Requiem for a Business Model
Open Source: You say goodbye and I say hello
Many years ago C. Gordon Bell, perhaps the foremost student of the computing industry, drafted a paper with a working title of something like “Microprocessors: Any Winning Avenues Left?”. In it he analyzed the trajectories of broad categories of then current technologies including cpu, memory, operating systems, etc and came to the conclusion, and I paraphrase here, that the then current generation of mini-computers was likely doomed. Quite the revelation since most of the players in this industry were still enjoying double-digit growth at the time. Now try as I might I haven’t been able to locate this draft but most of the concepts were published in an article that later appeared in Computer entitled “The Mini and Micro Industries”.
If you consider yourself to be a student of strategy and the computer industry and haven’t read this article, then you are in for a treat. Not because of Mr. Bell’s breathtaking grasp of the modern computer but rather his dazzling comprehension of the factors that determine success and/or failure of innovation within the context of prevailing business, market and economic circumstances. His concepts remain useful to this day in predicting what will and what won’t work when it comes to technology innovation, not to be confused with the use of computing in social or media contexts which is more a function of fashion innovation than technology innovation.
Way back when, Mr. Bell introduced some extremely valuable concepts that were probably not well understood at the time. For instance, he introduced the notion of economy of innovation, roughly the idea that rivals could not by themselves produce a comprehensive, all encompassing, special purpose architecture, to create a stand-alone solution for a specific, narrow problem. It was simply too expensive to advance a special purpose architecture to solve any given problem and that the market for such solutions were ultimately too small to support the necessary business infrastructure to make such companies successful. The exception to this would obviously turn out to be things like a Cray machine where every instance was basically built from scratch to do a few very hard things. This in turn led to the thesis that general-purpose architectures would be more successful than special purpose architectures because it was economically more advantageous to augment a general-purpose architecture with specific applications to solve a given problem than design a specific architecture, which could not address other opportunities in the market.
This same idea, economy of innovation, he also applied to existing and latent or proto-markets, existing markets being those whose value proposition was established and latent markets being those which could be addressed if the cost of computing could be sufficiently lowered to make fashioning solutions economically viable (i.e. profitable). This is where the microprocessor would win the field over the previous generation of component-based mini’s and for that matter mainframes. It would also reinforce the notion of specialization within the industry as software, component and value added resellers became instrumental in attacking these newer markets.
Now we have to remember where the computing industry was in its lifecycle to fully appreciate the value of the concept when it was first put forward. At the time there were a number of outfits attempting to bring to market special purpose solutions for very specific problems. Take Wang for instance, the maker of a word processing computer, quickly overwhelmed by a lower cost general purpose device with a software application added to it. Or Adage, an early instance of wire-frame, real-time graphics for use in CAD/CAM. We’ll come back to revisit this concept in a bit as things here have just recently begun to change.
Now, an interesting experiment would be to apply Mr. Bell’s concept of economy of innovation to the idea of say, open source software (OSS), to see whether or not it could still have been used to predicted an industry wide outcome. Both with respect to a stand-alone value proposition as well as a contributor to the creation of special purpose architectures.
First, open source software, leaving off the evangelical non-sense for a minute, had many of the attributes inherent in those technologies that exhibit an economy of innovation. As a competing substitute to commercially proven incumbent solutions it had a lot going for it. It was by its very nature produced by a community of coders and therefore cost and sustainability weren’t significant factors inhibiting its adoption. It was for the most part general purpose in nature, generic versions of proven broad categories of software solutions. As a commercial offering, it was backed by substantial players who would eliminate most of the risk of employing it in the enterprise. And yet when you look at OSS’s success at an enterprise or for that matter consumer computing offering it is little more than a footnote in the context of the broader industry and likely to stay that way.
What happened?
Well for starters, despite certain promising attributes, OSS in the enterprise and consumer computing markets lacked one key ingredient: the ability to generate new value by virtue of addressing heretofore unaddressed market segments. Without that OSS was pretty much dead on arrival. Why? You simply cannot create the industry ecosystem needed to make OSS successful unless you are truly addressing new markets that provide new value for customers. Yes, Redhat will continue to enjoy success so long as rivals simply don’t use their pricing power to limit the amount of share they can capture. But with Oracle’s acquisition of SUN how far off can that be?
So where has OSS demonstrated an economy of innovation that could make it a viable success?
Well, that’s obvious, in the creation of special purpose devices a la Android that attack new markets, mobile devices, previously unavailable to the rivals who seek to capture these customers. The only rub is that many of the incumbents of the current generation of enterprise and consumer computing can’t make the numbers work and therefore won’t be making this trip. Microsoft has been struggling to take a general-purpose operating system and stuff it in a tablet or cell phone or toaster for years and it just won’t work. Neither will its economics. Fifteen dollars a copy for a smart phone operating system is not going anywhere anytime soon. Hence, the recent announcement that Microsoft will be porting its software to ARM, special purpose, mobile device architectures. But why bother when you can simply use open source software?
So, is there a prediction out there waiting to be made using the ideas of economy of innovation?
You betcha. Thanks, Mr. Bell.