Browse by year:
June - 2007 - issue > Innovation@work
Focus on Architectural Innovation
Sanjeev Chopra
Friday, June 1, 2007
Whether it is venture capitalists, entrepreneurs, or fresh college graduates, everyone has a stake in understanding the innovation process. It is a myth that innovative ideas come from a stroke of genius or a flash of brilliance. No doubt, it helps to have some of those faculties, but one can also find opportunities for successful innovation by being aware of the different types of innovations and their impact on industry participants (customers, startups vs. incumbent vendors, etc.). The key difference between innovation and invention is that unlike invention, successful innovation is judged by its ability to earn a profit. Therefore, we can have no discussion about innovation without discussing the ability to profit from it.

In this article I will focus on two different types of innovation and submit to you that upstarts or startups have a better shot at success if their innovation is architectural in nature.

Architectural vs. Component Innovation:
Architectural Innovation: Architectural innovation refers to innovation in the architecture of a product that modifies or changes the way different components of the systems interact or link with each other. The different components of the system may also be modified within the new architecture (smaller form factor, lighter weight, etc) however, the key technologies at the component level remain unchanged. Some examples of architectural innovation:

1. ‘3.5’ disk drives were an architectural innovation over the ‘8’ or ‘14’ Winchester disk drives. The ‘3.5’ disks were smaller, lighter, and lower in cost (in terms of total cost and not cost per Megabyte) than the Winchester architecture.

2. Desktop photocopiers represent an architectural shift over the big stand-alone photocopiers.

3. Multi-core microprocessor architecture is an architectural shift over single core CPU architectures.

Component Innovation: Component innovation refers to innovations that improve the performance of one or more of the components that make up the system, leaving the overall system architecture largely untouched. Some examples of component level innovations are:

1. Continuous reduction of minimum feature size in transistors (going from 180nm to 135nm to 90nm, etc.) that improve the performance of an integrated circuit.

2. Grinding the magnetic head of a disk drive finer so that you can store more bits per square inch and improve the capacity of disk drives.

As it turns out incumbent firms (established market leaders) often operate in environments where the architecture of the product is largely fixed. In some cases the incumbents were once the early innovators who were able to establish their architecture as the dominant architecture in their product category. Examples of such dominant designs include, IBM with the Personal Computer, and Apple Computer in music downloads. However, once the architecture is fixed the incumbent firms differentiate themselves by their ability to improve system performance by creating component level innovations. In doing so, incumbents organize their internal management structures, communication channels, information filters, and problem solving capabilities around tasks that enhance their mastery of the components and routing recurring incremental innovations.

As necessary as the above activities are for the incumbent vendors to sustain their market leadership, this R&D orientation on component innovation slows down their ability to identify, and hence react to, architectural innovations. Architectural innovation by definition places a premium on exploration of new design and assimilation of new knowledge. Incumbent firms have trouble making this transition as the process of discovery and building new capabilities usually takes time and commitment. The incumbents often try to explain the differences between old and new technologies within the concept of the current architecture and fail to master the new before it is too late.

Historical examples of firms unable to cope with architectural innovations abound, such as; failure of Xerox (pioneer of stand-alone photocopier) to create a competitive desktop copier product in response to Canon, In semiconductor lithography, the emergence of proximity aligner lithography vendors replacing contact aligner lithography vendors, emergence of Sony in the portable transistor radio market in US based on technology originally pioneered by RCA and so on..

Looking forward, those startups that are attacking their industries with architectural innovations have a better shot at success. The incumbents can ‘out resource’ the startups when it comes to component level innovations. A few examples of companies out there who are attacking their industries with a new architecture include Structured ASIC vendors eSilicon or eASIC in silicon prototyping, companies creating viable business models around open source software, companies that are providing different ways to distribute content (YouTube), and social networking (Myspace).

The author is a visiting lecturer at University of Illinois, Urbana-Champaign. He can be reached at sanjivc@uiuc.edu

Share on LinkedIn