Slide 1 © Carliss Y. Baldwin and Kim B. Clark, 2004 Selfish Designs: What Computer Designs Need from the Economy and How They Get It Carliss Y. Baldwin Harvard Business School University of British Columbia October 14, 2004
Slide 2 © Carliss Y. Baldwin and Kim B. Clark, 2004 Four Points Designs “need” to become real –They become real by creating the perception of “value” Designs act as a financial force –In the process of becoming real, they can change the structure of an industry A Modular Design Architecture creates Options with Option Value –What is an “ORMDA”? –What does an ORMDA need from the economy? The economy bites back –ORMDAs are dangerous places to make a living
Slide 3 © Carliss Y. Baldwin and Kim B. Clark, 2004 What are designs? Instructions that turn knowledge into things Span all artifacts and human activities –Tangible, intangible –Transacting, contracting, dispute resolution –Government The wealth of an economy inheres in its designs
Slide 4 © Carliss Y. Baldwin and Kim B. Clark, 2004 The Indictment “Selfish” designs want to become real Their tool is human motivation –A user perceives use-value => willingness to make or willingness-to-pay –Designers and producers add up the users’ willingness- to-pay, subtract costs –The result is an asset => financial value Humans move mountains for financial value –Value operates “as a force” in the economy –Designs have captured the value force, and thus today we work to serve their needs
Slide 5 © Carliss Y. Baldwin and Kim B. Clark, 2004 Designs, when reified (made real), help humans to: –Survive –Interact –Create Humans reify and also improve designs –Complete them, make them, transport them –Pay for design evolution — this is a recent development Human-Design Symbiosis “Surprise and delight” Mutualism or Parasitism?
Slide 6 © Carliss Y. Baldwin and Kim B. Clark, 2004 Design Evolution creates “value forces” that can change the structure of an industry Look at the computer industry from
Slide 7 © Carliss Y. Baldwin and Kim B. Clark, 2004 Industry Transformation Andy Grove described a vertical-to-horizontal transition in the computer industry: “Modular Cluster” “Vertical Silos”
Slide 8 © Carliss Y. Baldwin and Kim B. Clark, 2004 Andy’s Movie The Computer Industry in 1980 Top 10 Public Companies in US Computer Industry Area reflects Market Value in Constant US $
Slide 9 © Carliss Y. Baldwin and Kim B. Clark, 2004 Andy’s Movie The Computer Industry in 1995 Top 10 Public Companies in US Computer Industry Area reflects Market Value in Constant US $
Slide 10 © Carliss Y. Baldwin and Kim B. Clark, 2004 Andy’s Movie—the Sequel The Computer Industry in 2002 Top 10 Public Companies in US Computer Industry Area reflects Market Value in Constant US $
Slide 11 © Carliss Y. Baldwin and Kim B. Clark, 2004 Turbulence in the Industry Departures from Top 10: Xerox (~ bankrupt) DEC (bought) Sperry (bought) Unisys (marginal) AMP (bought) Computervision (LBO) Arrivals to Top 10: Microsoft Cisco Oracle Dell ADP First Data Sic Transit Gloria Mundi… Sic Transit
Slide 12 © Carliss Y. Baldwin and Kim B. Clark, 2004 What changed?
Slide 13 © Carliss Y. Baldwin and Kim B. Clark, 2004 Design Architecture Small designs “just get done” by one person or a small team Large designs require architecture –“The design of the design process” –Forward-looking, future oriented –Analogous to physical architectures »Create and constrain” movement and search Major social technology, but not much studied
Slide 14 © Carliss Y. Baldwin and Kim B. Clark, 2004 Short History System/360: first modular computer design architecture ( ) –Proof of concept in hardware and application software –Proof of option value in market response and product line evolution –First ORMDA = “Option-rich Modular Design Architecture” –System software NOT modularizable »Fred Brooks, “The Mythical Man Month”
Slide 15 © Carliss Y. Baldwin and Kim B. Clark, 2004 Short History (continued) Bell and Newell, Computer Structures (1971) –General principles of modular design for hardware –Basis of PDP-11 design—another ORMDA Thompson and Ritchie, Unix and C ( ) –Modular design of operating system software (contra Brooks Law) –Over time, general principles for evolvable software design (Unix philosophy) Mead and Conway, Intro to VLSI Systems (1980) –Principles of modular design for large-scale chips
Slide 16 © Carliss Y. Baldwin and Kim B. Clark, 2004 Short History (continued) IBM PC (1983) –DEC PDP-11 minimalist strategy (exclude and invite) –+ Intel 8088 chip –+ DOS system software –+ IBM manufacturing –+ Lotus –A mass-market ORMDA
Slide 17 © Carliss Y. Baldwin and Kim B. Clark, 2004 ORMDAs and Value Migration in the Computer Industry, Significant Option- Rich Modular Design Architectures IBM System/360 DEC PDP 11; VAX IBM PC Sun 2; 3; Java VM RISC Internet Protocols (end-to-end principle) Unix and C; Linux HTML; XML(?)
Slide 18 © Carliss Y. Baldwin and Kim B. Clark, 2004 This was the puzzle Kim Clark and I began to tackle in 1987 Where was the value shown in the slide coming from? Designs, yes, but what part and why?
Slide 19 © Carliss Y. Baldwin and Kim B. Clark, 2004 A Modular Architecture “frees up” Design Option Value Split options, decentralize decisions, fragment control Evolution
Slide 20 © Carliss Y. Baldwin and Kim B. Clark, 2004 As scientists, we can visualize and measure modularity in design — after the fact DSMs, Design Hierarchies Methods are tedious, non-automated
Mozilla just after becoming open source Linux of similar size Coord. Cost = 30,537,703 Change Cost = 17.35% Coord. Cost = 15,814,993 Change Cost = 6.65% For fun: Comparison of different software systems with DSM tools
Mozilla just after becoming open source Linux of similar size Coord. Cost = 30,537,703 Change Cost = 17.35% Coord. Cost = 15,814,993 Change Cost = 6.65% Different organizations needed for different architectures One Firm, Tight-knit Team, RAD methods Distributed Open Source Development
Mozilla Before Redesign Mozilla After Redesign !!
Slide 24 © Carliss Y. Baldwin and Kim B. Clark, 2004 But modularity is only half the story—options matter, too “Creates” vs. “Frees up” The sad story of auto front-end modules Design options have “technical potential”, denoted Technical potential, , varies by system and by module Modularity in the absence of high option value is an expensive waste of time
Slide 25 © Carliss Y. Baldwin and Kim B. Clark, 2004 Low Medium Zero High Measuring Option Value Successive, improving versions are evidence of option values being realized over time—after the fact Designers see option values before the fact What do they see?
Slide 26 © Carliss Y. Baldwin and Kim B. Clark, 2004 Sources of option value in computer designs Moore’s Law— –Value of seamless, asynchronous upgrading –Applies to chips Amdahl’s Law “Make the frequent case fast”— –Value of ex post optimization –Applies to all complex artificial systems (“Build one and throw it away.”) Wilkes-Alexander-Clark observation “Valid perceptions of desires emerge through use”— –Value of ex post discovery, direct experience, play –Applies to all new artifacts
Slide 27 © Carliss Y. Baldwin and Kim B. Clark, 2004 In conclusion, an analogy… An ORMDA is like …
Slide 28 © Carliss Y. Baldwin and Kim B. Clark, 2004
Slide 29 © Carliss Y. Baldwin and Kim B. Clark, 2004 Where we are in the argument: Designs “need” to become real –They become real by creating the perception of “value” Designs act as a financial force –In the process of becoming real, they can change the structure of an industry A Modular Design Architecture creates Options with Option Value –What is an “ORMDA”? –What do “selfish” ORMDAs need from the economy? The economy bites back –ORMDAs are dangerous places to make a living
Slide 30 © Carliss Y. Baldwin and Kim B. Clark, 2004 Selfish ORMDAs “need” Lots of design searches Institutions of Innovation to –Complete the designs –Produce the artifacts –Transport/Distribute the goods Mechanisms for financing, selection, compensation, reward (an advanced economy…)
Slide 31 © Carliss Y. Baldwin and Kim B. Clark, 2004 Selfish ORMDAs “need” lots of design searches—and promise lots of $$$ Value Landscape of a minor ORMDA— Sun Microsystems Workstation circa 1992
Slide 32 © Carliss Y. Baldwin and Kim B. Clark, 2004 Technical Potential and Cost of Design Search Vary by Module
Slide 33 © Carliss Y. Baldwin and Kim B. Clark, 2004 Thus each module has its own “value profile”
Slide 34 © Carliss Y. Baldwin and Kim B. Clark, 2004 Institutions get built to exploit opportunities like these, which are “created” by the design architecture This is where the economy bites back!
Slide 35 © Carliss Y. Baldwin and Kim B. Clark, 2004 What are institutions? Firms and markets Transactions and contract types Rules and rights (eg, property rights) Stable patterns of behavior involving several actors operating within a consistent framework of ex ante incentives and ex post rewards –Equilibria of linked games with self-confirming beliefs (Aoki and game theorists)
Slide 36 © Carliss Y. Baldwin and Kim B. Clark, 2004 ORMDAs “need” institutions But not just any type will do In DR2, we argue that the “most suitable” institutional forms for an ORMDA are: –A modular cluster of complementary firms and markets with “own your solution” property rights –A community of cooperating user-developers with GPL-type property rights These forms are “good for the designs”
Slide 37 © Carliss Y. Baldwin and Kim B. Clark, 2004 Three ways to frame the institutional analysis Descriptive: What has actually happened in the ORMDAs we know about? Deductive: What do our models predict? Strategic/Normative: Faced with an ORMDA (and access to financial capital), what should “you” do?
Slide 38 © Carliss Y. Baldwin and Kim B. Clark, 2004 Faced with this ORMDA, what would you do? One module or many? In each module you chose, how many design searches? Which modules are most attractive?
Slide 39 © Carliss Y. Baldwin and Kim B. Clark, 2004 Lots of stories They all make sense When you see them play out, the moves are logical and in some cases “inevitable” But our strategic advice for managers and financiers today comes down to: –“plunge in,” –“get lucky,” –“watch out for Microsoft,” and –“get bought by HP”
Slide 40 © Carliss Y. Baldwin and Kim B. Clark, 2004 One story before we close— ORMDAs and New Industries Significant Option- Rich Modular Design Architectures IBM System/360 DEC PDP 11; VAX IBM PC Sun 2; 3; Java VM RISC Internet Protocols (end-to-end principle) Unix and C; Linux HTML; XML(?)
Slide 41 © Carliss Y. Baldwin and Kim B. Clark, 2004 The Bright Side of the ORMDAs
Slide 42 © Carliss Y. Baldwin and Kim B. Clark, 2004 But there was The Dark Side … $ 2.5 trillion appeared then disappeared in the space of four years!
Slide 43 © Carliss Y. Baldwin and Kim B. Clark, 2004 Bubble followed by a Crash A failure, not of the Internet’s design architecture, but of the institutions built on that architecture
Slide 44 © Carliss Y. Baldwin and Kim B. Clark, 2004 A victory for selfish designs? Good for the designs, not for the humans
Slide 45 © Carliss Y. Baldwin and Kim B. Clark, 2004 Other Cautionary Tales IBM System/360 and “plug-compatible” peripherals IBM PC vs. clones Sun Microsystems vs. Apollo Computer Dell Computer vs. Compaq Computer Each illustrates Perils of ORMDAs
Slide 46 © Carliss Y. Baldwin and Kim B. Clark, 2004 Unless we turn the ORMDA stories into science…
Slide 47 © Carliss Y. Baldwin and Kim B. Clark, 2004 Selfish designs will be in charge! Value-seeking design evolution— As we’ve seen it — the good, the bad, and the ugly…
Slide 48 © Carliss Y. Baldwin and Kim B. Clark, 2004 Remember Designs “need” to become real –They do so by creating perceptions of “value” Value is a powerful economic force –Which can change the structure of an industry The most powerful designs are ORMDAs ORMDAs are dangerous (but interesting) places to live Designs, institutions and strategies are still evolving
Slide 49 © Carliss Y. Baldwin and Kim B. Clark, 2004 Thank you!
Slide 50 © Carliss Y. Baldwin and Kim B. Clark, 2004 IBM System/360 The first modular computer design IBM did not understand the option value it had created Did not increase its inhouse product R&D Result: Many engineers left –to join “plug-compatible peripheral” companies San Jose labs —> Silicon Valley “Compelling, surprising, dangerous”
Slide 51 © Carliss Y. Baldwin and Kim B. Clark, —IBM wanted to be the sole source of all of System/360’s Modules
Slide 52 © Carliss Y. Baldwin and Kim B. Clark, — What actually happened: Entry on modules
Slide 53 © Carliss Y. Baldwin and Kim B. Clark, 2004 By 1980, 100s of firms made S/360 “plug-compatible” components
Slide 54 © Carliss Y. Baldwin and Kim B. Clark, 2004 IBM PC, early 1980s IBM was chasing Apple Created an ORMDA— for “fast evolution” Outsourced: –microprocessor to Intel –operating system to Microsoft Kept for itself: –BIOS –Final-stage assembly
Slide 55 © Carliss Y. Baldwin and Kim B. Clark, — IBM provided few PC Modules
Slide 56 © Carliss Y. Baldwin and Kim B. Clark, 2004 But then … Compaq reverse engineered the BIOS Chips and Technologies made “chipsets” Taiwanese clones had cheaper/better manufacturing Intel refused to second-source Microsoft sabotaged OS/2
Slide 57 © Carliss Y. Baldwin and Kim B. Clark, —IBM PC is the standard, but IBM makes no money
Slide 58 © Carliss Y. Baldwin and Kim B. Clark, 2004 Back up to 1980—Apollo Computer did not make the same mistake as IBM PC managers Keeps Design Control
Slide 59 © Carliss Y. Baldwin and Kim B. Clark, 2004 Then Sun came along… And did even less! How?
Slide 60 © Carliss Y. Baldwin and Kim B. Clark, 2004 Then Sun came along… And did even less! Design Architecture for performance Public Standards for outsourcing
Slide 61 © Carliss Y. Baldwin and Kim B. Clark, 2004 Result: ROIC advantage to Sun Sun used its ROIC advantage to drive Apollo out of the market
Slide 62 © Carliss Y. Baldwin and Kim B. Clark, 2004 Compaq vs. Dell Dell did to Compaq what Sun did to Apollo … Dell created an equally good machine, and Used modularity-in-production to reduce its production, logistics and distribution costs and increase ROIC –Negative Net Working Capital –Direct sales, no dealers
Slide 63 © Carliss Y. Baldwin and Kim B. Clark, 2004 Dell vs. Compaq 1997 Dell started cutting prices; Compaq struggled, but in the end had to exit. Like Apollo, they were acquired by HP!