The S-Curves and Technological Strategy

1
The S-Curves and Technological Strategy

• Henry C. Co
• Technology and Operations Management,
• California Polytechnic and State University

2
Recapitulation

• Technology -- A process, technique, or
  methodology -- embodied in a product design or in
  manufacturing/service -- which transforms inputs
  of labor, capital, information, material, and
  energy into outputs of greater value.
• Technology Change -- A change in one or more of
  the inputs, processes, techniques, or
  methodologies that improves the measured levels
  of performance of a product or process.
• Many growth phenomena in nature show an S
  shaped pattern -- Any single technical approach
  is limited in its ultimate performance by
  chemical and physical laws that establish the
  maximum performance that can be obtained using a
  given principle of operation.

3
Recapitulation (continued)

• Slope of technological trajectory gt slope of
  trajectories of customer need.
• Intel microprocessor speed increases about 20
  per year.
• Eli Lily Purity of insulin improved from 50,000
  ppm in 1925 to 10 ppm in 1980 (by about 14 per
  year).
• Manufacturers of hydraulic excavators increased
  by 15 per year the amount of earth their machine
  could heft in a single scoop from 0.25 cubic
  yard in 1948 to 10 cubic yards by 1974.

4
(No Transcript)
5
(No Transcript)
6
(No Transcript)
7
(No Transcript)
8
Component v. Architectural-Design Improvements

• Engineers manage improvements in overall product
  performance
• by interactively affecting the capabilities of
  components, and
• by refining or overhauling the products
  architectural design.
• To keep up with the industrys pace of
  improvement, technology managers monitor
  improvement trajectories of present and potential
  architectural technologies and the extent to
  which individual component technologies
  constitute actual or potential bottleneck.

9
A Relative Concept

• A read-write head can be viewed at one level as a
  complex system architecture, comprising component
  parts and materials that interact with each other
  within an architectural design.
• The head is a component in a disk drive.
• The disk drive is a component in a computer, in
  which a CPU, semiconductor memory, rigid and
  floppy drive, and I/O peripherals interact within
  a design architecture.
• The computer a component in an information-process
  ing system architecture.

10
Read-Write Head (Disk Drive) Technology
Substitution

• First Technology incremental improvements to the
  original ferrite-head/oxide disk technology
  enabled manufacturers to grind the heads to
  smaller, more precise dimensions.
• Second Technology thin-film photolithography
  displaced ferrite-heads in most disk drives
  between 1979 and 1990.
• Third Technology magneto-resistive heads.

11
Prescriptive S-Curves Strategy

• Becker Speltz (1983), and Foster (1986) urged
  strategists to track S-curves and to acquire or
  develop new technologies in time to switch to it
  when its performance surpasses the capabilities
  of the present technology.
• Hypothesis 1 The industrys leading incumbent
  firms were generally more aggressive in switching
  to new component technology S-curves, but there
  is no evidence that they gained any sort of
  strategic advantage.
• Hypothesis 2 In the disk drive industry, the
  technological change in which attackers (first to
  switch S-curves) have demonstrated strategic
  advantage have been architectural in nature.

12
Hypothesis 1

• Ferrite/oxide technology S-curve
• Disks coated with microscopic particles of
  magnetic metal oxide.
• Efforts to improve density within the particular
  oxide approach involved making the particles
  smaller and more uniform and dispersing them so
  that the maximum possible surface area on the
  disk was coated with magnetic media.
• Thin-film technology S-curve
• When disk engineers felt they had reached the
  limits of fineness, uniformity, and dispersion,
  they turned to thin-film deposition technology,
  attempting to coat substrates with extremely
  thin, continuous coatings of metals.
• Fujitsu and Control Data Corporation (CDC)
  launched development efforts to switch from the
  ferrite/oxide S-curve to the thin-film technology
  S-Curve in 1980 and 1977, respectively.

13
S-Curves For Ferrite/Oxide Technologies at
Fujitsu and CDC
14
Little Evidence That First-Movers Enjoy
Sustained Advantages.

• Actual or perceived limits can be circumvented
  through advances in less mature elements of the
  products design.
• Since thin film deposition technology was not
  quite ready yet when CDC and Fujitsu launched the
  development efforts of thin film technology, both
  firms pushed the ferrite/oxide technology to
  about 3 times the original perceived limits.
• IBM moved to thin film technology at 3.5 mbpsi in
  1979. Hitachi and Fujitsu rode the ferrite/oxide
  technology S-curve far longer (1987) and achieved
  27 and 30 mbpsi respectively.
• IBM, Memorex, Storage Technology, NEC, CDC, and
  Rodime switched to the thin film technology
  S-curve early, but there is little evidence that
  these firms enjoyed sustained first-mover
  advantages.

15
From Ferrite/Oxide to Thin-Film
16
From Ferrite/Oxide to Thin-Film (continued)

• Order of Adoption of Thin-Film Technology Versus
  Areal Density of Highest Performance 1989 Model.

The combined share of the total world market held
by the early adopters of thin-film technology
fell from 60 in 1981 to 37 in 1989. The firms
that switched later (Priam, Micropolis,
Miniscribe, Seagate, HP, Quantum, Toshiba,
Hitachi, DEC, and Fujitsu) saw their combined
world market share rise from 10 in 1981 to 33.
17
Hypothesis 2

• When new architectures emerged in the disk drive
  industry, entrant firms and first movers that
  adopted the new technology early enjoyed a
  decided advantage over industrys incumbent firms
  and generally were able to ride the new
  architectural technology to positions of industry
  leadership.
• Between 1973 and 1990, five successive
  architectural technologies emerged in the disk
  drive industry 14-, 8-, 5.25-, 3.5-, and 2.5
  inch diameter Winchester drives. The drives have
  become smaller, with less parts per unit.
• The advent of new architectural technologies in
  disk drives precipitated the downfall of the
  industrys leading firm.

18
Leading Firms Downfall

• CDC, the dominant 14-inch producer in the OEM
  market, was upstaged by entrants Micropolis,
  Priam, and Shugart in the 8-inch architecture.
• Seagate, Miniscribe, and Tandon entered to
  dominate the 5.25-inch generation, eclipsing the
  former leaders.
• Conner Peripherals and Quantum achieved similarly
  dominant positions in the market for 3.5-inch
  drives relative to the leaders in the in the
  5.25-inch architecture.
• Why were the established drivemaker able to lead
  the industry in developing component technology,
  while they were dethroned at points of
  architectural technology change?

19
Architectural Innovation

• The principal customers for the 14- and 8-inch
  architectures were makers of mainframe and
  minicomputers, respectively.
• Performance measures total capacity and access
  time.
• The 5.25-inch drives that emerged in 1980 had
  capacity of 5 mb and access time of 160 ms.
• The 14- and 8-inch drives had capacity of 100-500
  mb and access time of 30 ms.
• Industry leaders focused on component-level
  improvements that drove performance within the
  14- and 8-inch architectural framework.
• Along other dimensions of performance (capacity
  per cubic inch and cost), 5.25-inch drives was
  superior.
• In the emerging market application for disk drive
  (the desktop PC), the new dimensions of
  performance measures were important.

20
Architectural Innovation (continued)

• Once the new 5.25-inch architectural technology
  became established in its new market, the
  5.25-inch drivemakers found they were able to
  increase the capacity and speed of their drives
  at much faster annual rates than were demanded in
  the desktop market. Within a few years, 5.25-inch
  drives became able to compete with
  earlier-architecture drives in the minicomputer
  and mainframe markets on the original bases of
  performance measures (total capacity and access
  time).

21
A Different S-Curve Model of Architectural
Innovation

• The new technology (2) is deployed in a new
  application (B) wherein performance is defined
  differently than it had been in the established
  market, Application A.
• Technology 2 is in fact the superior performer in
  Application B and achieves a measure of
  commercial maturity there.
• At some point in this progression the new
  architecture becomes capable of addressing the
  performance demanded in the original market more
  effectively than the established technology.

22
Intersecting Performance Trajectories of
Successive Disk Drive Technologies
Average areal density of all models introduced,
in millions of bits per square inch. Bold entry
indicates year in which the architecture captured
over 50 of total industry shipments in 30-100 mb
drives. Underlined entry indicates year in which
the architecture captured over 50 of total
industry shipment in 100-300 mb.