Twenty years ago, IBM owned the computer industry. Then a disruptive little product called the Apple II came along, putting spreadsheets and word processing in the hands of individuals. It was small, it was cheap, it fit on a desk, and it sold like hotcakes.
IBM knew a threat to its monopoly when it saw one, and decided to kill the Apple II by putting out a personal computer of its own. Unfortunately, an audit of IBM's internal bureaucracy had revealed that IBM required nine months to ship an empty box, and IBM's PC offering had to be out within a year to have any chance of derailing Apple's momentum.
The solution was to outsource absolutely EVERYTHING, and assemble a product out of commodity off-the-shelf components available from third parties. Following standard procurement guidelines, IBM's engineers used only components available in compatible versions from multiple vendors, from disk drives to memory to the microprocessor at the heart of the machine. This is how Advanced Micro Devices got started cloning Intel processors, by the way: Intel had to license its design to another manufacturer in order to get the IBM contract. IBM had no intention of allowing any of its hardware suppliers to gain a monopoly that allowed the supplier to extract the kind of margins from IBM that IBM was used to receiving from its own customers.
The IBM PC backfired on IBM in at least three ways. First, IBM's entry into the PC market legitimized the personal computer as a business tool, immediately driving sales of Apple's product to record levels. Secondly, all a competitor had to do to clone IBM's PC was buy the same components from the same third-party suppliers IBM had used and reverse engineer IBM's one unique contribution (the BIOS chip). Compaq was the first to do this, quickly followed by a horde of imitators that made the PC what it is today.
Let me repeat that. The PC was successful because it was cloned. The clones were what drove Apple to the fringes of the marketplace, and they did exactly the same thing to IBM's PC offerings. Proprietary products are simply out-evolved in the Darwinian jungle of a multi-vendor commodity marketplace. The lumbering elephants with fat margins don't do very well against lean and hungry companies that do without sleep and slash prices to offer more, faster, and cheaper. The only way to stay ahead of the competition each day is to offer customers a better deal, and to improve that deal faster than the competition can. Commoditization isn't anti-capitalistic, as most monopolists would have you believe. It's the inevitable result of unfettered capitalism, what the free markets inherently strive for at all times. Commodity products are what customers WANT, that way they have a choice of suppliers.
IBM lost its monopoly on computer hardware because computer hardware became a commodity. Not only were the parts from different manufacturers interchangeable, but the finished PCs assembled by the different manufacturers were interchangeable as well. Instead of IBM swallowing up the PC, the PC swallowed up IBM, and eventually even IBM's high-end mainframe business came under attack by those little boxes under people's desks. That IBM even survived the loss of its monopoly (and the resulting collapse of its stock in 1992) is a testament to the strength of the company. These days, IBM is quite definitely "under new management."
I mentioned a third mistake IBM made with the PC, and it was a very simple one. They handed Microsoft a monopoly on a silver platter. IBM was a hardware company; it never even occurred to them that the software might actually be important. They were excruciatingly careful about all the hardware they used, but they didn't bother to get a second source for their operating system. (They had a second, incompatible operating system available for the PC, but that's not the same thing.) Microsoft's monopoly exists because IBM screwed up.
Since my space is limited, I'm going to skip over the 20 years Microsoft enjoyed a monopoly and get to the commoditization. The Internet commoditized data transmission the way the PC had commoditized computing hardware. Software is just data, the cost of distribution of software suddenly, literally, dropped to zero. Like IBM, Microsoft tried to consume the threat, by monopolizing access to the Internet with a proprietary Web browser. But like IBM's experience with the PC, Microsoft bit off more than it could chew.
Bill Gates' book The Road Ahead didn't even mention the Internet, but after Netscape's IPO Microsoft caught on fast. They decided that if they could control the Web browser people used to access the 'Net, they could own the 'Net.
But unlike the trail of Windows software companies that Microsoft had consumed or destroyed in order to grow faster than the PC market as a whole, the Internet had come from the Unix world. It wasn't inherently dependent on Microsoft, and it didn't want to become so. To put it bluntly, the Internet community counter-attacked by intentionally commoditizing the software side of the PC that IBM missed long ago. The Internet community created gigabytes of software, including versions of each of Microsoft's core products, and gave away the source code. For free.
Thousands of individuals co-ordinating their efforts through the Internet have reverse-engineered Microsoft's core products and put up the results on FTP sites for free download. The most direct response to Microsoft's attempt to claim the network clients was to lock up the network servers those clients spoke to. The open source "Apache" Web server now runs the majority of all websites, forcing Microsoft to adhere to standard network protocols despite its strangulation of development funding for Netscape's Web browser.
Netscape, which also emerged from the Internet community, understood the power of commoditization and opened the source code to its own Web browser in response to Microsoft's attack on it. As a result, the company -- once written off as dead due the monopoly leveraging that got Microsoft in trouble with the Department of Justice -- was still worth billions when purchased by AOL. The decline of Netscape's market share slowed even before the release of the result of this collaborative development, the Netscape 6 browser (beta-test versions of which are now available).
But the main thrust of the Internet's counter-attack was at the heart of the operating system monopoly IBM originally gave Microsoft. The Unix-based Internet consolidated the previously fragmented Unix market with an open source version of Unix, called "Linux." The first thing they did with Linux was install it on all their servers, to inoculate them against the spread of Windows.
Once the servers were secured (with open source Linux and Apache), the next attack was on Microsoft's future expansion opportunities. From "Beowulf" class supercomputers to embedded devices, Linux has rapidly spread almost everywhere Microsoft wasn't already established. As with all disruptive technologies, the attack on the existing technology's core business comes last, although with Michael Dell delivering the opening keynote speech at LinuxWorld Expo this summer, it can't be that far away.
Linux is out-evolving Windows the exact same way the PC out-evolved IBM's mainframes. Sure, there are some things Linux doesn't do as well as Windows yet. But improvements in Linux are measured in months, whereas new versions of Windows take years. As time goes on, the commodities simply develop faster.
The parallels run deep. Just as a PC is available from Dell, Compaq, Gateway, HP, or a thousand other companies, Linux is available from Red Hat, SuSE, TurboLinux, Mandrake, and a thousand other companies and websites. And just like a Dell box can be upgraded with a hard drive from Gateway, a copy of Linux installed from one of Red Hat's CDs can be upgraded with a package installed from a TurboLinux CD. Commodity, interchangeable parts. What Linux distributors like Red Hat are really selling is a service, which is exactly what PC makers like Dell sell if you think about it.
The irony of the situation is that the previous monopolist, IBM, is now one of the leading supporters of Linux. You want free software on commodity hardware, IBM is happy to sell it to you. They know the real money in a commodity marketplace is in service and support, and ongoing relationships with customers. That's where they intend to make a bundle.
Microsoft is free to jump on board the open source movement. But as IBM proved with the PC, embracing change without changing yourself just doesn't work. And the most interesting difference between IBM and Microsoft is that IBM earned its monopoly in the first place.
This Week's Duel
Related Links