Intel (Nasdaq: INTC) and Microsoft (Nasdaq: MSFT) share a similar uncertainty about their futures. Facing shrinking margins due to a gradual market shift toward smaller and cheaper "Internet devices," both have reacted by diversifying and trying to embrace the new networked market. If they can succeed, the higher volumes could easily offset the lower margins. Even though they're paying attention to the low-margin end, it may not be enough. The temptation to pursue high margins could easily leave them vulnerable to players who capture the higher volume and the revenue stream that goes with it.

The PC industry has always been defined by its commodity hardware. Multiple vendors construct equivalent packages of interchangeable parts, competing not based on the exclusivity of their solution, but on price and performance. A few vendors have carved out exclusive niches for themselves with Creative Technology's (Nasdaq: CREAF) Sound Blaster card and Iomega's (NYSE: IOM) Zip drive, but these domains tend to be transitory things, eroding with time. In each instance, their lasting advantage was more a matter of brand name than technology. In the end, Moore's Law commoditizes everything.

The two kings of the PC domain have been Intel and Microsoft. Intel has lived the reality of commodity hardware, and stayed on top primarily because it took advantage of its economies of scale, as the largest microchip manufacturer, to bring down price (by manufacturing in volume) and increase performance (by funding research and development) faster than any other vendor. Intel stayed on top of a commodity market by playing the commodity game better than anybody else, with a constantly improving product and a strong brand name.

Microsoft was the one company in this space with a proprietary advantage, its copyrighted software made from secret source code more closely guarded than Coca-Cola's (NYSE: KO) 7-X formula or Colonel Sanders' 11 herbs and spices.

When IBM (NYSE: IBM) put together the PC, it understood hardware monopolies intimately and ensured it would never be on the receiving end of one, by selecting only commodity hardware available from multiple vendors to go in its PC. But, IBM didn't understand the software side of things well enough to do the same there (although it tried, it thought Digital Research's CP/M and Microsoft's DOS were close enough to be interchanged), and Microsoft snared a monopoly that has defined its existence (and the software side of the PC) ever since.

The Internet threw both companies for a loop by changing the main use of the PC from a tool for running ever-more-power-hungry programs to a communications device. The incentive to buy twice as much hardware, to run twice as much software, quickly evaporated. All you needed was enough power to run a Web browser. The vast majority of new websites are still viewable on a five-year-old Pentium running a five-year-old copy of Windows.

Soon, Intel was forced to pull its attention away from the high end to put out the Celeron, an extra-cheap processor to satisfy the ravenous demand at the low end that was feeding revenue to its main competitor, long-dismissed Advanced Micro Devices (NYSE: AMD).

Microsoft didn't immediately notice its own rival, Linux, creeping up from the low end, but it did notice that once people had Windows 95, there was little incentive for them to upgrade to Windows 98 or beyond, as long as they could get new versions of Netscape. It fought viciously against Netscape, but won a Pyrrhic victory when it wound up in federal court on antitrust charges.

Just as PC volume eventually rendered mainframes obsolete, Internet volume is attacking margins in the PC industry. PCs under $1,000 are now the norm, and models under $500 are increasingly common. Windows 95 sold for almost $100. Windows Millennium (I can just see Bill Gates with his pinkie in his mouth: "I shall call it... Winni-ME!") sells for around $60. The price decline is not due to Microsoft wanting less revenue, but because it's under severe price pressure from its own established base. Linux aside, the main incentive is not to switch software at all.

Figuring that "if you can't beat 'em, join 'em," both Microsoft and Intel have decided to profit from the Internet. Intel is trying desperately to become the world's largest Internet service provider via its Global Web Hosting Services initiative, where it competes head-on with everyone from IBM to AT&T (NYSE: T). Intel is also building on its traditional Ethernet offerings to take on Cisco Systems (Nasdaq: CSCO) in the network connectivity arena with voice and data digital subscriber line (DSL) hardware and software.

Microsoft is trying to take on the Web, repeatedly naming Sun Microsystems (Nasdaq: SUNW) and Oracle (Nasdaq: ORCL) as competitors (and Intel as a partner) on the e-commerce page. And, its new initiative, within which it insists virtually all its future software development will occur, is centered on a language called C# (pronounced C-Sharp), which could be summed up by the slogan "because we didn't invent Java."

So, to find future growth, Intel is pushing against a half-dozen established companies (not even counting AMD nibbling at its core market), and Microsoft is pushing against dozens more (not counting Linux nibbling away at its core market, or side markets such as the gaming industry where Nintendo and Sony (NYSE: SNE) square off against Microsoft's X-Box).

The market as a whole is expanding, so there is room for more winners than losers, but it's still an uphill battle. Meanwhile, the temptation to focus on the highest-margin areas (such as the high-margin monster network servers powered by Xeons, running Windows 2000) causes both companies to waste effort on market niches with diminishing volume (as those monster servers are replaced by server clusters). Read The Innovator's Dilemma for an in-depth examination of this temptation.

The whole thing reminds me why I like Coke, which can keep doing the same thing over and over, year after year, and remain profitable without having to expand out of its niche. It may have some internal troubles each time upper management changes, but it doesn't have to find a new business model to survive and prosper.

Coke's not very exciting, and doesn't have the upside potential of its more active and flashier cousins. But to me, Rule Makers are the solid foundation on which to build the rest of the portfolio. I'm keeping my Coke shares, they help me sleep at night while my Intel shares are off to war.