Technology & Strategy Technology and Strategy

Transcription

1 Technology and Strategy Modularity, open interfaces, open standards and open source Michael A M Davies 8 April 2009

2 Strategy and technology Demand opportunity segments behaviour change diffusion and adoption chasm Technological infrastructure architecture parameters envelope and trade-offs innovation trajectory Co-evolution & transitions episode, era dominant design radical vs incremental architectural vs modular disruptive Business ecosystems, value creation and value capture niche lead/follow co-opetition inimitability focus on locus of value standards modularity Decision making common goals, balanced power focus on timely, objective facts embrace uncertainty several options 8 April 2009, Page 2

3 Two simultaneous challenges: create value, at the same time ensure that you can capture value Customers Demand Opportunity Business Ecosystem Applications or Outcomes Markets Offer Partners Technology Partners Operations Marketing Innovation Management Value creation Value capture Technological Infrastructure Turning Points Step changes y Trajectory shifts (disruption) Architectural Shifts t t x Time 8 April 2009, Page 3

4 Some terminology for timelines and transitions Epoch applies to the Era beginning of a new period Epoch marked by radical changes and new Episode developments - while era Event applies to the entire period Era Epoch Episode Episode Event Time Roget's New Millennium Thesaurus, First Edition (v 1.3.1) March 2009, Page 4

5 Era noun 1. a period of time marked by distinctive character, events and so on 1 2. the period of time to which anything belongs or is to be assigned 1 3. a period of time as reckoned from a specific date serving as the basis of its chronological system 2 1: Random House Unabridged Dictionary, Random House Inc : American Heritage Dictionary, 2000 Houghton Mifflin 26 March 2009, Page 5

6 Epoch noun 1. a period of time marked by distinctive character, events and so on 1 2. the beginning of a distinctive period in the history of anything 1 3. a point of time distinguished by a particular event or state of affairs 1 4. a notable event that marks the beginning of a period of history, especially one considered remarkable or noteworthy 2 1: Random House Unabridged Dictionary, Random House Inc : American Heritage Dictionary, 2000 Houghton Mifflin 26 March 2009, Page 6

7 Episode noun 1. a portion of a narrative that relates an event or a series of connected events and forms a coherent story in and of itself 1 2. an incident in the course of a series of events 2 1: American Heritage Dictionary, 2000 Houghton Mifflin 2: Random House Unabridged Dictionary, Random House Inc March 2009, Page 7

8 There are distinct stages in the battle for technological dominance Beginning of a new technological field, with an organization pioneering applied R&D First working prototype emerges First launching of a commercial product Clear early front runner appears One of the early designs becomes dominant Fernando Suarez, Battles for technological dominance: an integrative framework, Research Policy, Volume 33, 2004, pages March 2009, Page 8

9 These phenomena often result in characteristic {product industry} life-cycle (YMMV) Demand Opportunity Early ferment Lead users, early adopters - high payoff, low switching costs Dominant design emerges Early mainstream - usability, cost more important Incremental innovation Mainstream customers - soft factors, aesthetics Maturity Saturation, segmentation, customization Eclipse or renewal Find new needs or new customers Business Ecosystem Many entrants - diverse business models Decisive battles for leadership Intensifying competition, early consolidation Fierce competition, consolidation around majors and minors Technological Infrastructure Make it work - innovate on performance, diverse integrative designs Select optimal architecture, drive down costs, focus on ease of use Provide broader offer, rationalize portfolio, build complementary assets Develop broad portfolio, build platforms Search for new options 11 March 2009, Page 9

10 Reconciling these two models Demand Opportunity Early ferment Lead users, early adopters - high payoff, low switching costs Dominant design emerges Early mainstream - usability, cost more important Incremental innovation Mainstream customers - soft factors, aesthetics Maturity Saturation, segmentation, customization Eclipse or renewal Find new needs or new customers Business Ecosystem Many entrants - diverse business models Decisive battles for leadership Intensifying competition, early consolidation Fierce competition, consolidation around majors and minors Technological Infrastructure Make it work - innovate on performance, diverse integrative designs Select optimal architecture, drive down costs, focus on ease of use Provide broader offer, rationalize portfolio, build complementary assets Develop broad portfolio, build platforms Search for new options 11 March 2009, Page 10

11 GNU and Linux (Lead) user (from CSAIL at MIT) wanted control Developed key complements: compiler and editor license GPL/ copyleft Linus Torvalds: 91, <10k lines, Intel s 386 architecture Increased functionality Wider variety of hardware platforms Network effects Increasing complexity 8 April 2009, Page 11

12 Red Hat (Lead) user finds Linux difficult to install and use couldn t afford a Unix workstation The Halloween Release Initially inferior, did not appeal to mainstream, potential for rapid innovation, surpasses performance Combine build with distribution channel Key innovations installer package manager Time-paced release work back from ship date 8 April 2009, Page 12

13 Value capture where we re going to make a lot of money is services and maintenance programs.. if you don t have Red Hat Linux, then no one s going to buy support 8 April 2009, Page 13

14 Red Hat - update 2004 ranks first in value amongst CIOs Acquires AOL s Netscape server software Partners with IBM 2005, Enterprise Linux 4 wins OS and Server Product of the year at Techworld IPO in August 1999 shares triple ~2,200 employees ~$500 million in revenue, ~$4,000 million market cap 8 April 2009, Page 14

15 Red Hat s stock 8 April 2009, Page 15

16 Microsoft s take on Linux OSS poses a direct, short-term revenue and platform threat to Microsoft, particularly in server space the intrinsic parallelism and free idea exchange in OSS has benefits that are not replicable with our current licensing model and therefore present a long term developer mindshare threat. commercial quality can be achieved / exceeded by OSS projects OSS is long-term credible FUD tactics can not... combat it. Linux outperforms many other UNIXes Linux can win as long as services / protocols are commodities. OSS projects have been able to gain a foothold in many server applications because of the wide utility of highly commoditized, simple protocols. By extending these protocols and developing new protocols, we can deny OSS projects entry into the market. OSS evangelization scales with the size of the Internet much faster than our own evangelization efforts appear to scale 8 April 2009, Page 16

17 Why does open source work (economically, not technically)? Co-operate create value OS infrastructure, not a basis for differentiation all the users of that infrastructure have a reason to help you build it and ensure that it continues to develop. A modular system enabling decoupled development Expose interfaces and inner workings of modules Compete capture value Deliver complements Package sources and patches Provide professional services Sell hardware that runs Linux 8 April 2009, Page 17

18 Focus of Ubuntu shifting to usability Canonical, the leading backer of the Ubuntu version of Linux, this week said it would hire a team to help make open source software on the desktop more appealing and easier to use. The company plans to sign up designers and specialists in user experience and interaction to lead Canonical's work on usability and to contribute to other free and open source desktop-environment projects, including Gnome and KDE "We are hiring a team who will work on X, OpenGL, GTK, Qt, Gnome and KDE, with a view to doing some of the heavy lifting required to turn those desktop experience ideas into reality," said Mark Shuttleworth. Free Linux desktops should have "a user experience that can compete with Apple in two years", he said at the O'Reilly Open Source Convention last week. "He's bang on the money. Linux absolutely needs more usability," said Mark Taylor, founder of the Open Source Consortium. "Having said that, it's not that hard to find," he said, pointing to the strides made by the Gnome and KDE user interfaces. However, Taylor cautioned against the open source movement taking too rigid a line with developers on usability requirements. "I don't believe we need one desktop to rule them all," he said. Consumers needed a great user experience more than businesses did, Taylor said. IT managers were more likely to use Linux on servers than on desktops. Any desktop implementations they do work with were designed to lock the system down and keep the user within set applications and policies. "Even when they use a Linux desktop, delivering a user experience is not high on the agenda," Taylor said. Shuttleworth said that the freedom of open source software, where developers were free to develop as they wish, could lead to user interfaces that were "patchy and inconsistent" between applications and operating systems. For instance, Ubuntu itself is normally available with the Gnome desktop interface, but one version ships with KDE. Both Ubuntu versions include OpenOffice, which is based on the GTK graphics library. GTK is also used by Gnome, so OpenOffice in KDE would have a different 'open' dialogue to that on the desktop. ZDNet UK 8 April 2009, Page 18

19 Modularity is becoming more prevalent, increasing value creation, challenging value capture Falling costs of co-ordination make modularity easier Modularity, where it can be employed effectively, can accelerate value creation once dominant design established, hence stable architecture and modular interfaces and ultimate performance is not critical autonomous or modular innovations, in this context Very challenging for value capture loss of control for leaders rapid, diverse innovation revenues and value widely dispersed 8 April 2009, Page 19

20 Modularity noun 1. the use of individually distinct functional units, as in assembling an electronic or mechanical system 1 2. designed with standardized units or dimensions, as for easy assembly and repair or flexible arrangement and use 2 1: Random House Unabridged Dictionary, Random House, Inc : American Heritage Dictionary, 2000 Houghton Mifflin 8 April 2009, Page 20

21 Modularity decoupling When a product or process is modularized, the elements of its design are split up and assigned to modules according to a formal architecture or plan. From an engineering perspective, a modularization generally has three purposes: to make complexity manageable to enable parallel work to accommodate future uncertainty Modularity in the Design of Complex Engineering Systems, Carliss Y. Baldwin and Kim B. Clark, HBS Working Paper, January April 2009, Page 21

22 Product Architecture Integral Modular Karl Ulrich, The role of product architecture in the manufacturing firm, Research Policy Volume 24 (1995), pages April 2009, Page 22

23 Interfaces Modular Customer understands and can specify key parameters Can be measured and tested reliably and unambiguously Understand how variation affects system performance Market can function effectively Codified knowledge Difficult to protect Interdependent/ Systemic/Integral Associated with optimizing design for ultimate performance Unstructured technical dialogue Necessary information for market does not exist Management and integration most efficient coordinating mechanisms 8 April 2009, Page 23

24 IBM System/360 First modular computer, conceived as a family of computers different sizes suitable for different applications same instruction set standard interfaces for peripherals Design rules and decentralized development Central Processor Control Office defines rules each team full control over hidden elements Wildly successful, drove other players out of the market BUT undermined IBM s dominance in the long run - through emergence of plug-compatible modules 8 April 2009, Page 24

25 Nippon Denso can make 288 products from just 8 modules 8 April 2009, Page 25

26 Braun family of coffee makers 8 April 2009, Page 26

27 There are six modular operators that together enable a very wide range of system designs Splitting Substituting Augmenting Excluding Inverting Porting Separating systems into modules that interact across well defined interfaces Switching between components that perform the same function Adding a module to increase the functions of the system Removing a module to reduce the functions the system can perform Making an embedded function into a stand-alone module Moving a module from one system to another 8 April 2009, Page 27

28 The resulting systems can exhibit several different types of modularity Slot Component Swapping Direct Component Sharing Bus Bus 8 April 2009, Page 28

29 Modular platforms can be a very effective vehicle for diverse offers 2 May 2007, Page 29

30 Modular platforms can be a very effective vehicle for diverse offers 2 May 2007, Page 30

31 Strategic options Architect For system innovations, may require broad scope of activities at the outset Create design rules, define visible information Convince people this architecture will prevail As modularity established, lead the evolution of the business ecosystem Module player Conform to the architecture, interfaces and test protocols established by others Master the hidden information involved Rely on superior execution 8 April 2009, Page 31

32 Linux 0.01 MacCormack, Alan, John Rusnak, and Carliss Y. Baldwin. "Exploring the Structure of Complex Software Designs: An Empirical Study of Open Source and Proprietary Code." Management Science 52, no. 7 (July 2006) 8 April 2009, Page 32

33 The architecture of Linux is significantly more modular than the architecture of Mozilla MacCormack, Alan, John Rusnak, and Carliss Y. Baldwin. "Exploring the Structure of Complex Software Designs: An Empirical Study of Open Source and Proprietary Code." Management Science 52, no. 7 (July 2006) 8 April 2009, Page 33

34 And than OpenSolaris 8 April 2009, Page 34

35 or XNU 8 April 2009, Page 35

36 Mozilla before and after a purposeful re-design effort modularity to allow participation MacCormack, Alan, John Rusnak, and Carliss Y. Baldwin. "Exploring the Structure of Complex Software Designs: An Empirical Study of Open Source and Proprietary Code." Management Science 52, no. 7 (July 2006) 8 April 2009, Page 36

37 The payoff 8 April 2009, Page 37

38 Open source code has much lower propagation cost than conventional proprietary approaches 8 April 2009, Page 38

39 Financial Management Software 8 April 2009, Page 39

40 Key to value capture is focus on locus of value the virtuous virtuals have carefully nurtured and guarded the internal capabilities that provide the essential underpinnings of competitive advantage they invest considerable resources to maintain and extend their core competences [because without them] their strategic position in the network would be short-lived Henry Chesbrough and David Teece, Organizing for Innovation: When is Virtual Virtuous?, Harvard Business Review, August 2002, pages Attractive profitability seems to flow to the point at which unsatisfied demand for functionality, and therefore technological interdependency exists. Clayton Christensen and others, Disruption, disintegration and the dissipation of differentiability, Industrial and Corporate Change, 2002, pages April 2009, Page 40

41 Locus noun 1. a center or focus of great activity or intense concentration 1 2. a center or source, as of activities or power 2 1: American Heritage Dictionary, 2000 Houghton Mifflin 2: Random House Unabridged Dictionary, Random House, Inc April 2009, Page 41

42 Amdahl s Law: make the common case fast Amdahl s Law is concerned with the speedup achievable from an improvement to a computation affects a proportion P of that computation where the improvement has a speedup of S Amdahl's Law states that the overall speedup of applying the improvement will be God grant me the serenity to accept the things I cannot change (much); courage to change the things I can (a lot); and wisdom to know the difference. - Reinhold Niebuhr 8 April 2009, Page 42

43 Sun Microsystems intense focus on performance bottleneck 8 April 2009, Page 43

44 Interestingly, Sun Microsystems has become a major open source advocate Made Java open source OpenSolaris build a developer community derived from Unix System V only open version OpenSPARC processor architecture design, application development building community around it to advance it OpenStorage built around ZFS combine open source software with industry standard hardware 8 April 2009, Page 44

45 And Apple s OS X is built on open source (as is the code base for the iphone) 8 April 2009, Page 45

46 And Google s Android platform 8 April 2009, Page 46

47 Leveraging open source requires a disciplined approach, from legal and strategic perspectives Architecture Policies Practical and pragmatic procedures for evaluating open source options In-house development Drive how to access software Custom order Existing 3 rd party Open Develop own software, and release to open source community Encourage new open source community development [Google with Android] Integrate 3 rd party open source components Closed Develop own software and keep source and rights internal Pay for development as work for hire and keep rights internal Buy rights to use offthe-shelf 3 rd party software Source: Olswang Open Source Summit, 9 Nov 2007; private discussion and communication with Nigel Swycher of Olswang, and Heather Meeker of Greeberg Taurig, and others 8 April 2009, Page 47

48 The key concern with open source is inherited obligations, copyleft licenses Principal form of License General Public License GPL of the Free Software Foundation Main kinds of open source license Characteristics Copyleft most open; prohibits proprietary forks, strongly driving openness If you add code, and distribute software, you must grant unaltered GPL to combined program - Linux Kernel uses this (GPLv2) Apache Software License (ASL) Allows proprietary extensions or distributions Code issued with this license can be included on closed products Much of Android software uses this (Apache v2) Berkeley Software Distribution (BSD Unix) Few restrictions on what you can add and how you can limit openness of what you add Apple uses this 8 April 2009, Page 48

49 Resources April 2009, Page 49