Resolve and Fortitude : Microsoft's ''SECRET POWER BROKER'' breaks his silence

by Joachim Kempin

  Counterpunching DRI, we announced making an end run for a superior version with a six-month delivery horizon. The marketing folks did a good job comparing our planned advanced features favorably to DRI’s. A handful of journalists and politicians accused us of stalling DRI’s sales momentum with a containment strategy of announcing early. Comparing the timing of our announcement with those of other software companies, we were just following common practice. I further believed the development team—after Bill’s scathing e-mail—now truly understood the urgency to keep a promise.

  That same summer, we got a new president in Michael Hallman. I decided to test the new man. As long as I had been running OEM, even when reporting to Jeremy Butler, Jon Shirley had continued signing all contracts and amendments. To find out if the new guy wanted to follow in his footsteps, I asked my admin to save all documents needing presidential signatures for one month. I then sent her upstairs with a fully loaded office cart. As predicted, within minutes, Mike was on the phone. I listened patiently as he barked away, ruminating over my provocation. What the hell? I explained what Jon had practiced then, exploiting the example at hand, proposed he delegate signature authority to me. He agreed, without hesitation, empowering me to sign all OEM contracts. A smart move! Compared to Jon, he had limited knowledge of my business and its intimate details. Jon had added value in the process. Michael opted to free up his time. Compressing the approval chain had another positive side-effect: reducing turnaround time with customers.

  In general, he was easy to work with, but I received hardly any directions from him. With Jeremy constantly on the road, I met with Bill more frequently. A lot of employees were surprised when replacing Jon with an outsider. They rightfully assumed Scott Oki, my old boss, or Jeremy Butler would have been better candidates. Scott had quit the company earlier, reading the tea leaves correctly. By ’91 Jeremy was surprisingly on his way out as well. Constant traveling had stressed him out and severely impacted his health. The obvious choice for the job would have been Steve. One of our board members told me later that the board did not think he was ready for the job yet.

  Just after Mike joined us, I was invited to address the board for a second time. In addition to analyzing the tough DRI situation and the continuing OS/2 onslaught, the board wanted a ten-year OEM business outlook. I spent two weekends combing through our historical numbers and any forward-looking analysis I could find. In place of presenting just a simple growth chart for our future revenues, I developed a worst- and best-case scenario before deciding on a most-probable trend line. The board was impressed by the methodologies applied to back up my predictions. I still don’t believe anybody truly trusted my final conclusion as I predicted nearly ten billion dollars of OEM revenue for fiscal year 2001. Luckily, nobody threw me out or assumed I would still be in the same job and could be held responsible. In the end, though, the unexpected happened.

  Long-term planning like this was creeping into the company. The plan being demanded and repeated annually from then on, I considered it a total waste of time, considering the inherent competitive and dynamic business environment we lived in. Relevance and focus was asked for and not an endless contingency-planning exercise, which more than once led to politicized gamesmanship.

  Rather, I signed up more Windows customers. To gain additional penetration, I started a contest for the sales reps. With version 3.0 boosting performance significantly and allowing for multitasking, existing customers increased their commitments. All of the directly selling OEMs signed on. Our main holdouts were Amstrad and Compaq. Despite having signed on, IBM refused any public support. Not increasing prices helped our crusade as much as winning the hearts and minds of reluctant ISVs. This was where Steve excelled as he whipped our so-called evangelist support group into shape. Only Lotus, WordStar, and WordPerfect remained the lone holdouts exclusively seated in the OS/2 tent.

  CRISIS IN THE FAR EAST

  With Windows finally on an uptrend, I visited my Far East customers again. We had succeeded in obtaining commitments from them for our printer OS. In the fall of ’90, the project ran into serious trouble. Not only had we missed all promised deadlines, but also worse, the code itself was not stabilizing. Product management was seriously considering abandoning the well over one million lines already written. I decided to personally give customers the bad news. Not an easy task, considering the investments they had made and the loss of face our decision would cause for the Japanese execs who had signed on. My meetings with Canon and NEC had the singular goal of minimizing damages. Delegating the unpleasant task would have set a bad example. I knew these customers were unhappy even as we offered to give them the unfinished source code at no charge, allowing them to use it with no strings attached.

  During my stay in Tokyo, I met a second time with Sony’s management. Our meetings were always cordial. Their executives spoke English well, having spent mandatory time in Sony’s US subsidiary. However, the electronic consumer giant was no intimate friend of MS. Sony’s president, Mr. Idei, later expressed this belief as “MS wants to control us.” I still do not understand how he drew such an inaccurate conclusion. We wanted cooperation and openness; he wanted proprietary secrets and unmerited advantages. The overly polite pins-and-needles atmosphere in meetings I attended did not signal trust.

  Bill would have loved to have Sony on our side. He visited with her execs numerous times without achieving a relationship breakthrough. Sony was no doubt an innovative and powerful consumer electronics and entertainment company. If her top management would have opened the kimono just a little bit, the industry would have moved forward faster. Instead, maintaining distance, she included as many proprietary features in her PCs as possible at the cost of endangering the holy grail of compatibility. MS was eager to engage in media technologies with her where she had an undisputable lead. What I found was a company who reluctantly bought our OS products and was barely willing to work with us. Sony obviously had her own agenda and ambitions. MS found herself stuck in the role of being just another of her nonstrategic vendors.

  Before returning to Seattle, I stopped in Hong Kong. The local OEM group had organized a small Asian customer summit. I delivered a keynote speech and met with several key Asian manufacturers. What I heard in regard to losing business came as quite a shock. Therefore, I extended my stay and gave the Asian sales team a pep talk, repeating key OEM strategies and reemphasizing our business objectives. We were under severe attack by DRI, particularly in Taiwan. Several medium-size Taiwanese OEMs had abandoned us and signed up for DRI’s new DOS version. I made time to be interviewed by a local journalist who had, a couple months earlier, done the same with DRI’s executive responsible for Asia. He had named twenty-nine different companies in Southeast Asia, which he claimed were licensing from DRI exclusively. The journalist gave me a hard time, chiding me over how we had lost the edge and wanting to know what I planned to do about it. I told him in no uncertain terms would we make a serious effort to bring these customers back into the MS igloo where they, as I saw it, belonged. After the interview, I asked an aide to dig out the article he had mentioned. Fortunately written in English, it contained a treasure trove of information the DRI guy had foolishly disclosed. We immediately proceeded with mapping out a counterstrategy. We took the target list of prospects straight from the newspaper article. Whoever volunteered this should surely have been fired! Bragging rights over common sense!

  My last stop was Taiwan, where I attended the opening of our new facility. As part of our counterstrategy, we decided to bring in experienced support personnel from the United States to give it a running start. They made all the difference in the world. The smaller OEMs DRI’s executive had supposedly won over and boasted about hugely appreciated the sudden attention. DRI had no permanent feet on the ground, sporadically dispatching personnel from the United States. Being able to nurture the manufacturers locally daily was the key to welcoming them back into our realm. The result: six months after my HK visit, DRI had lost 90 percent
of her accounts on the island.

  At the end of the visit, I was joined by Bill to formally open our new sub. His visit to Taiwan was publicized when it was announced that he would be the keynote speaker at a local ISV conference. The idea to direct attention to the conference and increase attendance backfired; we received a bomb threat. Traveling with Bill and visiting additional customers meant I was included in all the security measures. Sweeping hotel rooms for bombs, changing cars frequently, and never venturing alone in public were but a few. Fortunately, it was a hoax. I admired Bill for not cancelling the trip. After experiencing all the hoopla and intrigue, achieving celebrity status like Bill’s was definitely not what I longed for!

  PIRATES OF THE WORLD

  On my trips, I made a point to personally investigate how small system builders were conducting their businesses and where their OS software derived from. I personally observed the incredible amount of piracy happening in Asia, South America, and to my surprise, Germany. Software theft seemed to have no meaningful moral or legal consequences. Any child could do it by just copying, and hardly anybody watched.

  Copies these crooks made did not look like the original software we advertised and sold. Honest buyers recognized the differences and gave the cheaters a hard time. To overcome authenticity objections, a network of illegal replicators sprung up and produced look-alike packages. The illegals often shipped them in small FedEx boxes avoiding discovery. We, in turn, engaged customs authorities to stop them. Their motivation to stop the nuisance: when a counterfeit package changed hands, the government missed out on import duties and taxes, just as we did in royalties!

  Our own product management was partially responsible for these thefts by not embedding a copy protection scheme that was difficult to break. In the early days, Bill had made his opinion about stealing software well-known. His comments caused controversy in the industry and with the trade press. In PRC, where piracy was most endemic, IP17 theft can be traced as far back as Confucius. One thousand five hundred years earlier, together with other Chinese philosophers, he had extolled the virtues of public ownership of IP. Modern day OSs and software apps free to anyone for the taking? We had quite a task ahead of us, reeducating a country with a citizenry of then nearly a billion people misguided for centuries.

  I put together a small task force to gain an in-depth understanding of our losses and challenged its team members to turn heedless thieves into honest customers. The team discovered straight off the bat that software replicators our OEMs used were part of the illicit supply chain. Many of them conspired by producing extra copies and selling them out the back door. A number of our own OEM licensees did the same. I estimated we only got paid for two out of three copies in use. This meant my business could have been 50 percent larger—a great motivation to personally lead the crusade against the IP terrorists.

  How did this get so out of hand? Our OEMs inserted backup copies of our software and a user manual into every PC box. They were supplied to them by duplicators of their choosing without any control from us. Here laid the crux. So my task force recommended the implementation of a proven concept used by the entertainment industry requiring duplicators to obtain a license from us. Our immediate concern was solving the mechanics of such a supply chain. Thousands of companies throughout the world reproduced our products. How would we select and license the right ones?

  Grant Duers, the team leader and an experienced logistic expert, pointed out that reducing the numbers of replicators would lead to larger production runs and lower repro costs. He had my ear. Investigating how costly these inserts actually were, he discovered every so often that they exceeded royalties paid to us. Our customers were being creamed! (In a footnote to our fact-finding mission, we enlisted the product groups to reduce documentation page counts.) This convinced me that if our moral intent would not find customers’ approval, cost savings for sure would! Intending to close loopholes and increase our ability to control the flow of goods, we asked our attorneys to include auditing rights and report requirements in our envisioned replicator licenses. We then made it mandatory to put an MS designed label, containing an encrypted serial number, on all packages.

  Implementing the new procedures enabled us to compare replicators’ unit reports with the ones OEMs submitted. It took no time before large discrepancies were discovered. Lag time between the reports often caused this and was easily straightened out. Deviations had other reasons: every so often, OEMs served the black market without paying us or reported fewer machines than they were actually selling! Sufficiently large deviations warranted pronto follow-up, and an audit via an independent accounting organization was typically the result. It required my approval. Ironically, most caught did not deliberately cheat. They had grown so fast that their data processing and internal-reporting capabilities lacked accuracy. My controller therefore made it her task to educate customers on proper data capturing and reporting methods. To her delight, collecting the proper amount of royalties due was, from then on, much more easily achieved.

  Other MS groups watched anxiously if the new measures made a difference, and they followed as soon as our lead showed progress. Aligning all sales groups and improving the safety system occupied our logistics group for decades. Next we established a network for issuing security numbers electronically, directly to the printing press. When installing our software, we demanded that end users type in the sixteen digits found on our security label. Still not adequate! The pirates soon cracked our algorithm. In turn, we made the math harder for them. The next logical step was to use a WW computer network, which could validate every single number from a central place denying multiple uses to cheaters. With the appearance of the Internet, that got solved as well. Not all holes were ever plugged sufficiently, leaving room for substantial leaks.

  Internally, my colleagues labeled me the antipiracy czar. In the press I was described as MS’s enforcer. None of these monikers offended me one bit. Successfully discouraging the thieving backdoor artistes was all that counted. As long as I ran OEM, I continued advancing these means with passion. Eventually, our products would be delivered on CDs or DVDs. We worked with one supplier, printing unduplicable (or so we thought) blanks with holographic images and distributed them to other replicators for final production. They gave our products a distinct and artistic look and again increased our security incrementally. The pirates countered with a design mimicking ours. In a fascinating game of global cat and mouse, we indeed won over time after several incarnations of hologram technology. The last task was to create piracy-proof sticky labels. They were not only hard to replicate but also nearly impossible to remove without damaging the carrier. Applying banknote printing techniques and special glue, we upped the ante.

  Protecting our products was without question a never-ending effort. Each of our actions provoked a reaction from the notorious scam artists. Even in this field of endeavor, we learned again that complacency was enemy number one.

  FALLING BEHIND

  MS Windows 3.0, in combination with MS-DOS 5.0, benefitted five companies totally committed to design and sell these types of PCs: AST18 Research, Dell, Packard Bell, National Cash Register Company (NCR), and Gateway (GW). The most interesting case study of these emerging movers is GW’s. Her founder, Ted Waitt, brought the company to life on the prairies of South Dakota with a strong belief in selling PCs directly to customers, bypassing the retail channel. Displaying his loyalty to the community he’d grown up in, he designed the Gateway shipping boxes with a flourish of Holstein cow pattern. Producing high-end PCs for gamers and power users, his company was soon widely recognized as bearing a strong brand. I got to know Ted well, and one day he approached me about licensing MS Excel and MS Word for his higher-end product lines. Supportive, Bill gave me the green light for a trial period. I was convinced we had found an ideal partner. Typical GW customers bought highly customized PC systems by phone. Why not try to sell them our Windows applications at the same time? GW could easily preinstall them and s
ave an interested buyer time and installation hassle.

  Teaming up, we expected to increase our Windows apps market share, reduce piracy, and give our products extra market octane thanks to Gateway’s advertising. For GW our willingness constituted a first in the US market, and I was hopeful that pioneering application bundles would boost her sales and visibility. I agreed to a three-month exclusive. Afterward I would unleash the opportunity on other parties. The reason for GW’s short contract term can be found in my desire to move slowly with so many inherent unknowns. After all, our Windows apps were by now part of our crown jewels—at par with our OSs. I offered GW a reasonable but not too aggressive price. The exclusive made this palatable.

  Market reaction to our experiment was measured skepticism. Considering the additional cost burden, other OEMs told me the initiative would never succeed. I disagreed based on my German experience and what we were doing in Japan. Retailers, our own retail sales, and our application product group were initially not persuaded either. I was stepping on a lot of toes, exploring untested territory. Jealousy over the additional OEMs’ revenues developed, and a kind of turf war ensued—politicking at best. Steve, in particular, wasn’t terribly enthused about licensing apps to OEMs in the US. Again, he was fully committed to “his” retail channel and had the merchant’s full attention and best interests in the fore. It was soon too late to be sentimental. Against all odds, the experiment developed into a stunning success. Within six weeks, we had GW’s competitors knocking at the door. At the end of the trial, GW begged me, to no avail, to extend her exclusivity. In its place, I convinced Ted to increase his commitment and sweetened the pot. Other OEMs trying to follow GW timidly barked at our prices but nevertheless experimented with a limited number of their own bundles.