Opportunity Knocks
IN THE SPRING OF 1983, I drove to work along state highway 149, a two-lane blacktop deep in dense East Texas forest. Twenty-five miles from downtown Houston, just where the isolated highway crosses a sandy, flood-prone creek, a shiny black office building would appear suddenly out of the thick pines and my heart would race a little. It was Compaq’s newly leased headquarters.
With hardly more than a hundred people, we needed only part of the modest building for our office staff, while others remained in our factory nearby. The atmosphere was electric, intense, and fast-paced. Nonstop discussions, debates, and on-the-fly decision making spilled from offices to halls to the parking garage. It had been only a few weeks since the first shipments of our portable, and it seemed to be taking off more strongly than even our most optimistic estimates had projected.
We were anticipating that IBM would soon bring out a hard-drive version of its PC, and on March 8 the announcement came about the IBM XT. We knew all our competitors would be racing to market with their own hard-drive products as quickly as they could. Our product’s portability made the technology issues trickier because these small hard drives, only 5¼ inches wide, were very fragile and could not survive the bumps and shocks a portable computer would surely experience. Since a core feature of Compaq’s market positioning was the exceptional ruggedness and reliability of our portable, meeting the high standards we had set was going to be a serious challenge.
MARCH 28, 1983, 9:00 A.M.
Jim, Bill, and I are well into one of our almost daily meetings. This one’s in Bill’s sixth-floor office with a view of the dark green forest below. We’re deep in conversation about our next product, our response to the IBM XT.
Jim says, “We’ve come up with a solution to the hard-drive ruggedness problem, but there are some trade-offs involved. Instead of a 5¼ inch-wide hard drive, we can use the newer 3½-inch drive and shock-mount it inside a metal enclosure that is the size of a 5¼-inch drive.”
I ask, “What trade-offs?”
Jim replies, “The smaller drives aren’t in production yet, so we’ll have to wait awhile before we can get enough units to begin shipments of our product. Also, we need to do more reliability testing to be sure it’s ready for production. And they cost more.”
Bill frowns. “We really can’t afford to be too far behind our competitors or the dealers will start carrying some of them instead of us. And our price needs to be in line with IBM’s. Isn’t there any way to make the 5¼-inch drive work?”
Jim grins. “Sure, if you don’t mind waiting a year while we design a bigger case.”
I say, “I’m particularly concerned about a reliability problem popping up because it’s so new. What do we know about the manufacturer?”
Jim answers, “The name is Rodime. It’s a small company based in Scotland, but it’s been building 5¼-inch drives for a while.”
We continue discussing the issue for a good while. There is no straightforward solution. Running out of time, I finally say, “I don’t like taking such a risky path, but I don’t see any better alternative. We’ve got to have a hard-drive version of the portable, and there’s no way to make it rugged without shock mounting a 3½-inch drive. But we have to be sure the product is reliable before we ship it, even if it means delays.” On that point we all agree.
“Shock mounting” a hard drive, using rubber spacers to absorb energy when the unit is bumped, seems extreme today. The three of us had experience working with the small Winchester drives from our days at TI and knew just how fragile they were. We were also very serious about maintaining our reputation for ruggedness and reliability, because we knew that such a reputation was hard to achieve—but easy to lose.
A few weeks later, the three of us met on an issue we had not anticipated. While going through the reverse-engineering process with the new XT, our engineers found that it would not run all the IBM PC’s software used in tests for our own compatibility. While it ran most of the software correctly, the fact that any of the programs had a problem was surprising.
It was hard to understand how IBM could have allowed this to happen. Adding a hard drive to its PC was a straightforward engineering task that should not have required significant changes. It seemed IBM simply didn’t place a high priority on testing the XT for backward compatibility with its PC. That meant some of its customers who upgraded from a PC to an XT would have to buy newer versions of some software because the old ones wouldn’t work. Free updates didn’t exist at that time.
We briefly considered whether we should duplicate the incompatibilities of the XT, but quickly concluded that didn’t make sense. Customers would want to be able to use their existing software when they bought our hard-drive product. We used our compatibility technology to enable ours to run all the XT software, as well as all the PC software.
OCTOBER 25, 1983, 11:00 A.M.
With less than six months of development and not quite a year since introducing our first product, I’m standing behind a podium at Tavern on the Green in New York City’s Central Park. A red cloth is covering our new computer as it sits on a pedestal next to me. After making a few introductory remarks to the assembled press and industry analysts, I lean over, pull the cloth away, and let it fall.
Revealed is a computer that initially looks identical to the original Compaq Portable. I say, “It looks the same, but the Portable Plus has significant differences. Most important among them is a built-in, 10-megabyte hard drive.”
Then I hold up the new device for inspection. “Compaq has solved the problem of hard-drive fragility by going with a new 3½-inch drive and using a new packaging approach.”
Step by step, I point out how the Portable Plus fits together, showing how the much smaller, 3½-inch hard drive is housed within a strong 5¼-inch aluminum cage and is triple shock-mounted.
“The drive’s small size means it weighs less than its larger cousins, so there’s much less energy from its motion for the shock mounting to absorb. It’s virtually immune to the jarring that comes with portability.”
Later I discuss compatibility. “Apparently other PC companies often confuse the compatibility issue, thinking it only means running a certain version of Microsoft’s DOS, whereas most popular PC programs require a much deeper level of compatibility.
“Compaq’s definition of true compatibility means being able to run all IBM PC and XT software off the shelf without modification. This capability has been a fundamental premise of our company since its beginning.”
Near the end of the event, answering a reporter’s question, I lay out what will become a central theme in the battle for the future of the industry:
“IBM has set a standard that all the third-party hardware and software companies have adopted. So many customers are using software and hardware based on it, the standard doesn’t really belong to IBM anymore. That sounds strange, but if you step back and think about it, even if IBM didn’t support it there are so many hundreds of thousands of users out there that it will continue to be the standard of the industry. If IBM comes out with a new architecture, it will have to fight on its own to set a different standard.”
I get this question often during the next few years, and although my answer remains the same, my confidence in that viewpoint keeps increasing.
To most observers, the Plus seemed to be an expected and incremental move; in reality, it represented two major advancements in PC technology beyond what IBM was offering. Both advances were precursors to key strengths Compaq would continue to leverage in the years to come.
By being the first to solve the problem of ruggedness in hard drives, we further separated ourselves from the pack of clones and continued to steadily build our reputation. Not just for better products than our lower-priced competitors, but actually for better products than IBM. I began to communicate my long-term goal to Compaq’s management team this way: “If a buyer can have any PC on the market, we want them to choose a Compaq over IBM. We want Compaq to be the one they desire.”
After investing a tremendous amount of money and resources to make our portable totally compatible with the IBM PC, we ended up with something even more valuable and unique: We developed a proprietary technology that enabled us to make the Plus totally compatible with both the PC and XT. This was the first instance of complete backward compatibility. No other PC company had this capability.
Since the modern PC era started with the original IBM PC, the first opportunity for backward compatibility to exist in a PC came with the introduction of the XT. When some of the existing programs didn’t run correctly, it didn’t create a big stir. There was no noticeable criticism because the expectation hadn’t been set yet. The opportunity for us to demonstrate our proprietary compatibility technology was knocking on the door, and we gladly welcomed it in. Backward compatibility, or the lack of it, would become a much bigger deal over the next few years.
When the Compaq Plus was first introduced, retail dealers didn’t understand what backward compatibility meant—they had to see it in action. They had already noticed some of the existing software didn’t run on the XT. When Compaq sales personnel showed them the same software running on the Plus, they were quite impressed.
They could easily have ignored the issue and let it quietly pass. Generally, the dealers liked working with us and liked selling Compaq’s portables, because that made them less dependent on IBM. A lot of them began to promote this unique capability as a sales tool.
Then the press picked up on it and a few stories appeared. One quoted a Compaq dealer as saying, “Compaq is more compatible with IBM than IBM.” It sounded at first like nonsense, but it was true. We were more backward compatible than IBM, and we were the only company that was.
Even if IBM noticed, it probably didn’t see the news as important. IBM was the industry leader and was selling all the PCs and XTs it could build. It was the company that defined how computers were supposed to work. Tiny companies in Texas did not!
With the introduction of the Compaq Plus, we set a clearly defined paradigm for bringing new technology and innovation to the market. Each new product would need to have advancements in performance and features, clear differentiation that would justify its existence. And it would have to run all existing software.
Ironically, many of our competitors were complaining about how constraining standards were. Some said standards meant the end of innovation and the industry had to make a choice. But we knew better. The emerging industry standard would enable innovation, as long as a company understood how to innovate without creating incompatibilities. No one did this better than Compaq.
In late 1983, another pivotal point in the creation of a true industry standard occurred. Microsoft contacted us about an important matter: They wanted to know if Compaq was getting into the software business. They had noticed our dealers were selling more of our version of MS-DOS than Compaq computers. Customers were buying our software to use on other PCs.
We assured Microsoft that we had no such intentions. Then we licensed our version of MS-DOS to Microsoft to make our life simpler. Our engineers had found and fixed hundreds of incompatibilities, so our version of MS-DOS had become very different from the one Microsoft was distributing. Every time Microsoft issued a new version, we had to make so many changes that the time and cost became prohibitive. By licensing our version to them, the distribution version we and others received would already have those changes.
It appeared we were giving away the “family jewels,” because all our competitors would receive a more compatible version of MS-DOS. There was a significant delay, however, between the time we sent Microsoft a new version of our DOS and when our competitors received it and integrated it into their products. Each time we introduced a new advancement in one of our computers, it was fully backward compatible on announcement day. It would take months before that version of MS-DOS found its way into a competitor’s computer, so we were still able to maintain our reputation for delivering the most compatible PCs.
Licensing our software to Microsoft also established the process by which we would later spread an important attribute of the industry standard, backward compatibility, across all PC brands. Compaq was the only company with the technology to make its new products fully backward compatible. Our efforts to educate customers on the value of backward compatibility during the next three years would lead to a strong expectation of it and be critical to making the industry standard more independent of IBM. Making sure that all compatible PC makers delivered backward-compatible products played an important part in creating that expectation.
Compaq was the only company with the technology to make its new products fully backward compatible.
The fact that we licensed our software to Microsoft has remained a secret until now.
As 1983 came to an end, the Compaq Plus was seeing high demand and winning many awards. It was also about to play a key role defending against IBM’s first direct attack.