XBOX
By 1998/99, planning for an MS entertainment console had gone into overdrive. MS, the ultimate software company, was boldly venturing into the sharply different dimensions of being a hardware vendor. Yes, we had produced computer mice and ergo keyboards before, but a fully functioning computer system had never been on our list of offerings since ’86, when we had abandoned an early game console in Japan.
Since then, the competitive landscape for game consoles had changed radically. With Nintendo in the lead, Sony had launched PlayStation, immediately making life difficult for all competitors. Keeping a wary eye on her progress, Bill was concerned that this alternative computing platform could conquer the living room. Sony’s division was run by Ken Kutaragi, an energetic and impressive entrepreneur and business executive. He had equipped his console with an advanced, lightning-quick proprietary graphic processor that was far beyond anything found in PCs, attracting a huge number of leading-edge game developers. Nintendo with her GameCube and Sega with her Dreamcast system felt pummeled as gamers left opting for the newcomer.
We landed a deal with Sega, making Windows CE her OS of choice. It got us a foot in the door. Yet Bill insisted on gaining a larger piece of the action by launching a full-blown game console of our own, stopping Sony’s march. To justify his ambitions to the board and the shareholders, he wondered if we could offset predictable console losses with game-software profits. Ed Fries, a MS veteran and the genius behind MS’s immensely popular flight simulator program, made a convincing case confirming Bill’s idea, and soon the race was on. Promising to attract and recruit a critical mass of independent software vendors for the new platform, Ed sealed the deal. Short on detail and long on vision, MS style.
I was stunned to hear about this new endeavor. My main concern: how much would our OEMs appreciate our move into producing a complete computer system potentially endangering their PC business? We had enough daunting challenges already; why take on yet another where initially huge losses were expected? I further questioned our ability of sourcing parts cheaply enough and managing production aspects in a cost-effective and timely manner. The idea of still-to-be-created game software coming to the rescue for this hardware adventure did not strike me as a solid business foundation. Questions regarding the software being delivered on time and us being able to create hot-enough titles crossed my mind. To accomplish our profitability goal, we first of all had to sell lots of consoles at probably substantial losses for years to come.52
Despite my customer-related concerns, Bill was dead set, brimming with optimism and perfectly willing to gamble with shareholder’s money. The ambitious project soon went one step further with Nathan Myhrvold, our chief technology officer and resident propeller-head, insinuating we could eventually transform our game console into a full-blown PC. I shook my head. Fortunately, cooler heads prevailed. At least for the moment.
In the end, Bill listened and offered me a chance to approach several OEMs, swaying them to build and launch a game console on their own account using our blueprint. We committed to supplying a special Windows edition along with a promise to develop game software exclusively for the new platform. We talked to Dell, Compaq, Hewlett-Packard, Acer, and Nippon Electric Company (NEC). Thanks but no thanks! They had studied the Sony and Nintendo business models in detail and didn’t believe they had the ability to compete profitably. Leaving us to stick our neck out, alone! We should have listened closely. After this setback, there was talk of buying Sega or Nintendo, neither faring too well, but no deal was ever seriously attempted.
Just after I left the OEM post in 2001, Xbox was launched. The project had gone through several iterations and numerous delays. MS had a hard time designing the box properly and sourcing the components at acceptable prices, shocking AMD when deciding to go with a special low-priced Intel Pentium CPU. Ed Fries did a super job recruiting ISVs, but even then, games for the new platform were coming along slowly. ISVs were outright skeptical if we would pull it off and were busy writing PlayStation games. Buying one of them enabled us to publish one big hit, Halo, a supersuccessful war game, yet not enough to compensate for mounting losses. In the following ten years, the company lost well over $ 10 billion on Xbox before breaking barely even. Not the type of success I would have aimed for.
One last aspect to our hardware story: Rick Thomson, who had been in charge of building all MS hardware products, joined the Xbox team in ’99 to build the new platform. As he left the company a year later, frustrated by internal politics, our hardware group experienced compounding difficulties. New management came in and decided to abandon the OEM mouse business I’d been nurturing all along. I was deeply disappointed but no longer had the passion to fight for its revival. The little critters would from now on be supplied by other manufacturers while MS would focus entirely on the supposedly more profitable retail sector. Opportunity lost!
LATE CHALLENGES
My travel in 2000 was mainly to developing countries such as South Africa, China, and India. I was not satisfied with our progress in India in particular and decided to visit the country for the first time ever. The leading IT Company to visit was Hindustan Computer. Founded in ’76 and still mostly focused on India but with plans to branch out into Africa, she had moved beyond selling hardware and transformed herself into a top-notch IT consulting and engineering company. Bill recommended I talk to Jai Chowdhry, her CEO and founder, calling him one of the smartest people he had ever met. After my arrival, he graciously gave me an educational overview of India’s fast-growing IT industry and the local PC market in particular. Typical of a developing country, India’s PC business was firmly in the hands of small system-builder shops. Among the relatively few larger manufacturers, HCL was the most eminent. To stay competitive with the screwdriver outfits, the smaller manufacturers sold their PCs mostly without any software and let the pirates find a way to take over at the point of purchase. HCL, with her established brand name and her primary target of serving enterprise customers, was a noble exception to the rule.
I visited India a couple of weeks after Bill Clinton had toured the country in a formal head-of-state visit. My local team had planned the visit following his route, including some of the same hotels and restaurants. You do live well as president of the United States. By contrast, the country’s poverty rate was staggering. The other much-too-apparent observation I’m compelled to mention was the tremendous pollution I observed. Mexico City was bad, and Beijing may be close behind, but nearly any large Indian city tops them in spades.
I found the business climate in India a touch aggressive and tense, though the people I met were sharp and highly educated with advanced business experience. As a result of my visit, we teamed up with several OEMs, incenting them to promote genuine software bundles. It helped move the needle in the right direction for a while, but not as much as in the People’s Republic of China. The main reasons as I saw it: the government had no stakes in these companies and was not supportive of anybody who wanted people to pay for software.
South Africa (SA) was my next stop. I had first visited the country in ’94, shortly after the apartheid regime had been removed from power. I had gotten to love its wild primal aspects, which still existed in an ethereal and natural state as long as you were willing to travel into the veld. Unfortunately, her poverty rate had remained way too high even as her new government had made long strides in easing some of that pain.
The local PC market was dominated by a company called MUSTEK, with 40 percent market share. Ran by her founder, the Taiwan-born and Pittsburg State–educated David Kan, her huge share was a result of the apartheid boycott, which had prevented overseas PC manufacturers from branching out into SA. David arrived there from Taiwan, incorporating in ’89 and later floating his company on the local stock market. His Taiwanese connections enabled him to import enough PC components to make an impressive run. I had met this brilliant and energetic businessman several times in the past, and we developed a close friendship. David ha
d no ambitions to compete on a global scale, but he had cleverly branched out into other parts of Africa proper. To shore up his core business, he established a profitable component-distribution business including Windows, helping us to reduce piracy in this remote part of the world.
One reason for my current visit was to review the progress of a special project we had launched exclusively in SA. The idea was the brainchild of our MBA whiz kids. I had predicted failure but was interested enough—along with a skeptical Steve—to let the experiment proceed. Despite of only a slim chance for success, we opted for gaining valuable experience in exploring an alternative way of selling MS Office.
Priced at about $300 down there, a lot of buyers in this developing country were hesitant to shell out that much money for personal productivity software. Instead of requiring the full price up front, the whiz kids proposed to allow customers to pay in monthly installments. Their plan called for OEMs to conveniently preinstall the product like an OS, and if end users agreed to purchase it permanently, a low monthly fee would be charged to their credit cards. To limit our risk and exposure, the small and remote SA market had been selected for a test run. My reason to be skeptical about the idea was twofold. First and foremost, credit card ownership was not exactly widespread in SA, the Internet usage was low, and last but not least, how would we police the customer’s failure to pay up?
David Kan, together with other locals, had agreed to participate in the trial. The initial sales had been slow but promising. The OEMs had gotten a small sales bounty, and as predicted, payments beyond the initial ones had been hard to obtain. Not being eager to deal with public outcries when cutting off people’s right to use MS Office, local management was weary. Still in SA, I sent e-mails back home, asking to abandon the project and extend amnesty to current participants. In retrospect, we were ahead of our times; most likely we’d simply chosen the wrong country to launch the experiment.
For quite some time Bill and I had been talking about trying something similar with Windows. Exploring the concept in detail, we soon discovered how hard it was to implement and enforce payments. In theory, we could have charged OEMs zero dollars for preinstalling Windows and given PC purchasers thirty days of free usage. Agreeing to pay us an annual fee would have kept their PCs functioning. Once modeling such an annuity concept, I discovered, despite an obvious first-year revenue shortfall, the gain derived from no longer having to extend volume discounts to OEMs, and adding a small finance charge could boost per-unit revenue.
What spoke against the idea? Not every end user had Internet access, and any other way of implementation would have been awkward, inefficient, and probably impossible. Nor did all Windows customers possess a credit card. The final issue breaking the camel’s back concerned enforcing the annuity deal. Windows, like any OS, remained essential to keep a computer running, unlike an antivirus program, for example. Would we put our money where our mouth was and enforce nonpayments by making PCs dysfunctional? Instead of disabling them completely, implementing an automatic slowdown and reducing the number of applications to run simultaneously53 were other technical possibilities. Undoubtedly, any of these measures meant creating a lot of unhappy PC owners. Last but not least, hackers could go to work and repair the inflicted damages; in the end, it was just software, and even if encrypted, it could have been cracked. We even considered selling an encrypted hardware dangle with every Windows copy, which turned out to be too expensive to manufacture but might be a way to implement a stronger protection and enforcement scheme today. Our final conclusion then: wrong thing to do, dream on about OS annuities, end of discussion!
Back home I had to attend to an unexpected crisis stemming from the antitrust trial. Documents the Feds had confiscated during their investigation had been posted on the Internet for public perusal. Naturally, they were a treasure trove for people trying to gain insights into our internal discourses. Best compared to WikiLeaks. The ones published had been scrutinized by our attorneys, ensuring confidentiality. Redacting three million documents, nevertheless, was prone to errors.
One occurred when a report from a negotiation with Compaq was not well-enough guarded. The lowball Windows price mentioned in the erroneously published document immediately caught the eyes of Kelly Guest, Dell’s corporate counsel. Freely but mistakenly assuming it reflected Compaq’s actual royalty, he immediately apprised Michael Dell and Kevin Rolling, Dell’s president, of his findings, which at once raised a howl of despair and disfavor. Constituting about half of what Dell was actually paying, the news made the execs understandably furious, signaling a de facto huge disadvantage.
Welcome to the world of most favored nation clauses where trouble is sure to abound. Their origins stemmed from treaties between nations in which signatories agreed to grant each other the same favorable terms as offered to others. Similar clauses had found their way into commercial contracts. I hated them with all my heart and soul. Yet in supertough negotiations, we had deigned to extend them but so far to only three customers. (Proving we had way less pricing power than the judge was made to believe.)
Under a most favored nation clause, we guaranteed customers the absolute best royalty. Guaranteeing this to Compaq, the top volume shipper had for now no apparent risk. Making sure we could honor such a promise, I instructed my attorneys that anytime a similar clause was headed toward another agreement, they needed to obtain my advanced approval. A couple of years later, Dell got wind of her neighbor’s favorable contract term, and Kelly Guest—why not?—asked for reciprocal treatment. With Steve trying to climb into bed with Dell’s execs, there was no denying.
With Compaq outselling Dell by 30 percent, I was adamantly against extending her prices to Dell and asked our attorneys to find a way to soften the absolutely-best-price guarantee. Under the modified clause, Dell could only obtain Compaq’s price when she bought the same number of Windows copies. Dell understood the finesse but nevertheless called us on the carpet. A painful meeting with Steve commenced. Between Kelly Guest arguing eloquently, Michael Dell playing hurt, and Kevin Rollins seeming pissed, we were in a grand and dramatic pickle. Wanting their cooperation, Steve had no desire to stand up to them. I passionately wanted our contract language to prevail. In the end, none of my arguments got through. I got plainly overruled. We left with me sporting bruises and bad feelings but nevertheless agreeing to apply the reduced royalty rate going forward. Kelly Guest’s political gamesmanship triumphed, having judged the situation correctly. I had met him several times before, got to know him quite well, and had a great deal of respect for him. He had pulled a fast one. I swallowed my pride.
When Steve and I recapped the situation shortly thereafter, I angrily told him to consider himself from now on Dell’s chief account manager. The sales reps were better off dealing with him directly when micromanaging this account. He laughed it off.
Steve’s behavior was in stark contrast to Bill’s. Both had a difficult time saying no to customers. Right after I got the OEM job, I watched Bill struggling through price negotiations though managing to stay involved, if obliquely. After we’d had a chance to chat at length about how to avoid putting him on the spot in the future, he changed his tune. From then on, whenever customers asked him directly for price concessions, he redirected them point-blank: “You’ll have to work that one out with Joachim.” A decision he hopefully never regretted, allowing him to remain firmly in the driver’s seat while empowering me to obtain the best possible deals. Auftragstaktik at work!
SEARCHING FOR A SUCCESSOR
All these little incidents added up. Having to endure Steve’s emboldened management style and observing the command philosophy drift away from Auftragstaktik with iceberg resolution, a tactical standard which had served MS so well, I longed, at last, to leave. My favorite successor was still Richard Fade, who was by now back in my group, working diligently and hard as ever despite the unresolved drama in his family. Steve agreed with my judgment, but as we seriously approached our preferred candidate, he
remained unwilling to commit long-term. A concern to me and a much larger one to Steve, desiring to build a lasting team. In the fall of ’00, after a visit to Europe, Steve had made up his mind, telling me he would like to see Richard Roy, the current German country manager, to succeed me. Another German, what was so great about them? I asked Steve to allow me an opportunity to contact my friends in Germany before giving an opinion. Their feedback left me unimpressed, ranging from “The guy is too political and not decisive enough” to “He’s not well liked by a number of customers and employees alike.” Respecting my well-placed informants, I advised Steve not to appoint him. Reluctantly, I offered to remain on board in the event we failed locating a suitable long-term replacement.
Steve wouldn’t change his mind. I was unsure why. Not trusting my judgment any longer, being totally convinced of his own, honoring an already-made commitment, or was he just plain playing boss? The next step for me was to invite the chosen one to my—actually his—fiscal year ’02 planning meeting. A good test for the candidate. We prepared well, and as he showed up, I told my direct reports what Steve’s intent was. The meeting went well, but my guys were not impressed by the newcomer’s attention to detail or by his general business sense, echoing my German sources. Back in Redmond, I reported our collective observations. There was still time to reconsider, but after a couple days, with rumors now flying, my boss announced him as the next leader of the OEM division. I promised a smooth transition. Richard Fade, sidelined, did not mind remaining steadfast and cooperative.
Resolve and Fortitude : Microsoft's ''SECRET POWER BROKER'' breaks his silence Page 31
