1 Technology S-Curve

2 Outlines l Abstract l The usefulness of technology S-curve at the industry level l The limitation of S-curve at the individual firms level l Summary l More discussions

3 l The uses of S-curve at the industry level :  The description of the magnitude of improvement  The prescriptive S-Curve theory l Product performance results from:  Component technology  Architectural design What is “S-Curve”? S-curve can provide convincing explanations of why alternative technologies have made substantial inroads against currently dominant technology?

4 The Position on S-curve Corresponding to BCG Market Share Growth HIGH LOW Product performance Time or engineering effort ? ?

5 The Limitation of S-Curve l From the point of view of a manger within a single firm, could the S-curve be the prescriptive tool for new component technology development? (at the individual firm level)  The observed maturation of a technology maybe the result, rather than the cause, of the launch of an alternative development program.  Nobody knows what the natural, physical performance limit is in complex engineered products.  The flattening of S-curve is a firm-specific, rather than uniform industry, phenomenon.  Extending the conventional technology S-curve, rather than switching S-curves?  By improving the architectural system  By applying effort to less mature element of the system

6 Magnetic Rigid Disk Drives l Hard Disk industry :  During 1970~1989, the improvement was steady, averaging 34% per year  With time as the horizontal metric, no S curve pattern of progress is yet apparent.  Measure total industry revenue as a proxy for engineering effort

7 Using S-Curve to Prescribe Development of New Component Technologies l The risk to switching to a new S-curve.  Cost more and take much longer time l When to manage the switch from one component technology to another?  Engineers sensed they were approaching the physical limit of ferrite cores before  With a process used in integrated circuit manufacturing, thin-film photolithography, they can create much smaller, more precise electromagnets on the head.

8 l The areal density was pushed to about triple the level at which seems initially to have planned to abandon technology. l Is 30 mbpsi Fujitsu reached in 1987 the “real” natural limit of ferrite heads and oxide disk? Two S-curves for Ferrite-Oxide Technologies at Fujitsu and CDC The observed maturation of a technology maybe the result, rather than the cause, of the launch of an alternative development program.

9 Points at which Thin-Film Technology was Adopted by Leading Manufacturers, Relative to the Capabilities of Ferrite-Oxide Technology at the Time of the Switch

10 l Only 5 of the 15 firms shown actually leapt above the convention technology. l Conventional technology progressed far further than anyone expected. l Different competitors switched S-curves at different points. l Little evidence show that companies switched S-curve early enjoyed attacker’s advantages. Points at which Thin-Film Technology was Adopted by Leading Manufacturers, Relative to the Capabilities of Ferrite-Oxide Technology at the Time of the Switch

11 Relationship between Order of Adoption Thin-Film Technology and Areal Density of Highest Performance 1989 Model There is no correlation between order of adoption and rank order of density  Entrants enjoy no attackers advantage.

12 IBM & HP

13 Relationship between Order of Adoption Thin-Film Technology and Areal Density of Highest Performance 1989 Model l Entrants enjoy no attackers advantage. l No systematic differences exist in how firms respond to potential maturity in component technology. (EXHIBT 8)  IBM, switching to advanced component technology  HP, relying upon  Incremental improvement in established component technologies  Refinements in system design Switching to new S-curve is not the only option.

14 S-curve of Architectural Innovation l Different from S-curve of component innovation  Architectural technologies indeed follow S-curve patterns! l Timely S-curve switching seems critical when confronting architectural technology change. l Not only technological dimensions but also market innovation.

15 Comparing Prescriptive S-curve and S-curve of Architectural Innovation

16 Comparing Prescriptive S-curve and S-curve of Architectural Innovation (con’t)

17 Conclusions l The application of S-curve at a managerial level seems to be very ambiguous. l There is more than one way to skin the cat. l There was no clear evidence of any first mover benefits or “attackers’ advantage.” l Comparing with architectural technologies. 1. Switching to new component technology S-curve early results in no competitive advantage 2. Switching to architectural S-curve enjoys powerful first-mover advantage