Modularity (Or Not!) and Strategy: the Case of Bicycle Drivetrain Components

Modularity (or Not!) and Strategy: The Case of Bicycle Drivetrain Components

Sebastian Fixson MIT Sloan School of Management [email protected]

Jin-Kyu Park Florida International University [email protected]

Outline

• The Puzzle: Three Questions

• The Data: The Story of Bicycle Drivetrain Components

• The Analysis: Our Interpretation

• The Outlook

1 Background (very brief)

• Study of Technological Change and its Implication has a long history in various Disciplines – Economics, Sociology, and Technology History • e.g., Bijker, 1995; Nelson & Winter, 1982; Rosenberg, 1982; Sahal, 1981 – Technology Management and Strategic Management • e.g., Anderson & Tushman, 1990; Henderson & Clark, 1990; Macher & Mowery, 2004; Utterback, 1994

• More recently, the specific role (and effect) of modular product architecture on industry structure and competition has been of particular interest – Baldwin & Clark, 2000; Schilling, 2000

• And most empirical studies suggest a general migration towards higher levels of modularity – Baldwin & Clark, 2000 – MacCormack, Rusnak, & Baldwin, 2004 – Jacobides 2005 – Shibata, Yano and Kodama 2005

The Puzzle: Three Questions

• What is the Directionality of Product Architecture Change?

• What is the Directionality of Causality between Product Architecture and Industry Structure Changes?

• What is the Origin of these Changes?

2 The Situation – as mostly described

Non-IS Causes

Modular PA1 PA2 Integral

Product a c Architecture Direction of Causality? Underlying Mechanisms? b d

Disintegrated IS1 IS2 Integrated Industry Industry Structure gh

Non-PA Causes

Fig. A.1.1. Product Architectures and Market shares (Road 1984)

Brake Lever

The Data: Shifter

The Industry Derailleur

Before and Segments Freewheel

After Chain

Hub

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Market Share

Shimano SunTour Campagnolo Others

Fig. A.2.7. Product Architectures and Market shares (MTB 1990)Campagnolo Fig. A.1.7. Product Architectures and Market shares (Road 1990) Syncro (3%)

SunTour SunTour Sachs ARIS AccuShift AccuShift II (1%) Brake Lever (11%) Brake Lever (15%)

Shifter Shifter

Derailleur Derailleur Shimano Shimano SIS+HG Campagnolo SIS+HG (78%) (57%) Syncro I Segments Segments Freewheel (28%) Freewheel

Chain Chain

Hub Hub

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% © Sebastian Fixson Market Share Market Share Shimano SunTour Campagnolo Others Shimano SunTour Others

3 Industry Data: Concentration within Bicycle Drivetrain Segments

Road Bicycles Mountain Bicycles 1.0 1.0 Shifter Shifter Derailleur Derailleur Freewheel Freewheel 0.8 Brake 0.8 Brake Chain Chain Hub Hub 0.6 0.6

0.4 0.4 Herfindahl Index [10^4] Herfindahl Index [10^4]

0.2 0.2

0.0 0.0 1984 1985 1986 1987 1988 1989 1990 1984 1985 1986 1987 1988 1989 1990

Industry Data: Concentration across Bicycle Drivetrain Segments

Road Bicycles Mountain Bicycles 1.0 1.0

N/A 0.8 0.8 Type 6

Type 5

0.6 Type 4 0.6 Type 3

Type 2 0.4 0.4 Integration Index Integration

Integration Index Type 1

0.2 0.2

0.0 0.0 1984 1985 1986 1987 1988 1989 1990 1984 1985 1986 1987 1988 1989 1990

4 t-1 = 1980 t-0 = 1985 t1 = 1988 t2 = 1990 Power Gear Brake Power Gear Brake Power Gear Brake Power Gear Brake Functions Transm. Shifting Actuation Transm. Shifting Actuation Transm. Shifting Actuation Transm. Shifting Actuation Function-Component Allocation Scheme PA PA Index 1 3 2 1 3 4 1 3 4 1 3 5 1

Dimension 1 Dimension Index 2 1 1 1 2 2 1 2 2 1 2 2 1

Interface Characteristics Strength 6 6 N/A 7 7 N/A 7 7 N/A 8 7 N/A Industry Leader PA Irreversibility 5 4 N/A 5 4 N/A 5 4 N/A 5 4 N/A

Dimension 2 Standardization 3,3 3,3 N/A 8,8 8,8 N/A 6,6 6,6 N/A 4,4 4,4 N/A

Function-Component Allocation Scheme Product Architecture Product PA PA Index 1 3 2 1 3 2 1 3 4 1 3 4 1

Dimension 1 Index 2 1 1 1 1 1 1 2 2 1 2 2 1

Interface Characteristics Strength 6 6 N/A 6 6 N/A 7 7 N/A 7 7 N/A Rest of Industry PA Irreversibility 5 4 N/A 5 4 N/A 5 4 N/A 5 4 N/A

Dimension 2 Dimension Standardization 3,3 3,3 N/A 5,5 5,5 N/A 7,7 7,7 N/A 8,8 8,8 N/A

Shifter 2,750 * 2,670 4,420 4,200 Derailleur 2,870 * 2,660 4,420 4,220 Freewheel 2,580 * 2,700 4,170 4,180 RB Brake 1,800 * 1,800 3,250 4,210 Chain 1,650 * 1,700 3,260 4,190 Hub 1,000 * 1,150 2,400 4,150

Shifter N/A 4,180 7,460 6,810 Derailleur N/A 4,350 7,620 6,810

[Herfindahl Index] [Herfindahl Freewheel N/A 4,140 7,000 6,550 MTB Brake N/A 3,630 5,420 6,790 Chain N/A 2,110 6,170 6,550 Within Segment Concentration Segment Within Hub N/A 1,310 3,060 6,380

Type 1 5.6% * 13.4% 38.8% 94.2% Type 2 16.3% * 19.4% 25.9% 3.5% Type 3 20.5% * 17.2% 15.0% 0.6%

Industry Structure RB Type 4 19.1% * 17.2% 10.2% 0.0% Type 5 24.2% * 17.9% 6.1% 0.0% Type 6 8.8% * 11.9% 0.0% 0.0% N/A 5.6% * 3.0% 4.1% 1.7%

Type 1 N/A 7.0% 46.8% 94.3% Type 2 N/A 20.9% 23.4% 4.1%

[Integration Index] Type 3 N/A 18.6% 4.3% 0.0% MTB Type 4 N/A 27.9% 13.8% 0.0% Type 5 N/A 16.3% 5.3% 0.0% ©SebastianAcross Segment Concentration Fixson Type 6 N/A 2.3% 0.0% 0.0% N/A N/A 7.0% 6.4% 1.6%

* = Data from 1984 t-1 = 1980 t-0 = 1985 t1 = 1988 t2 = 1990

Data Summary Integral Product Architecture Modular t Integrated Industry Structure Industry Disintegrated

5 The Analysis: Consequences of Product Architecture Change

Systems Firms Improved Systems Firms Product withdraw to Systems try to compete Architecture Niche Markets or Performance with own Systems Change exit the Industry More integral Mechanism 1 Function- Increase in Component Synergistic Allocation Mechanism 3 Specificity Systemic Performance difficult to copy Interface Strength Mechanism 2 Decrease Interface in Competitors’ Irreversibility Network Size

Lower Component Reduced Component Firms Interface Firms try to form Component disappear from Standardization Alliances & JVs or Compatibility the Industry exit the Industry

The Analysis: Our Interpretation

• Product Architectures can shift from (more) modular to (more) integral states (at least temporarily)

• If this shift is successful, it can have powerful implications on the nature of competition in an industry (and ultimately on the industry composition)

• It appears as if the existence of cross-module knowledge is beneficial to create integral (systemic) innovation

6 The Outlook

• Implications for Industry Studies Research – This is one powerful link between Technical Change (in PA) and Industry Structure (Industry Architecture) – More Data needed to better understand Contingencies – Importance of Measurements!

• Implication for Firm Strategy – These Decisions are not ‘only’ Engineering Decisions, they are fundamentally Strategic in Nature!