GAITHERSBURG, MD (Nov. 24, 1999) -- On Friday, I reviewed The Innovator's Dilemma, a book about disruptive technologies that encourage start-up companies to replace large established firms. On Monday, I talked about "use value," which is discussed in greater detail in Eric Raymond's book The Cathederal and the Bazaar (which I'll be talking about some more today). Yesterday, I explained how network effects can create natural monopolies, and discussed the first way I know of to compete with this type of monopoly: participating within the existing network to share the market.

A disruptive technology can also crack a monopoly, by replacing the existing network with a new and larger one. A larger network will consume a smaller network operating in the same space, because the value of the network increases exponentially with the size of the network. But a disruptive technology doesn't (initially) operate in the same space; it starts by finding a new market with wants and needs that the old technology cannot meet.

Because a disruptive technology can't initially compete directly against the existing technology for the original uses even without taking the network effects into consideration, the only successful way to market one is to carve out a new "roaches under the floorboards" niche taking advantage of the new technology's distinctive advantages. This new niche is a market the existing technology can't fit into, and is usually too small and low margin for the existing companies to care about anyway. But it gives the disruptive technology a protected base within which to develop, and which can finance an upscale attack into the existing value network. Since the disruptive technology can eventually be made to serve the same purposes as the old technology, and it has its protected base to draw from as well, it forms a larger network than the old technology.

The disruptive technology currently attacking Microsoft (Nasdaq: MSFT) is the Open Source development model. Eric Raymond's classic paper "The Cathedral and the Bazaar" has recently been published by O'Reilly as part of a book by the same name. It explains how software can be developed not by a team of highly trained professionals, but by a loose association of hobbyists organized through the Internet as a kind of fan club. It also discusses why this software development model often beats the pants off of the old way of writing software.

While traditional software development is motivated by the sale value of that software, Open Source development is motivated by the software's use value. People who use the Apache Web server naturally want it to be the best Web server it can be. So, if they are capable of improving it, many of them will -- in their spare time or when they encounter a problem they need to fix. All that time spent on hold to tech support for proprietary software companies is spent actually finding and fixing the bugs and limitations of Open Source software. The availability of the source code to Apache gives the technically inclined individuals and organizations using Apache the ability to do whatever they like with it.

Sharing the improvements through the fan club doesn't cost Apache's users anything, since they didn't plan on selling those improvements anyway. They just want a good Web server. The fact that it's a free download is merely a bonus. What they really get is control over the tools they use to run their business -- high quality tools created by people motivated by how well they work and not how well they sell.

Apache is the server behind the majority of the websites on the Internet. In that niche, Open Source is the dominant paradigm. To examine Open Source as a disruptive technology, we need to look at a clean room clone of Unix that started as the Free Software Foundation's GNU project and was completed (well, first made independently usable; Open Source development and improvement never ends) by a graduate student at the University of Helsinki in Finland named "Linus Torvads." You may have heard of it; it's called Linux.

Linux started out in a "roaches under the floorboards" niche. People who couldn't afford what AT&T and others charged for Unix but wanted to use it anyway reverse engineered the thing and made it run on PC hardware. Then, Linux became popular to turn old, discarded, underpowered PCs into file servers and print servers. As the Internet (created by and full of Unix machines) became popular, this reverse engineered Unix clone was a cheap way to use a PC as a Web server, gateway, or firewall to hook up to the Internet.

Microsoft's high end is the desktop, and that's the last market Linux will take over. From the low-end machines that other operating systems were too inefficient or expensive to use on, Linux jumped straight to high-end multi-processor machines. As the Web became important and expensive hardware was thrown at it, Linux was moved over to that expensive hardware where the lean and mean efficiency required to run at all on the cheap junk was just as valuable to squeeze every drop of performance out of the expensive stuff.

Again, this was fringe territory as far as Microsoft was concerned. Although they have tried to push Windows NT on the high-end for almost ten years, it has never shipped in volumes even approaching that of its desktop home ground. Microsoft may have been unwilling to yield any niche, but Linux definitely attacked where it was weakest.

Linux has also had success with embedded systems, and has virtually driven Windows CE from the field. Microsoft may have unlimited funds with which to develop their products, but Open Source products simply cannot be starved for cash. Their development is funded with their users' time and effort, not with revenues from any sale. Use value again, not sale value.

If you remember FUD from yesterday, Linux is un-FUDable. As long as the users have the source code, development will continue. And how can it go out of business if it never made any money? Companies make money off of Linux, of course. Red Hat's (Nasdaq: RHAT) IPO shot to a multi-billion dollar market capitalization, and hundreds of established companies like IBM and SGI are betting heavily on its success. But like dating services, they seek to cash in on an existing phenomena. They dip their mill wheel in a river that's already there, and could go on without them just fine.

Like all companies performing an upward retreat (however grudgingly) in the face of a disruptive technology, Microsoft's profits will probably continue to increase right up until the end as it tightens its focus on the areas that make it the most money. The desktop is Microsoft's home ground, its strongest niche, and where it makes the most money. It will be the last niche Linux penetrates, and the one Microsoft will fight hardest to keep.

According to the Gartner Group's estimate, Linux passed the 10 million user mark during 1998 and was growing at 212% annually at the time. Assuming it slows to 100% annual growth (which it has maintained since its introduction in 1990), it will surpass the Windows installed base in about three years. At that point, the network effects will favor Linux and hinder Windows.

Strangely enough, many of us in the software industry view the trial as an interesting supplement to the main action. Microsoft IS an abusive monopoly, and unrestrained perhaps it could force the entire computer industry into stagnation. But we're busy working on its replacement.

- Oak

What do you think?
Come post your thoughts about tonight's report on our Rule Maker Strategy board. Alternatively, if you have ideas, analysis, or questions about a particular Rule Maker company, take a seat at the roundtable discussion on our Rule Maker Companies board. Finally, if you're new to all this stuff, the Rule Maker Beginners board is the place to get your questions answered.