Product and design decisions that affect switching

costs

Information Rules Spring 20008

Classification of Lock-In

• Durable purchases and replacement: declines with time

• Brand-specific training: rises with time• Information and data: rises with time• Specialized suppliers: may rise• Search costs: learn about alternatives• Loyalty programs: rebuild cumulative usage• Contractual commitments: damages

Information Rules Spring 20009

Durable Purchases

• Telephone switches, Mainframe, OS • After-market sales (supplies, maintenance)• Depends on (true) depreciation• Usually fall with time due to depreciation• Watch out for multiple pieces of hardware

– Supplier will want to stagger vintages

– Contract renewal

• Technology lock-in vs. vendor lock-in

Information Rules Spring 200010

Brand-specific Training

• When personnel are trained

• General training/brand specific training

• How much is transferable?

• Software, an obvious example

• Competitors want to lower switching costs– Borland’s Quattro Pro help for Lotus 123 users– MS Word and WordPerfect help

Information Rules Spring 200011

Information & Databases

• Data files– Insist on standard formats– S/W and database– Whether information can be easily ported over

to another system

• Zip - CD - DVD Transition

Information Rules Spring 200012

Specialized Suppliers

• If durable equipment or S/W is highly specialized, it will be hard to find alternatives

• Pentagon: Joint strike fighter project– Structuring competition among suppliers– Boeing, Lockheed Martin, McDonald

• IBM– Dual sourcing– Intel and AMD

Information Rules Spring 200013

Search Costs• Consumers’ Search Cost

– Psychological costs of change– Time and efforts– Risk to customers

• Suppliers’ Search Cost– Promotional cost– Cost of actually closing the deal– Cost of setting up a new account

– Risk to suppliers• Example of Risk: Credit Cards

– $100 million in receivables is worth about $120 million– Market valuation of “loyalty”

Information Rules Spring 200014

Loyalty Programs• Constructed by firm (artificial lock-in)

– Frequent flyer programs– Getting more popular in E-Commerce– Keep track of history sales: consumer information

• Personalized Pricing– Gold status

• Example: Amazon and Barnes and Noble– Amazon Associates Program v. B&N's Affiliates

program

• Add nonlinearity?

Information Rules Spring 200015

Suppliers and partners

• Bilateral, or two-sided lock-in

• Railroad spur lines

• Customized software– Game for the Nintendo 64 platform– S/W for Apple computer

Information Rules Spring 200016

Follow the Lock-in cycle

Brand Selection

SamplingLock-In

Entrenchment

Information Rules:

A Strategic Guide to the Network Economy

Networks and Positive Feedback

Information Rules Spring 200018

Old and New

• Industrial Economy– Populated with oligopolies– Economies of Scale

• Information Economy– Temporary monopolies– Economies of Networks

Information Rules Spring 200019

Important Ideas

• Positive feedback

• Network effects

• Returns to scale– Demand side– Supply side

Information Rules Spring 200020

Positive Feedback

• Strong get stronger, weak get weaker

• Negative feedback: stabilizing

• Makes a market “tippy”

• Examples: VHS v. Beta, Wintel v. Apple

• “Winner take all markets”

Information Rules Spring 200021

Sources of Positive Feedback

• Supply side economies of scale– Declining average cost

– Marginal cost less than average cost

– Example: information goods, Automobile industry

• Demand side economies of scale– Network effects

– In general: fax, email, Web

– In particular: Sony v. Beta, Wintel v. Apple

Information Rules Spring 200022

Network Effects

• Real networks: – Fax machines, compatible modems, email

• Virtual networks– Mac users, CD-ROM driver, Nintendo 64– Computer (both S/W and H/W) buyers are picking a

network, not simply a product. E.g. user group

• Number of users– Metcalfe’s Law: Value of network of size n

proportional to n2

• Importance of expectations

Information Rules Spring 200023

Lock-In and Switching Costs

• Network effects lead to substantial collective switching costs

• Even worse than individual lock-in

• Due to coordination costs

• Example: QWERTY

Information Rules Spring 200024

Don’t Get Carried Away

• Network externalities don’t always apply– ISPs (but watch out for QoS)– PC production

• Likelihood of tipping– See next slide

Information Rules Spring 200025

Likelihood of Tipping

Low ScaleEconomies

High ScaleEconomies

Low DemandFor Variety

Unlikely High

High DemandFor Variety

Low Depends

Information Rules Spring 200026

Chicken & Eggs

• Fax and fax machines

• VCRs and tapes

• Internet browsers and Java

Information Rules Spring 200027

Igniting Positive Feedback

• Evolution– Give up some performance to ensure

compatibility, thus easing consumer adoption

• Revolution– Wipe the slate clean and come up with the best

product possible

Information Rules Spring 200028

Evolution

• Offer a migration path– Failure of CBS

• Examples– Microsoft– Borland v Lotus

• Build new network by links to old one

• Problems: technical and legal

Information Rules Spring 200029

Technical Obstacles

• Compatibility/Performance Trade-off

• Use Creative design

• Think in terms of system (NBC/CBS)

• Converters and bridge technologies– One-way compatibility– Office 97/95– Boland Q-pro/Lotus 1-2-3

Information Rules Spring 200030

Legal Obstacles

• Need IP licensing

• Example: – Sony and Philips CDs– Amazon.com’ banner ad

Information Rules Spring 200031

Revolution

• Users will bear the switching cost when production is so much better than what people are currently using

• Groves’ law: “10X rule”

• But depends on switching costs

• Example: Nintendo vs. Sega

Information Rules Spring 200032

Openness v. Control

• “Open” approach: offering to make the necessary interfaces and specifications available to others

• “Control” approach: keeping your system proprietary

• The goal is to maximize the value of your technology, not control

Information Rules Spring 200033

To maximize the value…

• Your reward = Total value added to industry x your share of industry value

• Value added to industry– Depends on product and– Size of network

• Your share– Depends on how open

Information Rules Spring 200034

Openness

• More cautious strategy than control

• Full openness– Anybody can make the product– Problem: no champion

• Alliance– Only members of alliance can use– Problem: holding alliance together

Information Rules Spring 200035

Control

• Control standard and go it alone

• A strategy for Market leaders: AT&T, MS, and Intel

• If several try this strategy, it may lead to standards wars

Information Rules Spring 200036

Generic Strategies

Control Open

Compatible ControlledMigration

OpenMigration

Incompatible PerformancePlay

Discontinuity

Information Rules Spring 200037

Performance Play

• Introduce new, incompatible technology

• Examples– Palm Pilot– Iomega Zip

• Attractive if– Great technology– Outsider with no installed base

Information Rules Spring 200038

Controlled Migration

• Compatible, but proprietary

• Examples– Windows 98– Pentium chips– Upgrades and update of S/W programs

Information Rules Spring 200039

Open Migration

• Many vendors, compatible technology

• Little switching cost for customers

• Examples– Fax machines– Modems

Information Rules Spring 200040

Discontinuity

• New technology, but incompatible with existing technology

• Supplied by many vendors

• Examples– CD audio– 3 1/2” disks

Information Rules Spring 200041

Lessons on Lock-in

• Switching costs are ubiquitous

• Customers may be vulnerable

• Value your installed base

• Watch for durable purchases

• Be able to identify 7-types of lock-in

Information Rules Spring 200042

Lessons on Network and Positive feedback

• Positive feedback means strong get stronger and weak get weaker

• Consumers value size of network• Works for large networks, against small

ones• Consumer expectations are critical• Fundamental tradeoff: performance and

compatibility

Information Rules Spring 200043

Lessons, continued

• Fundamental tradeoff: openness and control

• Generic strategies– Performance play– Controlled Migration– Open Migration– Discontinuity

• Lessons of history