Exploring the Limits of
the Technology SCurves – Component
Technologies
Clayton Christensen
Technology S-Curve
Growth
Product
Performance
Emergence
Time or Engineering Effort
Maturity
Technology S-Curve
• It has become a way of thinking about technological
improvement over a period of time
• Theory
– Early stages improvement in performance is slow
– As the technology is understood and diffused,
rate of improvement increases
Technology S _Curves
• S-Curve - used at the industry
level
– Incumbent firms are concerned
about refining existing
technologies
– They lose their positions of
dominance to new entrants
Typologies of technological
change
• Architectural change
– Rearrangement in the way components are
relate to each other
• Using the motors and fan blades coming up with a
table fan
• Modular change
– Fundamental change in the technological
approach employed in a component where the
architecture is left unchanged
• Changing the type of motor in a ceiling fan
Typologies of technological
Change (Contd)
• Incremental change
– Improvements in component
performance
• Better quality RAMs, memory chips
• Radical innovation
– Change in architecture and new
approach in the component level
Using the technology S-Curve at
the firm level
• Why is harder to get
performance improvement as a
technology reaches maturity?
– Scale phenomenon (things get too
large or too small)
– System complexity
The Disk Drive industry
• What is the performance measure for
incumbents?
– Areal recording density
• When resources spent in engineering
improve the performance of a
technology, there is less of an
incentive to switch to alternate
technologies.
Component and Architectural
technologies
• Disk Drive System
– Component Technology
– Read write technology on the disk drive
• Ferrite and Oxide Technologies Vs. Thin Film Heads
• Incumbents prefer to work on existing technologies and
make incremental changes that bring performance
improvements
– Resource Rich companies invest in radical
component improvements
• Thin film heads cost IBM over $1 billion and took
over 10 years
• Architectural technology
• Cost significantly less and can be developed in
lesser time
Timing of adoption
• Thin film technology replaced the Ferrite heads
– The time at which different firms switched to the
new component technology varied over a 10 year
period
– The extent of performance improvement was
also different
• IBM was one of the early movers (1978)
• Hitachi and Fujitsu switched much later in the
mid 1980s
• Switching to a new technology did not
improve the performance
Timing of adoption
(Contd)
• No relationship between timing of adoption
and performance improvement
• Early adopters no clear improvement
in storage capacity
• Later adopters able to work with the
technology and improve performance
• Companies had different strategies in the
way they adopted component technologies
– Some companies like IBM choose to
switch to new technologies
– HP preferred to improve existing
technologies
Incumbents vs. New
Entrants and S-Curves
• Incumbents are more likely to
succeed wrt changes to component
technologies
• New Entrants seldom succeed with
Component technologies
• The story is different with
architectural technologies
• Component technologies reinforce
existing competencies
• Architectural technologies look at
competencies with a different lens.