Resolve and Fortitude : Microsoft's ''SECRET POWER BROKER'' breaks his silence (12 page)

AIMING

June of ’91 delivered a bombshell.
A
pple,
I
BM, and
M
otorola announced their
AIM
alliance. The intentions of the triumvirate, firmly
AIM
ing at MS’s head, sent shockwaves through the PC industry, rattling its foundation. Bill’s earlier phobia, expressed after the OS/2 divorce, had been entirely warranted! The alliance was the brain child of Jack Kuehler, our old foe. It proposed a new computing platform incompatible with Intel’s and based on IBM’s brand-new RISC CPU–christened PowerPC. Its advantage was supposedly its price, and Motorola announced that it would act as a second-source supplier. The PC, as the world knew it, seemed doomed. Fantasy or reality, the press was already convinced.

The menacing aspect for MS was the coalition’s intent to create a brand-new OS—a cutting-edge object-based masterpiece. Obsoleting MS-DOS, Windows, OS/2, and NT all together! The maverick mouthpiece announcing—in his usual aggressive style—Apple’s side of the deal was Mike Spindler, my old boss and now Apple’s CEO. An exceptionally brilliant guy who did not always have his temper or his mouth under control! We had nicknamed him Dynamo. Not mincing words, his speech extolled the wholesome benefits of good old-fashioned, vigorous all-American competition principles at their best—
AIM
ed, unfortunately, squarely at our core! How many times has history described this? Among the great, the greater the challenge, the loftier the response! Go ahead, Mike, wake a sleeping titan.

The only real product the alliance had to show was IBM’s prototype, the not-ready-for-prime-time CPU. On the software front, there was no mock-up cardboard cutout, or even faint smoke and mirrors. The announced OS was the purest form of vaporware. A dream and a bet! MS announced MS-DOS version 5.0 with a six-month time horizon. The alliance announced hers twelve to eighteen months in advance. MS did not run whimpering to the Feds about a wildly premature, empty, and deceptive revelation, nor did any of IBM’s competitors or DRI. They applauded, and the press found nothing faulty to report.

Soon thereafter, I joined, Bill, Paul Maritz, and Steve to assess the situation. We concurred at once about the announcement’s deadly intent, though we were of differing opinions on just how much true and dry gunpowder the newly formed alliance actually possessed. Paul expressed skepticism about its longevity. He did not believe in Apple and IBM being able to work with each other. We all agreed that the new RISC chip would probably see the light of day, and Apple would probably purchase the CPU for building a new generation of Macs.

The ambitious OS pronouncement sparked a heated discussion. Based on his working experience with IBM, Paul did not believe that she and her enlisted software partner, Taligent, could deliver a viable OS in the projected time frame. As usual, Steve was diving for the panic button with Bill not far behind. These two still loved to create paranoia, yet it did not work in today’s closer quarters. Neither Paul nor I bought their arguments. The alliance had just too many divergent goals. What benefit would IBM receive from seriously altering the PC landscape? Wasn’t having Apple as future partner a cutthroat detriment? An Apple entry with an identical architecture would definitely hamper her already-thin margins further. Would Apple really buy an OS developed by IBM and give up her proprietary advantages? Dream on! Would this alliance be able to attract enough independent software vendors and lure them away from Windows? Possible—considering the resources of the three giants! I remained as skeptical as Paul, but crafting a tactical response was not, at this juncture, in my job description.

Shortly thereafter, Bill covered our bets, announcing a special Windows NT version for IBM’s PowerPC platform. IBM strangely promised OS/2 and a UNIX version. Sun chimed in that her UNIX,
nomen
Solaris, would be available as well. While many people had begun accusing MS of being an unassailable monopoly, this event laid bare our unending vulnerability. The AIM alliance had the clear and immediate potential to stand the current PC world on its ears. With her meanwhile impressive one billion dollars in annual revenue, MS was a dwarf compared to the united powers of these three entrepreneurial monsters. Competition was vigorously alive and vitally well, absolutely threatening MS’s livelihood. The Federal Trade Commission’s investigation of MS simply continued. It quietly, powerfully, and malevolently swept through the MS campus, unbidden and beyond the reach of logic or reason, like a ghostly ground fog, unleashed by the futility of our competitors and our very government itself. Taxpayer funded. As if AIM were nonexistent.

I took the public relations circus around the alliance with a grain of salt, and so did the always-pragmatic Paul. We had experienced plenty of other threatening announcements before; they did not necessarily translate into timely execution or genuinely competitive products! The paper they were written on was cheap and expendable.

One year later, the alliance derailed when ambitious OS plans did not materialize and Apple couldn’t agree with IBM on hardware specifications. The Power RISC chip did make it into the market in proprietary IBM and Apple systems. MS withdrew NT. IBM never finished an OS/2 version and focused solely on UNIX. Apple bought her CPUs only from Motorola. Mike Spindler lost his CEO post. No one ever heard from him again. Amen!

The other company who announced an OS for the new architecture was Be Inc. She was run by Jean-Louis Gassée, Apple’s former French country manager, who had helped me contract Multiplan for Apple’s IIe European launch in early ’83. After leaving Apple, where he succeeded Steve Jobs as leader of the Mac division, he entered the OS fray. Eager to compete with and dethrone MS, he developed BeOS for IBM-compatible PCs. Unable to find sufficient customers for his baby, he blamed what else? My business practices. Jean-Louis, a smart man, should have recognized his core strength—great French showmanship. Not enough to pull his company through. Bill and I applauded him for his admirable efforts to gain a foothold. He earned a second chance, but IBM’s proprietary UNIX versions eventually ate his lunch.

Still worried about the importance of IBM’s new architecture, Bill initiated, right after AIM blew up, several informal efforts exploring a renewed partnership with IBM. He was rejected. Tenacious Jack Kuehler made sure of that and responded, courting our competitors and forming hostile alliances with Lotus, Borland, and Novell. I felt bad for Bill, but IBM was by no means ready to give up on OS/2, and it showed. The wounds of her current management team were too deep and needed time to heal and, really, a new leader to arrive. The only good news: few of our OEMs reported serious demand for OS/2. Observing this, I noticed IBM’s close allies were betting on the wrong horse—ultimately destroying themselves.

ASSIGNMENTS AND THREATS

After my boss Jeremy Butler left the company in the fall of ’91, I again reported to the president. In addition to managing all non-US business, he had also been in charge of our product-localization group. The main task of this group was devoted to publishing rules and educating developers on how to write easily localizable code.

To improve our localization efficiency and reduce delays, management wanted to reengineer the current process by integrating localization specialists back into the core-development groups. Jeremy did not finish the integration project before he left. Lucky me! I was the gifted soul the burden now fell upon. I shouldered the task as a thought-provoking challenge. Keeping my eyes on OEM, I delegated most of the gritty work to the group’s current management team and accomplished the integration in just six months, well ahead of the twelve-month goal.

Soon thereafter, another strange assignment fell into my lap. Oddly, Steve asked me to visit with DRI’s executives, exploring a technology exchange. The idea was to obtain the rights to her allegedly superior DOS technology for use in future MS versions. Steve had reportedly met Dick Williams, DRI’s CEO, at a conference in the UK. He seemed open to pursue such an exchange. Promising to follow up on their conservation, he made me the point man. I considered my assignment an odd and unwanted mission. My attempt could be misconstrued as apparent collusion between competitors. In using DRI’s technology in future MS-DOS versions, antitrust lawyers might accuse us of eliminating a competitor. I did not understand why we wanted this technology in the first place, or why I was picked to consummate such a deal. With a lurking suspicion, I departed for Utah to visit the hyena’s den.

As I arrived at DRI’s facilities, my doubts did not disappear. Dick Williams was present together with one of his attorneys. I was alone, and the right thing for me would have been to turn around and return with my own legal support. Instead I stayed on and made a vague proposal along the lines I discussed above. I found out that the technology in question was available for $30–$40 million. Steve had instructed me that he would not pay DRI in cash. In its place, he was willing to extend DRI a reseller license for MS-DOS as payment in kind. The total trade-in value he envisioned was $5–10 million. Hinting that their asking price was roughly five times higher than I had expected brought laughter; the deal was off.

I loathed the idea of having DRI as a trusted second source. Against all popular sentiment, I wanted her to be around and compete. DRI’s independent effort was the best insurance for my customers to receive constantly improving MS-DOS versions. A compelling reason to keep her alive and our development guys on their toes!

Did Steve send the wrong guy? You bet he did. He should have gone himself. His intentions were only loosely communicated and had too many inherent shackles. In addition, I was biased about DRI ever becoming a reliable partner. My compromised initiative can be summarized as denying DRI a deal bordering disobedience. Allowed under the doctrine I was operating under, considering the ultimate edict voiced by Bill: beat DRI fair and square in the market—much easier to follow and fun to execute!

I cannot exit ’91 without mentioning two other key events taking place. The expansion of American Online (AOL) is one; Novell buying DRI the other. AOL’s key contribution was to expand its online services everywhere, from serving the gaming community to enabling PC home users to exchange e-mails and access rich online information. Networks of this kind already existed, but most of them either were not as easily accessible or charged relatively high fees. Initially, we applauded AOL—just one additional reason to buy a PC. As reality set in, top management viewed AOL’s endeavor as a lost opportunity for cash-rich MS.

Novell’s buyout of DRI that year was a serious and potent threat. Raymond Noorda, Novell’s CEO since ’83, allegedly coined the word
co-competition
in describing the PC software industry. It suggested that even rivals enjoy a natural concordance to attain ultimate successes. His company originally sold servers. Novell’s breakthrough came when she developed a network interface card for PCs and corresponding software. The software, later called NetWare, enabled PC servers to easily connect to each other. During the ’80s, Novell transformed herself from a mostly hardware company into a pure software company, an impressive accomplishment. Until then, other server and network card providers regarded Novell as a competitor. Dropping the server hardware from her sales portfolio, Novell succeeded in making NetWare the leading PC network operating system. By ’91, NetWare’s lead appeared unassailable with MS-LAN-Manager and several UNIX versions together having less than 10 percent market share.

Thanks to my group limiting DRI’s success as an OS supplier and after the failed buyout, her CEO, short of cash, looked for a suitor. As a perfect merger candidate, Novell had the money and her CEO the zeal to make MS’s life harder. When I heard about Novell’s purchase of DRI, Bill and I were profoundly alert. Ray’s famous vindictiveness toward MS accompanied by Novell’s cash hoard and her much larger sales force strongly suggested a prolonged and more vicious firefight, for DOS market share was about to commence. In selling NetWare to our OEMs, Novell already had existing relationships with a lot of my customers. I feared her sales force could easily leverage these and prove plenty successful selling DRI’s product line in unison. Bill agreed.

As the merger unfolded, I was urging the MS-DOS group to finally deliver a version that could be embedded and function in read-only memory (ROM). DRI had successfully sold one for a long time. I considered this product segment her last bastion to fall. Not needing a hard disk to function, it had found its ways into the industrial controller market and emerging handheld devices. Not a huge segment then, but with the accelerating miniaturization of electronic components, its expansion was a foregone conclusion. The reason why I eagerly wanted to create a foothold, establish our brand, and take business away from DRI. I convinced Bill that this was an urgent, exciting, and obtainable opportunity, and with his help, the product group eventually delivered.

Luckily, my nightmares did not become reality right away. I never fully understood why Novell did not hammer the holy hell out of us. The machinations of executing the merger were undoubtedly one reason for her failure. Novell’s core business had mostly retail focus. A smaller amount was done through enterprise-focused OEMs. DRI had hardly any retail presence and sold directly to smaller non-brand-name OEMs. The salespeople in Novell and DRI had different skills and served different distribution channels. The grand tactical error: her sales management missed the opportunity to create synergies between them.

As we attacked Novell/DRI in her last stronghold, we found scant resistance. One and a half years later, despite—or because of—the merger, we had neatly swept a significant portion of the embedded business away from under her. No serious effort was made to stop us. No discounts, no marketing, nothing. It was too easy, I thought. Frankly, I was disappointed, having expected a bloody battle.

No attack on Windows either. No taking out a license, as IBM had done, from Apple to get over the patent hurdle. No cloning attempt of Windows either. Novell certainly had the resources and the skills to attempt such a feat. She could have succeeded in launching an alternative and nicely polished Windows-like product. Did Ray Noorda have other motivation for acquiring DRI?