Technological Evolution Charles Weber PSU-ETM EGMT 510/61

1
Technological Evolution

• Charles Weber
• PSU-ETM
• EGMT 510/610 INNO
• Innovation Management

2
Technological Evolution Questions of Interest

• Why do light bulbs look like they do?
• Why does an ergonomic computer keyboard look like
  it does?
• Which features of light bulbs and ergonomic
  computer keyboards were inevitable and which were
  path dependent? (Gould, 1988)

3
Technological Pioneering(Rosenblum Cusumano,
1988)

• Key Questions
• What is technological pioneering?
• How did we end up with VHS as a video format?

Prof. Michael Cusumano MIT Sloan School
4
The Evolution of Industries(Utterback,
Abernathy, Suarez)

• When is the best time to enter into an industry?
• How does the incumbent generally respond to the
  challenge from an entrant?
• Both ecological models and evolutionary models
  explain the mortality of firms in an industry.
  Why should we prefer one to the other?

Prof. James Utterback MIT Sloan School
5
Entering New Businesses(Roberts Berry, 1985)

• Key Questions
• Which product markets should a corporation enter?
• How should the company enter these product
  markets to avoid failure and maximize gain?

Prof. Edward B. Roberts MIT Sloan School
6
Entry MechanismsFrom Roberts Berry (1988)

• Most entries constitute substitution threats.

7
Optimum Entry Strategies(Roberts Berry, 1985)
8
The Incumbents Response to Substitution (Cooper,
Schendel, Utterback)

• In every industry studied, the old technology
  continued to be improved and reached its highest
  stage of technical development after the new
  technology was introduced.
• Ice making versus refrigeration
• Steam ships versus sail ships
• Most incumbent firms tried to participate in the
  new technology, but most who did so failed.
• It is common for the incumbent firms to emphasize
  the shortcomings of the new technology.
• Why do we have the electric chair?

9
Dominant Designs (Abernathy Utterback, 1978)

• Key Questions
• What is a dominant design? Cite examples.
• Under what circumstances does it occur?
• What type of innovation occurs before the
  introduction of the dominant design? Thereafter?
• Economics, a functionalist discipline, would
  suggest that the optimal design would become
  dominant. Why does this not always occur?
• Are dominant designs inevitable?

10
About Dominant Designs(Abernathy Utterback,
1978)

• A dominant design.
• Constitutes industry agreement on the attributes
  of a design that holds overwhelming market share
• Tends to be an accumulations of prior
  innovations
• Is not the best design in individual performance
  categories
• The DC-3 was not the largest, or fastest or
  longest-range aircraft (A/U 1978, p. 29)
• The best overall design? (early Utterback)
• The DC-3 was the most economical large, fast
  plane able to long distances. (A/U 1978, p. 29)
• The best available design that meets market
  demands? (Utterback, 1994) (e.g. Microsoft
  Windows)
• An adequate design that is supported by major
  players? (e.g. the IBM-PC)

11
Examples of Dominant Designs(Utterback Suarez,
1992)
Question Do the introductions of these designs
truly constitute the seminal events in the
history of the above industries?
12
Design Hierarchies and Dominant Designs
(Utterback Suarez, 1992)

• The evolution of designs is path dependent.
• Dominant designs are not necessarily optimal.
• e.g. QWERTY keyboard
• Can you cite other examples?

13
The A/U Model of Innovation(Abernathy
Utterback, 1978)(Utterback Suarez, 1992)

• Three Phases Fluid, Transition and Specific
• Introduction of dominant design .
• Triggers transition to specific phase.
• Triggers the shift from product to process
  innovation
• Increases the mortality of firms within an
  industry.

14
The Fluid, Transitional and Specific Phases of
Industrial Evolution(Abernathy Utterback, 1978)
15
Entry and Exit of Firms in the Typewriter
Industry (Utterback Suarez, 1992)
16
The Optimal Time of Entry

• In the very early stages of an industry entry is
  risky.
• The odds of successful entry are highest prior to
  the emergence of a dominant design.
• A few years after the emergence of the dominant
  design entry is virtually impossible.
• Only substitutions have a chance of being
  successful.

17
Dominant Designs and the Mortality of Firms
(Utterback Suarez, 1992)

• Arrows indicate date of dominant design.
• The IC industry has no dominant design.

18
Determinism versus VoluntarismThe Relevance of
Management

• Organizational Ecology
• (Hannan, Freeman, Boeker)
• Organizations choose niches and adapt to them.
• Organizational mortality is proportional to the
  number of competitors in an environment.
• Most competitors die early.
• Organizations become rigid.
• Management can do little to respond to a shift in
  the organizational environment.

• Evolutionary Models
• (Schumpeter, Tilton, Penrose, Nelson, Winter,
  Gould)
• Managers make decisions that influence the
  organizational environment.
• Organizations can become rigid.
• Management can make decisions in response to
  environmental shifts.
• Some of these decisions are irreversible.
• The evolution of industries is path dependent.

19
Organizational Ecology versus the
Abernathy-Utterback Model

• Both predict firm rigidity and a high firm
  mortality in mature industries
• Organizational ecology does not base argument on
  industry structure.
• The A/U model provides an in depth insight into
  industry structure that is based on innovation.

20
Mapping Creative Destruction(Abernathy Clark,
1985)

• Key Questions
• Abernathy and Clark classify innovations along
  two axes and into four categories. What are they?
  
• Why do different kinds of innovation occur at
  different times in the history of an industry?

Prof. Kim Clark Dean of Harvard Business School
21
Types of Innovation (Abernathy Clark, 1985)

• Architectural Innovation
• Defines basic configuration of product and
  process.
• Establishes the technical and marketing agendas
  that will guide subsequent development.
• Lays down the broad architecture of the industry.
• Market Niche Creation
• Matching customer needs with refinements in
  technology.
• Regular Innovation
• Builds on established technical and production
  competence
• Is applied to existing markets and customers.
• Can have strong cumulative effect.
• Radical Innovation.
• Renders established technical and production
  competence obsolete.
• Yet is applied to existing markets and customers.

22
Key Innovations in the Auto Industry

• Industries evolve in phases.
• Architectural innovation comes first
• Others types of innovation follow.

23
The Transilience Map(Abernathy Clark, 1985)

• Two Axes
• Technology/ Production
• Market/ Customer
• Four types of innovation
• Regular
• Revolutionary
• Niche creation
• Architectural

24
Routines and Rigidity (Nelson, Winter, Leonard)

• Key Questions
• How do routines affect innovation?
• Can organizations become too rigid to innovate
  effectively?
• Do evolutionary theories completely describe
  creative destruction?

Prof. Dorothy Leonard Harvard Business School
25
Creative Destruction

• Organizations learn by executing routines and
  procedures (Nelson Winter, 1982).
• Knowledge becomes specialized and internalized
  (converted from explicit to tacit). (Nonaka,
  1994)
• Organizations can become rigid to the core.
  (Leonard, 1992)
• Existing routines, old knowledge and perhaps
  existing organizations may have to be destroyed
  for new routines new knowledge to be created.
• Can you give an example from your experience?

26
Technology Trajectories(Dosi, Sahal, Moore)

• Key Questions
• What is a technology trajectory?
• Are there guideposts for technological
  trajectories?
• What is Moores Law?
• Why has Moores Law kept on going for over 30
  years?
• Is the (Inter)national Technology Roadmap a
  moving guidepost for Moores Law?

27
The S-Curve as a Technology Trajectory (Foster,
1986)

• Slow initial innovation rate
• Accelerates
• Until it reaches diminishing returns

28
Paradigms and Trajectories(Dosi, Research Policy
11, 1982)

• Technological Paradigm a model and a pattern of
  solution of selected technological problems based
  on selected material technologies.
• Technological Trajectory the pattern of normal
  problem solving activity (progress) on the ground
  of a technological paradigm.
• Continuous improvement along dimensions of merit.
• Why to technological trajectories keep on going?

29
Features of Trajectories (Dosi, 1982)

• Continuous improvement along dimensions of merit
• Morphogenesis (Sahal, 1985)
• The form of a system must be approprieate to its
  size.
• change in size mandates change in shape.
  Complementary trajectories
• Accumulation of technology and knowledge
• Mutual Exclusion difficulty from switching from
  one trajectory to the other.
• Technological guideposts, which set direction for
  innovation? (Sahal, 1985)

30
Morphogenesis in Innovation(Sahal, 1985)

• Technologies are continuously made to become
  larger or smaller.
• The relationship between size and structural
  requirements changes.
• Change in size of an object
• Accompanied by change in size of components
• Necessitates changes in materials of
  construction.
• Introduces various complications in its structure
• Learning by scaling

31
Examples of Morphogenesis(Sahal, 1985)
32
Complementarity

• Complementors neither buy form nor sell to each
  other.
• Together their technologies enable a third.

33
Technology Drivers

• One technology spurs on the development of
  others.
• Example Innovation in microprocessors drives
  innovation semiconductor processes, which drives
  innovation in process equipment.

34
Moores Law (Moore, 1975)

• Density of semiconductor chips increases by
  factor of four about every 30 months.
• Performance increases and price decreases
  accordingly.
• This trajectory has been ongoing for close to 40
  years.
• Semiconductor feature sizes shrink and process
  technology improves accordingly.
• Continuous miniaturization induces exponential
  growth in investment. (Moores Second Law)
• Roadmap committee (SRA-ITRS) was assembled to
  make the continuation of Moores law affordable.
• Is the ITRS a guide post for innovation in the
  semiconductor industry?

35
ArchitecturalInnovation(Henderson Clark, 1990)

• Key Questions
• How does the meaning or the term architectural
  innovation in Henderson Clark differ from the
  Abernathy Clarks definition?
• According to Henderson Clark, what is the
  entrants primary source of competitive
  advantage?
• Henderson Clark predict architectural
  innovation favors the entrant. However, since
  1990 the incumbents have dominated the
  photolithography industry. What happened?

Prof. Rebecca Henderson MIT Sloan School
36
Henderson Clarks (1990) Framework for Defining
Innovation
37
Architectural Innovation andthe Failure of
Established Firms(Henderson Clark, 1990)

• Organizational subsystems in established firms
  continuously improve particular technological
  subsystems (modules).
• Modular innovation may require an adjustment in
  the relationship between modules (architectural
  innovation).
• Established firms are notoriously bad at
  architectural innovations. They have difficulty
  managing inter dependencies.
• Existing structures interfere with architectural
  innovation.
• Entering firms without pre-existing,
  module-specific organizational structures have
  competitive advantage.

38
Attributes of Photolithography Tools (Henderson
Clark, 1990)
39
Questions Concerning Architectural Innovation

• Henderson Clark predict architectural
  innovation favors the entrant.
• However, since 1990 the incumbents have dominated
  the photolithography industry. What happened?
• Possible explanations
• Incumbents are better at managing complementors
  (Tripsas).
• Incumbents have mastered architectural
  innovation.
• Other factors like the investment capacity are
  more important than architectural innovation.

40
Conclusions Regarding Technological Evolution

• The evolution of industries is path dependent.
• These paths (trajectories) can continue for
  decades.
• The outcome of technological evolution can be
  suboptimal.
• In general small entrants are more flexible than
  large incumbents.
• Incumbents respond to threats by improving the
  existing technology.

41
Unanswered Questions

• How do technologies diffuse into markets?
• Evolutionary models suggest various rates of
  change but no radical transformation of
  industries. Yet these transformations occur. How?