How Do Firms Compete When Faced with Architectural Changes ? : Lessons from the Optical Storage Media Industry

This study explores how firms respond to changes in product or process architecture. Product or process architecture is the design rules of a system which can be described as the pattern of interrelations between components. Firms must prepare adequate knowledge in a specific architectural condition, so changes of architecture cause serious managerial problems for firms, called modularity traps (or integrity traps). A case study of the optical storage media industry gives some hints for overcoming such traps. First, firms have to alter their strategies and business domains when architectural conditions are changing. That is, in a modularity condition firms have to specialize, and in an integrity condition firms have to coordinate some activities. Second, and more importantly, firms have to retain their component and system knowledge, in order to maintain competitiveness both in specialized activities and in integrated activities.


Architectural change causes serious managerial problems
'Vertical integration to horizontal specialization' has been one of the characteristics of industrial competition since the 1990s.Many firms thought about 'outsourcing' and 'refocusing', to improve competitive performances.In fact, in electronics industries, traditional vertically integrated firms lost competitiveness and networks of specialized firms took over them (Sturgeon, 2002).In computer Nakagawa 2 industry, specialized firms such as Intel or Microsoft who can make each component of a computer, replaced integrated firms such as IBM (Grove, 1995).
In semiconductor business, instead of vertical integration, specialization in design (fab-less) and that in manufacturing (foundry) became the dominant form of organization.
Those changes were caused by modularity.
Modularity is the decoupled condition of product or process architecture.When architecture changes towards modularity, each component or each business activity can be done independently, and firms can get competitive advantage by focusing their resources on specific business areas.
Modularity is the essentials of the trends towards specialization in these days.
But architecture not only goes modular but also goes integral.When architecture return from modular to integral, specialized firms, who gained competitiveness in modularity, do not respond to integrity and lose competitiveness.Kusunoki and Chesbrough, 2001 called  Architecture is one of the perspectives of artificial systems.'Artificial' means human-made, and 'system' means composed of more than one component.So an artificial system is a thing that consists of some components made by humans.It includes not only concrete objects, such as automobiles or mobile phones, but also immaterial properties such as economic system or human relations (Simon, 1996).
'Architecture' as a characteristic of a system can be described as the pattern of interrelations between components (Baldwin & Clark, 2000).We can distinguish the architecture of systems into two types by the strength of the interdependence of their components: integral and modular.A modular architecture specifies independence (a decoupled structure) of each component.An integral architecture specifies interdependence (a tightly coupled structure) of each component (Ulrich, 1995).
The automobile, the quality of which is realized by integrating numerous parts, is a typical integral architecture product.On the other hand, the personal computer (PC), whose components, such as the MPU, DRAM, or HDD, are developed and produced independently, is a typical modular architecture product.
Firms must have adequate knowledge and capabilities for their confronting product or process architecture.In other words, different types of knowledge are required in different types of architecture (Ulrich, 1995).When the system has an integral architecture (consists of tightly coupled subsystems: Simon, 1996) we need mutual coordination and architectural knowledge (knowledge about the system) in order to produce, develop or operate that system.When the system has a modular architecture (consists of nearly decoupled subsystems) we do not need knowledge of the system but we need special, detailed component knowledge (Henderson & Clark, 1990, Sanchez & Mahoney, 1996) (Figure 1).
Because of these relationships, value chain structure can be laid out in accordance with the product or process architecture (Figure 2).In integral architecture, a firm should have architectural knowledge, so it will choose a vertically integrated form of organization, whereas in modular architecture the firm should take a specialized form in order to focus its resources on a limited area (Sturgeon, 2002).

When the architecture changes
Next we will consider changes in the product or process architecture.As we saw in the previous subsection, the architecture decides the value chain structure.Thus, changes in architecture cause changes in the value chain structure.
When the architecture of an industry is relatively stable, it is easy for firms to take adequate strategies and organization design.In modularity, a firm should specialize their domain where they have core competence.In integrity, a firm should integrate its business activities.However, when the architecture changes continuously, the problems faced can become difficult.In turbulently changing competitive environments, known as hypercompetitive (D'Aveni, 1994) or fast-clockspeed industries (Fine, 1998), the structure of interactions between business activities can change continuously and violently (Figure 3).Within such a turbulent architectural change, how do firms decide on their strategies and organizational designs?4).
So far, we considered the problem from a theoretical perspective.From the next section, we think about it from an empirical point of view.In order to certify how a firm can keep competitiveness in a turbulently changing value chain structure, we adopted one hypercompetitive industry as a unit of the study, that is, optical storage media industry.
Whether strategies mentioned above gives firms competitiveness or not will be clear from the case studies.

Research method
This paper is based on one detailed industry case Research (a research company in Japan).Also, I used several secondary sources such as newspapers, catalogs and some academic papers and books to triangulate the facts. 2We piled up those data and made up the industry level analysis.
Optical storage is a system of reading, erasing and storing data by the optical technology.It is famous as CD or DVD.This system is made up of two devices: the drive and the media.The drive reads, writes and erases the data on the media (Figure 5).The media and drive are physically separate from each other.The business of the optical media is also separate from that of the drive.
Writable optical storage media, which we analyzed here, are optical media on which information can be written.They are known as 'CD-R' (compact disc-recordable) or 'DVD-R' (digital versatile disc-recordable).
In the case study, we focused on the conditions of process architecture of OSMW industry.We analyzed the process architecture, because the production process of optical storage media is process-intensive, so the industrial structure is affected by process architecture more than product architecture.To make the analysis simple, we picked up three important processes among the whole production processes.I picked out three activities; 'final product production,' 'organic dye production,' 'manufacturing equipment production.' Organic dye is the most important material in making writable optical storage media, because the specifications and the qualities of organic dye mainly decide the lifetime of final product and the precision of data writing.The manufacturing equipment production is also important, as it affects on final product's cost and quality.Furthermore, I added the 'final product sales' into the analytical unit of the case study, because the relationships between production and sales are also the important factor for the settings of industrial structure.The flow of those activities is

Case studies: Writable optical storage media industry
The writable optical storage media (OSMW) industry started in 1989, when the first CD-R was produced.After that, the industry developed step by step, and in 2004, the OSMW industry yielded sales of about 4.5 billion dollars and produced 14 billion units of products (Figure 7).

The first stage of OSMW industry
The OSMW industry started in 1989 when Sony and Taiyo Yuden developed and produced the CD-R.At that time Sony mainly commercialized the CD-R drive and Taiyo Yuden did the CD-R media.Taiyo Yuden's concept of CD-R media business was that 'everyone can make his own original CD at a relatively low price,' so the target of the CD-R was the consumer market.But that plan did not go well.
So the CD-R standard did not spread into the consumer market in the early 1990s.In other words, the early CD-R business had no customers.Firms had to find customers other than consumers.
Therefore, the early entrants into the OSMW industry looked for demand through their own

The way to modularity
In the late 1990s the process architecture of OSMW industry went modular.On the production side, some technological innovations led to modularity.
Gradually, functional interrelations between    In organic dyes, Ciba Specialty Chemicals obtained a monopolistic share by selling highly equipment-independent dyes.In manufacturing equipment production, firms who had developed in-line equipments early received many orders and enjoyed high profitability.Before Ciba Specialty and process equipments makers produced those materials and equipments, firms could not make CD-R media without dyes and equipments adjusted to each other.
And Japanese and American existing firms did not have sold those integrated dyes and equipments.So 5 For the reason behind the bending of the experience curve, see Abell and Hammond (1979) or Shintaku (1994).
huge capacity, firms have to realize strict specifications by integrating material technology and process technology.For example, the quality of the writing layer on DVD media, where the data are recorded, is determined by the interaction between injection machine performance and the organic dye's specifications.In addition to realize strict specifications, the users had wanted DVD media to be better quality than CD-R media.In 2002 a DVD recorder for TV diffused into the market as one of digital home appliances but people came to dislike the writing errors of DVD media that occurred when they recorded TV programs, as VCR had rarely failed to record.So DVD media manufacturers tried to improve the product quality by adjusting the materials and equipments to respond to customer needs.
On the other hand, the relationships between production and sales did not change drastically.
Though users want quality more than low cost, basically the target market-the consumer-did not change.Sales companies did not have to alter methods of sales and distributions.They did not need mutual interactions with production and they focused on sales activities only, that is, marketing, branding or building distribution systems.
In this integration phase, Mitsubishi Kagaku became a leading company in the OSMW industry.
Mitsubishi Kagaku had integrated all its activities vertically in early 1990s, but it specialized in sales when the industry went modular.And then, Mitsubishi Kagaku succeeded in the technical integration of writable DVD in 2002.The profit rate of Mitsubishi Kagaku had not been announced, but about a half of our interviewees answered that Mitsubishi Kagaku is the best performer in the OSMW industry.Now Mitsubishi Kagaku had about a 20% share in final product sales, which is the largest in the world.And it had monopolistic share in organic dyes, but it did not have DVD production capacity.How did Mitsubishi Kagaku succeed in the DVD business?Mitsubishi Kagaku did not internalize business activities into its domain, and kept specialization in limited areas.But, Mitsubishi Kagaku succeeded in integration of the process architecture.To realize Mitsubishi Kagaku's 'recipe' was an aggregate of know-how for making good quality writable DVD media.By using that recipe, Mitsubishi Kagaku and an equipment maker jointly developed the dyes and the equipment.Mitsubishi Kagaku (and the equipment producers) obtained a large share in dye business (or equipments business).In addition, Mitsubishi Kagaku sold the dyes, the equipment and the recipe as an integrated production system only to some exclusive producers.They could produce better DVD media by that integrated system than by using dyes and equipment bought separately from markets.Mitsubishi Kagaku bought better quality DVD media produced from those exclusive producers and sold them.In this way Mitsubishi Kagaku succeeded in the DVD business.
After Mitsubishi Kagaku's leap forward, some firms followed Mitsubishi's ways and tried to integrate architecture.Among them TDK succeeded in integration of the process architecture.TDK also kept knowledge for integrated production system, so TDK could follow Mitsubishi's way.Taiyo Yuden was the third successful company.Taiyo Yuden chose final product manufacturing business in the late 1990s.Then Taiyo Yuden did not compete directly with Taiwanese makers.Taiyo Yuden focused their business on differentiated niche where the customers needed above normal product quality, so it kept integrated production way.So Taiyo Yuden could respond to the change of the architecture when the writable DVD media emerged.Taiyo Yuden would realize integration in DVD business easier than Mitsubishi Kagaku because only it had to do was to continue the same way of business as it did in CD-R business.But Taiyo Yuden did not have sales functions, so they could not earn more than Mitsubishi Kagaku.To think from the opposite side, Mitsubishi Kagaku's success was attributed to the creation of effective business model that Mitsubishi Kagaku could exploit whole value chain by possessing the upper end (organic dyes) and the lower end (sales to final customers).Specialized firms that did not participate in Mitsubishi's or TDK's group lost competitiveness.
Ciba Specialty Chemicals continued organic dye development by itself in the DVD business, but their dyes did not satisfy the customer's quality needs.
Firms who developed only equipment also lost market position.In DVD final production, CMC or Indian company MBI (Moser Baer India), who placed importance on cooperation with other companies, improved their performance.They introduced Mitsubishi's or TDK's integrated system and got orders from them, so they could produce better DVD media and achieved a relatively stable competitive environment.However, firms that did not cooperate with others were placed in a position of fierce competition.They could not make good quality DVD media, so they had no way to sell them at lower price.
In the DVD sales business, although Mitsubishi improved its share and the sales specialists showed a decrease in profits, a general survey of the situation did not change drastically.Memorex and Imation still kept their profit by specializing in sales, because performance in this area was decided by sales capability, regardless of the strength of these companies' other capabilities.

Discussion and conclusion
In the previous section, we observed the OSMW industry from a process architecture perspective.
And we confirmed that the changes in process architecture drove changes in the value chain structure.In the early 1990s process architecture was Finally the contributions of this paper will be discussed.This paper's novelty exists in its dynamic view of competition.Past research has suggested strategies in modularity (Baldwin & Clark, 2000, Sturgeon, 2002) and those in integrity (Takeishi, 2003).But there has been little research on how to compete when the architecture changes.Kusunoki and Chesbrough (2001) and Fine (1998) dealt with the dynamics of architecture, but they did not reach a solution to it.This paper has tried to find out how to compete in the change of the architecture.I consider this the main contribution of this study.
Some researchers have already reached the conclusion that knowledge should be kept widely regardless of the domain of a firm's business (Brusoni & Prencipe, 2001;Takeishi, 2003).Those researches insisted that knowledge wideness is important from static analysis.This paper extends the discussion from static to dynamic.That is, here we insist that to keep knowledge leads to competitiveness in a dynamic changing environment.
In such turbulent environments, firms have to achieve strategic flexibility in order to respond to the changes (D'Aveni, 1994).From this study, we can say that keeping knowledge would be one of the bases of strategic flexibility.
this problem as 'modularity trap.'Then, how do firms respond to such two-way architectural changes?When architecture changes to modular, firms should specialize.But, specialization disables a firm to respond to reintegration.To think about this dilemmatic problem, we analyze optical storage media industry.2. What is architecture?: Literature review 2.1.Architectural perspective of the product and process

Figure 1 .Figure 2 .
Figure 1.Relationships between knowledge and architecture type of architecture

Figure 4 .
Figure 4.When the product or process architecture changes, how do firms respond to that? how do firms respond to that?

Figure 5 .
Figure 5. System of optical storage

Figure 6 .Figure 7 .
Figure 6.Flowchart of important activities of OSMW industry organic dye production manufacturing equipment production efforts.For example, Taiyo Yuden sold their media to corporate and government customers such as NASA who needed fast, high-volume, stable data storage technology.Kodak sold CD-Rs as storage media for digital photographs (photo-CDs). 3To increase production volume, firms had to find usage and customers for CD-Rs, so the process architecture between sales and final product production seemed to be integral in this sense.Not only the customers, but also the production technology, were different in each firm.Every entrant adopted a different material (organic dyes) and different production technologies (manufacturing equipment), which were developed through each firm's own efforts.Manufacturing equipment and materials (especially organic dyes) were functionally interactive, so firms had to develop and adjust them to improve the final product quality.In order to differentiate in quality and adjust to the customers' several needs, firms made CD-R media with original technology that used original organic dyes and original equipment.To meet the architecture described above, the 3 For the early phase of the OSMW industry seeNakajima (1998).entry firms, Taiyo Yuden, TDK, Mitsubishi Kagaku, Mitsui Kagaku and Kodak selected a vertically integrated form.No firm specialized in any one activity.Every firm used original dyes and manufacturing equipment, made high quality products, and sold them to corporate customers.This vertical structure stayed in place until 1996.
materials and equipment came to be clarified, and development activities for both materials and equipment could be done independently.Ciba Specialty Chemicals developed the organic dyes, by which firms could produce good CD-R media using any equipment.Similarly, some equipment companies produced the manufacturing equipments, called in-line equipments, which yielded good quality CD-Rs stably in any environment and with any material.By using these dyes and equipments, 'everyone can produce CD-R, if he can get the materials and equipments.No knowledge of equipments and materials is required.' 4he market for CD-R media also changed in the late 1990s.From 1997, CD-R drives for personal computers were developed and spread rapidly, as personal computers were being sold all over the world.As a result, consumers started to use the CD-R at last.The market changed rapidly from corporate to consumers.In the early 1990s corporate customers bought good quality CD-R media, but in the late 1990s the consumers bought low price CD-Rs with famous brand names on it; they were not so concerned about quality.In addition to such qualitative changes, quantitative change also heavily affected sales activities of the firms.In the past, firms had sold millions of CD-R media, but now they sold from several hundred million to a billion CD-R media.Sales activities needed large administrative functions, marketing professionals and a great distribution network.These capabilities were specialties for sales, so sales activity became functionally independent from production activity.

Figure 8
Figure 8 summarizes changes of the architecture.Technological and market changes led to modularity of business activities.All the interactions between activities were no longer needed.In this modularity phase, specialized firms

Figure 8 .
Figure 8. Business architecture went modular in the late 1990s

Figure 9 .
Figure 9. Trend of writable optical storage media production by nationality

Figure 11 .
Figure 11.CD-R media sales share in 2000

Figure
Figure 12.Writable DVD media production

Figure
Figure 13.Reintegration of business architecture

Figure 14 .
Figure 14.Business model of Mitsubishi Kagaku