The Open Industry Standard Wins
AFTER THE SYSTEMPRO ANNOUNCEMENT in late 1989, the outcome of the Micro Channel versus EISA war finally became clear. EISA, and with it the open industry standard, was going to win. It’s difficult to put an exact date on when it won. Most wars end with the signing of a treaty, but there was no proclamation about the winner of this one. However, there would be no more serious challenges to the reign of the open industry standard throughout the remainder of the PC era.
Few people understood the real significance of the final battle. Many were aware of the “bus war” between IBM and the rest of the PC industry. Many more were aware of the war between IBM and Compaq for technology leadership of the PC industry and, later, for outright market leadership. But these were surrogates for a much more fundamental and far-reaching conflict, the outcome of which would determine the competitive nature and structure of the PC industry for decades to come.
Some would argue that the open industry standard was destined to win from the beginning. That it was so powerful and made so much sense that it could not be defeated. That rationale, however, grossly underestimates the power of IBM as a company and as a brand.
Before 1981, the PC industry had consisted of a diverse group of rebels and upstarts. Its leaders created relatively strong brands and viewed themselves as completely disconnected from the rest of the computer industry. But as soon as IBM entered the market in August 1981, they learned what a really strong brand could do. IBM quickly became the market leader, even though it had introduced a product made from off-the-shelf parts and showed nothing special except a label with “IBM” on it. Yet it remained in the PC industry leadership role for the next thirteen years.
Some would argue that the open industry standard was destined to win from the beginning . . . That rationale, however, grossly underestimates the power of IBM as a company and as a brand.
Very likely, IBM would still be the PC industry leader today if it had fully understood the phenomenon that developed as a result of the popularity of its first product, and if it had chosen to lead the open industry standard instead of trying to end it. That would have been its strongest strategic move by far.
Most likely, though, IBM executives didn’t believe there really was an industry standard. As far as they were concerned, a bunch of clones had copied IBM products. To be fair, the entire computer industry viewed it that way at first, calling it the “IBM-compatible market” and even the “IBM standard.” By late 1983, however, I had introduced the term “industry standard” and later, “industry-standard architecture.” If IBM executives noticed at all, they probably viewed these terms as marketing hype or our wishful thinking.
As is often the case, IBM’s great success led to an overconfidence that caused it to miss what was really happening. IBM viewed the rise of the clone PC companies as the same result that had occurred in the past, when it failed to protect one of its products with enough proprietary technology. So IBM naturally concluded the solution was the same one it had successfully used every other time: come out with a replacement product that is more protected and quickly put an end to the clones.
Initially it positioned the PS/2 and Micro Channel so there could be absolutely no cloning. But fairly quickly, IBM backed off and said it would license Micro Channel to the clones for a license fee and a royalty of 5 percent of sales. Since many of the clones made less than that in profits, the effective result was the same as not allowing cloning. However, this way looked better, especially to the United States Justice Department.
IBM could have continued leading the industry long term if it had chosen to introduce something like EISA instead of Micro Channel. Compaq and the others would have had no choice but to follow, even if we had to license the 32-bit part of the bus. There would have been no “what IBM should have done” alternative for us to use against it. And adding a strategic move that made so much sense to its already incredible brand would have made it seem even more invincible.
But that was not the way IBM executives thought. They must have believed they could get all the clones to follow them to the Micro Channel; after all, the clones had done so with every major technology advance up until then, except one. IBM unintentionally had given Compaq the opportunity to introduce the first 386 PC ahead of it. When Compaq’s 386 PC became a great success, it clearly demonstrated there were many customers willing to follow a company other than IBM to a major new processor, as long as it was compatible with the industry standard. But instead of worrying about this inconvenient fact, they probably assumed that the PS/2 would crush Compaq’s 386 and prove that getting out in front of IBM was a mistake.
Even though introducing the proprietary—but incompatible—Micro Channel wasn’t their best strategic move, IBM still could have succeeded if it hadn’t let Compaq get ahead with the 386. Ironically, we wouldn’t have had the reputation and credibility to successfully organize and lead the EISA coalition if we hadn’t succeeded with the first 386 PC and demonstrated the importance of backward compatibility with the industry standard. In addition, if IBM had made it less costly for clones to license the Micro Channel by charging a more palatable royalty of, say, 1 percent, IBM would likely have lured the clones to follow it, thereby retaining tight control of the technology and, therefore, the industry. If it had put both of these together, IBM would almost certainly have succeeded in eliminating the open industry standard and continued its reign for a very long time.
When IBM first introduced the PS/2, almost everyone focused on the impact the computer would have on the clones. It was correctly believed that IBM was driven by a desire to reduce competition and regain the significant market share it had lost to the clones. What was generally overlooked was the potential impact on Intel and Microsoft if the PS/2 had become the new standard.
While IBM was jointly developing OS/2 with Microsoft, there is no doubt that its long-term goal was to completely control the operating system, effectively cutting Microsoft out of the picture. Without the cash flow from its position as sole supplier of the industry-standard operating system, Microsoft wouldn’t have been able to persist in pursuing—and eventually taking the lead in—several other key software market segments.
Intel’s position as the dominant supplier of processors for industry-standard PCs was threatened by virtue of it having granted IBM manufacturing rights to the 286 chip as part of an equity investment in the early ’80s. It was reported that IBM was planning to use its own 286 chips in the PS/2, but due to chip manufacturing problems, it had temporarily backed off. If it had succeeded in making the PS/2 the standard, it would have been in a position to gradually cut Intel out of the picture as well.
One plausible explanation for why IBM lagged a year behind Compaq in delivering a 386 PC was that it may not have planned on using the 386 at all. If the PS/2 had become the standard using its 286 chip, IBM could have easily introduced a proprietary next-generation processor instead of the 386. This would have completely cut Intel out of the picture and put IBM in the position of supplying processor chips to all Micro Channel licensees.
Once EISA was established as the industry-standard bus, the “Gang of Nine” took control of future changes in system architecture. With IBM no longer in a dominant position, Intel alone controlled advances in the microprocessor; it no longer had to worry whether IBM would change to a different processor in its next product. And so long as each new Intel processor was compatible with the previous one, it would automatically be accepted as the new industry-standard microprocessor. The same was true for Microsoft and its operating systems. As a direct result of our success in defeating IBM’s Micro Channel, each company gained unfettered monopoly positions in the industry standard for the long term.
Compaq certainly gained a lot as well. But when we succeeded in keeping the playing field level by saving the open industry standard, we ensured that we would have to play on that level playing field too. We knew how to do that better than anyone, and as a result continued to thrive. Compaq’s management team had grown up in the industry-standard environment. We had matured and gelled as a team, and in my opinion were one of the best in the industry.
Compaq Management team, early 1990. Clockwise from left: Daryl White, Jim Eckhart, Bill Fargo, Murray Francois, Bob View, John Gribi, Kevin Ellington, Gary Stimac, Jim Harris, Mike Swavely, Ben Rosen, Rod Canion, and Eckhard Pfeiffer.
AFTER TEN YEARS AT THE HELM, I left Compaq at the end of 1991. Parting was bittersweet. In truth I was somewhat burned out by the intense, nonstop pace of those ten years, especially during 1991. That year we were caught off guard by an economic recession, by six of our top ten dealers merging into three, and by an unexpected strengthening of the dollar. The result was Compaq’s first quarterly loss since 1983 and our first layoffs ever. Laying off hundreds of dedicated people weighed very heavily on me.
Jim left Compaq at the same time. Jim had led the design of the original Portable and built an engineering team that was second to none. He established the design discipline in Compaq that was fundamental to achieving and maintaining our reputation for quality and ruggedness. Jim was also my sounding board and advisor throughout the entire ten years.
But before we left we had to deal with the question of strategy. I took the strategy team off-site several times during the summer and early fall of 1991 to carefully analyze our strengths and weaknesses, and decide if and how we needed to change. We decided we could no longer stay with just high-end products and prices. It became clear that we could leverage one of our most valuable assets, our strong upscale brand, to help enter the low end of the market. So in late summer we started a “crash” project to design a low-cost computer that still had differentiating features.
Eckhard Pfeiffer took over as CEO when I left and led Compaq to launch the low-price product line in June 1992. He and his team did an amazing job of taking the basic, low-cost design from the “crash” program and parlaying it into an incredible total of sixteen new products. The new machines included: low-priced desktop PCs, Compaq ProLinea; low-priced notebook PCs, Contura; and upgradable desktop PCs with advanced graphics and audio capabilities.
The announcement sent shock waves through the industry worldwide. No one expected Compaq to make such an aggressive move and to have such aggressive pricing. The strength of the low-end product launch shifted momentum back to Compaq, and its sales growth accelerated again. Revenues nearly tripled over the next three years, growing from $4.1 billion in 1992 to $10.8 billion in 1994. In that pivotal year, Compaq became the leading supplier of PCs worldwide, while IBM quietly announced that it would no longer be producing the PS/2.
As the tides ebbed and flowed, in 2001 Compaq achieved sales of over $33 billion and merged with its old ally from the PC wars, Hewlett-Packard. Another economic recession had come along and both companies were struggling. The little company that wouldn’t be pushed around by IBM had run its course, as the combined company took the Hewlett-Packard name. The Compaq brand today lives on in a somewhat subdued role as HP’s consumer PC product line.
Compaq’s legacy also lives on in its impact on the PC industry and all our daily lives. The many advantages of the industry standard are still with us, giving us broad choices of PC features and brands at incredibly low prices. Just as importantly, if Compaq hadn’t succeeded in stopping IBM and preventing Intel’s loss of its monopoly position in processors, Intel wouldn’t have earned such enormous profits through the ’90s and, as a result, wouldn’t have been able to advance chip and processor technology as rapidly. If that hadn’t happened, we almost certainly would not have the technology we have today that drives the smartphones and tablets we all enjoy.