JDSU's Critical Mass of Networked Scientists
I want to thank Apollo for his comments and questions. Correctly noting that one premise of my argument was that JDSU, through acquisitions, has achieved critical mass through owning a proprietary network of applied scientists and engineers. Apollo adds that no company owns its talent, which can leave for startups or companies offering more rewards. He asks, "How does one measure the strength and totality, as it relates to the total intellectual talent available?" He is not satisfied simply with measuring the pace of innovation in modular development, as this is very indirect surrogate for his proposed ratio of JDSU talent to total talent.
Both Apollo's criticism and question deserve answers
The heart of the Gorilla Game is Moore's (RFD, p. 52) phrase, "proprietary open architectures with high switching costs." A December series of posts on this thread about JDSU and a recent spate of posts over the weekend have been directed to the issue of what Moore means by "proprietary." First, there is universal acceptance that patents are proprietary; but what about "know how?" Second, a crucial distinction has been made that a gorilla's proprietary power is manifested through its control of the value-chain. That control is exercised by selecting the high margin plum cakes for itself, leaving the crumbs to weaker members in the value-chain, and disrupting competitors by changing its "standardized" product, changes that eliminate competitors or make them waste resources. Third, this power is given to the Gorilla because the excess demand in the mass market requires standardization on the Gorilla's solution. So, at present, we have a one-member class of "proprietary," (which I believe needs expanding,) and an observable effect of manifested gorilla power in the disruption of competitors that FOLLOWS from a mass-market-demand-based standardization on the Gorilla's product. Notice that I just wrote "Gorilla's product," deliberately leaving out the modifier, "proprietary" and I capitalized "follows" to emphasize its temporal precedence.
In my post 27995, I introduced a distinction between Proprietary Power and Innovative Power to facilitate a graded analysis of "power" by suggesting the usefulness of "degrees of power." (Simultaneously, Merlin was posting about the importance of "degrees of power" among Gorillas.) If we can make some useful distinctions among degrees of power, we can better calibrate degrees of Gorillahood or Kinghood. Staying within the spirit of the Gorilla Game, we are trying to move from a "category" to a "continuum" of power, to find variants of power other than proprietary that permit some degree of control over a value chain. If we succeed we can begin to calibrate degrees of power in a way that widens the potential field of Gorilla's, or, if you prefer, distinguishes among Kings, or given your acceptance, among Kingillas and Gorkings.
Is there a critical mass of scientists?
Although it would be interesting to know, I don't believe computing a ratio of JDSU's scientists divided by total size of the scientific pool is required by my argument. I contend that JDSU has acquired a critical mass of scientists/engineers that will produce continuing innovation in components and, better yet, modules of increasing complexity.
In this instance, a critical mass produces an explosion, to ride the metaphor for all its worth, of creative innovation that its competitors cannot equal and which produces sustainable innovative power. The Laws of Increasing Returns and of Accelerating Returns introduced by Brian Arthur and Ray Kurzweil, respectively, are the grounds for the claim. The explosion is triggered by the cross-fertilization of ideas across sciences and scientists. Because JDSU has many diverse products drawn from diverse sciences, and a critical mass of scientist, the pace of innovation increases at an accelerating rate. Innovation begets more innovation. Building on prior innovations with innovative tools, recursively, creates new order and accelerate returns. This acceleration is not only a function of order building on order but also of the network effect of communicating scientists, who, before the acquisition, were previously not in open communication. It only takes one new scientist to add one new idea or new skill to solve a previously intractable problem. Creative Wisdom emerges from many informed people who communicate as a network, like this board, like all science, like democracy.
Recall the progress in SiC LED brightness that Cree made when it acquired Nitres' knowledge (know-how) of GaN epitaxy. Now increase the scale across multiple materials times multiple products times many scientists times multiple sciences. Modularity in itself requires the sharing of ideas among diverse specialties across disciplines. Such a sharing of ideas, coupled with an R&D budget that doubles in absolute size each year, surely produces creative movement up the multiple learning curves required in this field that is both younger, with much more learning yet to come, and far more complex than semiconductor manufacturing.
Acquisitions keep a company on the edge, between stultifying order and chaos. Most mutations and innovative ideas come from the fringes of an ecological niche in evolution and from the fringe of start-ups in technology. Uniphase was selling helium neon laser for reading portable bar codes before Kevin Kalkhoven envisaged a new Uniphase. That bold idea incited three key acquisitions that began the cornering of scientific talent. In 1994, Uniphase acquired United Technology's Photonic Research Division; in 1997, it acquired IBM's laser operation; and in 1998, it acquired Philip's optoelectronic group. The merger with JDSU Fitel brought together the largest producer of active with the largest producer of passive optical products. Called a crown jewel, OCLI manufactured more than 80% of the highly complex thin film filters. ETEK added specific packaging expertise and a relationship with Bookham. Chronos, with its MEMS, was called the most important acquisition yet. Several smaller specialty firms were also added within the last year for their technologies and innovative scientists. SDLI, who acquired three firms this year, when added to JDSU strengthens the amplifier program, adding 1700 more people to the network of potential creativity.
Currently, JDSU has 17,000 employees in 30 locations around the world. Last November, Kalkhoven said they had 700 scientists. When JDSU was born it had a portfolio of over 250 patents. Talented scientists and engineers are hired not only to implement solutions but also to create them. When you have a field that is developing as rapidly as DWDM, you stay on the edge of new knowledge. You want people that know how to learn from experience so they can learn how to take innovation to the next level.
Most new scientists/engineers will want to work at JDSU because more exciting things are happening there, and the stock price is appreciating, lifting the value of their options.
It is true that entrepreneurial scientists can leave to start a company. To successfully attract venture capital, they must have an idea that can generate an order of magnitude solution, at least 10 time faster, better, cheaper.. If they leave, however, I'd bet they would hope that their company would be acquired by JDSU, just as SDLI preferred them to GLW. Time to exit from venture capital financing is also accelerating.
Start-ups, yes; true competitors, no! It is too late to start a new company from scratch that can compete with the depth and breadth of JDSU. You can't find the expertise, work around the patents, take the testing time to develop products that meet high availability standards, and the like. They already have a lock on breadth of components that insures a lock on modules. Any new company seeking to equal JDSU's breadth and depth of components would crawl while JDSU flies ahead. Of course, you cannot duplicate the modular strategy if you do not have the components or the scientists to integrate them.
Demand-Based Standardization
Finally, I argue that standardizing on products during the Tornado is demand-based, a demand based on pragmatist adoption. The Internet is the mother-of-all-tornados as well as the stimulus for the mother-of-all-value-chains. Standardizing on a product fills a role in the value-chain. A controlling architecture can be an emergent form of power. On the one hand, QCOM had the patents before they became a Gorilla. They had to demonstrate the effectiveness of CDMA to build demand. On the other hand, demand-based standardizing on a product could also lead to developing an architecture that may become increasingly powerful and, perhaps, proprietary. That is the scenario, I outlined earlier as potentially following from JDSU's modular strategy. In complex systems like tornado-markets, causality is not necessarily linear nor ordered as simply as billiard-ball cause and effect.
I hope this helps.
Don |
