Hide Notes 

Slide 1: Open Networks: Theoretical and Policy Lessons from the Internet François Bar
Slide 2: Modeling network change • 1. Digital networks – Network configuration independent from physical layout – Control separable from ownership • 2. Hypotheses – New network uses / innovation patterns – New network evolution dynamics • 3. Evidence – Corporate networks – Electronic marketplaces • 4. Implications – Understanding network transformation – Policy lessons
Slide 3: 1. New network possibilities • Digital: bits, ubiquitous logic, and plentiful bandwidth • Programmable networks: network configuration written in software A given network supports multiple communication patterns Network control becomes flexibly separable from network ownership
Slide 4: Four broad media categories live interpersonal delayed mass
Slide 5: Media archetypes live interpersonal delayed telephone mail mass television newspaper
Slide 6: Distinct technical infrastructures live interpersonal delayed telephone Switched network copper wires mail Packet switched trucks, planes, trains mass Broadcast network radio waves television broadcast network print and deliver newspaper
Slide 7: Distinct actors live interpersonal delayed telephone AT&T, RBOCs mail USPS FedEx, UPS mass CBS, ABC, NBC, Fox cable channels, cable networks television Publishers newspaper
Slide 8: Distinct policy regimes live interpersonal delayed telephone Common carriage Price regulation Universal service mail Common carriage Govnt ownership mass License television Free speech no prior restraint newspaper
Slide 9: live interpersonal delayed telephone mail answering machine voice mail conference call mass television newspaper
Slide 10: live interpersonal delayed telephone telegram overnight delivery mail mass mailing mass television newspaper
Slide 11: live interpersonal delayed telephone mail targeted newsletters several editions per day mass television newspaper
Slide 12: live interpersonal delayed telephone mail many channels television taped programs reruns newspaper mass
Slide 13: Internet as convergence live interpersonal delayed telephone ip telephone mail email mass web-cast television listserv newspaper
Slide 14: old distinctions become irrelevant live interpersonal delayed telephone ip telephone CU-seeme web mail email software updates listserv newspaper mass web-cast television
Slide 15: Characteristics of the new space? live interpersonal delayed "Open" means: •Internet as software overlay on existing networks •multiple communication patterns, invented by (lead) users •Interconnection becomes key mass
Slide 16: Three-layer network model Applications Control Hardware
Slide 17: Control is flexibly separable Applications Control Hardware
Slide 18: Applications Control Applications Applications Control Hardware Hardware
Slide 19: SBC Applications Control Hardware Applications AOL Applications Control Hardware
Slide 20: Earthlink Applications Control Hardware Applications Applications Control Hardware
Slide 21: Corp. LAN Applications Control Hardware Applications Applications Control Hardware
Slide 22: Open networks' new possibilities • Multiple diverse physical networks can be interconnected • Multiple communication patterns can be overlaid on a given network • Network owners no longer are the only actors who can do that • Users can control and configure the network they use  New drivers of network evolution
Slide 23: 2. Open networks: theoretical implications • Mechanisms driving network evolution • New ways to support and structure economic activities – Organizations – Markets
Slide 24: Hypotheses • Users –esp. Lead users– increasingly drive network innovation. • Iterative exploration process • Organizations and networks co-evolve • Network architecture matters • Various modes of intermediation yield different modes of innovation creation and diffusion
Slide 25: 3. Evidence • Case studies / User Roundtables Including: Bank of America, HP, General Motors, Levi Strauss, Bechtel, Cisco, Quantum, Peoplesoft, Solectron, Silicon Graphics,… • Internal corporate networks • Inter-organizational networks • Electronic markets
Slide 26: Corporate Networks Evolution
Slide 27: Network evolution cycle: automation 1. automate use network to do old things in new ways organization unchanged, new infrastructure deployed network as tool
Slide 28: Network evolution cycle: experimentation 1. automate 2. experiment without fundamentally changing the network infrastructure in place after the automation phase, experiment with new applications marginal changes in organization and in network network as lab (bricolage)
Slide 29: Network evolution cycle: re-configuration 1. automate limits of the old network have beenexperiment 2. reached organization and network change simultaneously network articulates change 3. re-organize
Slide 30: Network Evolution: start a new cycle 1. automate a new network has been deployed, upon which new tasks can now be automated new network generation 2. experiment 3. re-organize
Slide 31: A succession of cycles 4. automate 6. re-organize 2. experiment 5. experiment 1. automate 3. re-organize
Slide 32: A succession of cycles… made of smaller cycles 4. automate 6. re-organize 2. experiment 5. experiment 1. automate 3. re-organize
Slide 33: Key features of the network evolution cycle 1. automate Iterative Cumulative 2. experiment End-user driven Structured learning Embedded knowledge Path-dependent 3. re-organize
Slide 34: Co-evolution network 1. automate 2. experiment organization 3. re-organize
Slide 35: Co-evolution requires Network flexibility on ti ica ity p l il ap xib automate 1. fle 2. experiment on ti ra y gu ilit fi ib on ex 3. re-organize c fl
Slide 36: Co-evolution requires Network flexibility on ti ica ity p l il deploy ap xib automate a range of applications over a 1. stable network configuration le f 2. experiment n simultaneously - change network configuration to tio a provide new applications ur ity g il fi ib and nx 3. re-organize applications that take c o fle -invent new advantage of new configuration possibilities
Slide 37: Network flexibility on ti ica ity p l il ap xib automate 1. fle application 2. experiment layer on control layer ti ra y gu ilit fi ib 3. re-organize on ex c fl physical layer
Slide 38: Learning: using vs. doing 1. automate learning by using 2. experiment learning by doing 3. re-organize
Slide 39: Learning: who learns? 1. automate learning by using network user 2. experiment network provider 3. re-organize learning by doing
Slide 40: Corporate network generations 19 95 -? automate experiment web portfolio ? re-organize 5 19 85 -9 automate experiment hybrid network re-organize experiment 5 19 75 -8 automate private network re-organize experiment automate public network re-organize
Slide 41: Electronic Commerce
Slide 42: E-commerce promises: Structural change and efficiency • • • • Entry barriers shattered Small is powerful Disintermediation Frictionless economy
Slide 43: The story so far... • Entering is easy, staying is difficult • Large players dominate, consolidation • New intermediaries emerge, old intermediaries adapt • Stickiness provides traction, friction-free is slippery
Slide 44: Mapping E-commerce deliverable transaction & payment marketplace infrastructure
Slide 45: Mapping E-commerce l na o nti e ve rc n Co mme Co conventional deliverable transaction & payment marketplace infrastructure C-good Conventional marketplace
Slide 46: Mapping E-commerce l na d t io de ce ven rce et-ai er n N mm Co mme o Co C conventional electronic deliverable transaction & payment marketplace infrastructure C-good C-good Conventional marketplace Electronic infrastructure
Slide 47: Mapping E-commerce l na d t io de ce irect erce ven rce et-ai er nd I on me mm N mm Cm -co o E C Co conventional electronic deliverable transaction & payment marketplace infrastructure C-good C-good C-good E-payment Conventional marketplace Electronic marketplace Electronic infrastructure
Slide 48: Mapping E-commerce l na d t io rce de ce irect erce ect ven rce et-ai er nd ir mme Do I on me mm N mm Cm o -c o E-C E C Co conventional electronic deliverable transaction & payment marketplace infrastructure C-good C-good C-good E-good E-delivery E-payment Conventional marketplace Electronic marketplace Electronic infrastructure
Slide 49: Mapping E-commerce Creating new Marketplaces conventional electronic C-good C-good C-good E-good E-delivery E-payment marketplace Conventional marketplace Electronic marketplace Electronic infrastructure
Slide 50: Not so new: Network-aided commerce • Enhance the flow of market-related information • Inform price changes and relay them ("more dynamic pricing") • Fast, cost-effective demand and supply matching • Speed up transactions • Reach new customers • Lower sales and marketing costs • Lower procurement costs
Slide 52: Really new: The network is the marketplace The network has become the place where fundamental market processes take place: – discovery – matching – negotiation – transaction – (sometimes) delivery
Slide 53: Network configuration determines Marketplace architecture • Who can play? – Buyers, sellers, third parties – Equal or differential access • By what rules? – Exchange (bid/ask), auctions, catalog, brokerage,... – level or biased ⇒ network control yields market control
Slide 54: Multiple marketplaces • Consolidate existing market relationships and market power (e.g. auto, Dell) • New trading and distribution spaces (e.g. Napster, content syndication) • Aggregate buyer power (e.g. procurement) • Walled gardens (e.g. AOL-TW, broadband cable, I-mode, WAP)
Slide 55: + execution / fulfillment E-marketplace Neutral exchanges Buyers’ agents Existing channels automation Sellers’ agents Content & Community portals - Discovery +
Slide 56: Airline reservation systems automate experiment Orbitz, SABRE (travelocity) re-organize automate experiment Internet direct sales, multiple marketplaces re-organize automate experiment SABRE as open marketplace re-organize automate experiment re-organize SABRE, closed system
Slide 57: 4. What this means • User-driven innovation is now a core driver of network change • Economic organization and network infrastructure co-evolve (true at various levels of analysis: small groups, firm, firm clusters, markets, national economy) • Network architecture matters economic control, innovation creation and diffusion, cross-learning, intermediation.
Slide 58: Policy implications • A generic problem: – Microsoft Windows as "control layer" – Broadband internet access – Interactive TV – Wireless • Allocation of network control becomes a key policy issue • Open networks foster broad-based innovation