Who Fatally Wounded Microsoft? It Was Bill Gates.

Why has Microsoft found itself in third place, behind Apple and now Google, in the mobile segment of computing?

When you stop to think about it, it’s shocking.

At one point, Microsoft crowed about how Handheld PC, Palm-Size PC (then Pocket PC), then Windows Mobile, then Windows Phone Edition, would rule the world.

Microsoft owned the desktop, so domination of mobile technology was only logical. It wouldn’t surprise me that Microsoft considered it was just about its birthright too.

So how is it that Apple is dominating mobile devices? How is it that Apple is set to revolutionize all of computing within a few weeks with its iPad?

The seeds of Microsoft’s destruction were planted over a decade ago, in a decision made by Bill Gates himself, when he was still the final word on all things Microsoft.

After the break, an excerpt from Breaking Windows: How Bill Gates Fumbled the Future of Microsoft by David Bank, published in 2001.

I won’t set it in Blockquote to make it easier to read.


Many executives inside Microsoft thought Ballmer’s reorganization didn’t go far enough in untangling the company’s many charters. Harel Kodesh, an Israeli-born engineer in charge of Microsoft’s efforts on non-PC devices, had undergone an almost religious epiphany. Kodesh had helped develop Windows CE for “consumer electronics” such as handhelds, cell phones, set-top boxes, and the voice-activated “Auto PC” for car dashboards. At first, Kodesh subscribed to Gates’s notion that, as with PCs, a uniform operating system across many different electronic devices would attract software developers, who generally target their efforts on the highest-volume platform. Because Windows CE used the same tools and Win32 interfaces as Windows for the PC, Microsoft’s dominance in the PC market would give it additional leverage in the non-PC world. Windows CE was a key element in the Windows Everywhere strategy.

But it hadn’t worked out that way. Microsoft seemed permanently behind in the handheld market, where Palm was established as the software of choice for “personal digital assistants.” And Microsoft was being outflanked by a consortium of cell phone makers, led by Nokia, which established the Symbian joint venture with a British software maker, Psion, to create a specialized operating system far Web-connected cell phones, a potentially huge market.

Kodesh realized that he, and Microsoft, were both on the wrong track. The vaunted Windows leverage had turned out to be a liability. Kodesh had met Innovator’s Dilemma author Clayton Christensen at a symposium at Harvard. He came back to Redmond with a hundred copies of the book for his thousand-person group. The “information appliance” was exactly the kind of disruptive technology Christensen warned about, Kodesh argued. The new devices needed to be small, cheap, and most of all “cool.” That meant they should be precisely tailored to the needs of particular consumers for particular uses. Microsoft’s software was like a Swiss Army knife, adequate for many things but optimized for none. It was the only software that could work in handhelds, phones, and set-tops, but it wasn’t the best software for any of those categories.

To Kodesh, Microsoft’s initiatives were driven by the dictates of the company’s technology, rather than the wishes of consumers. An electronics maker like Casio could focus on delivering a product with just the right features for consumers; Microsoft insisted on squeezing PC applications like Word and Excel into the miniature devices.

Kodesh had a radical idea. Microsoft should keep Windows CE for handheld devices but start from scratch on new software for cell phones and all other non-PC devices. Microsoft should start from the other direction, asking manufacturers what they could build for, say, ninety-nine dollars and offering to give them the software they needed, rather than only the software the company already had. The choice was familiar: Would Microsoft embrace the new market for non-PC devices, create an autonomous company-within-a-company to tackle it, and do anything it took to win? Or would it shackle the new devices to its traditional businesses?

In September 1999, Kodesh wrote a memo to Gates and Ballmer under the heading “Starting from Scratch.” We need to kill Windows CE for those categories, he argued. Win32 is not an advantage; it’s a tax on device design. It served to further Microsoft’s strategy but not to help consumers. Given all their other alternatives, electronics manufacturers wouldn’t pay the tax. Kodesh wanted to take a small group of developers and work solely on developing the best software for information appliances, unconstrained by the needs of the rest of the company.

Gates rejected the suggestions. “It’s very disappointing you feel that way,” he told Kodesh. “We don’t have time to start from scratch.”

Kodesh left Microsoft several months later. [pages 179-180]

— end of quoting.

37 responses to “Who Fatally Wounded Microsoft? It Was Bill Gates.

  1. The idea of shackling an old business to a new business rings true in so many industries right now.

    Watching media companies try to get the web to pay for operating a printing press, for example.

    A culture of innovation is something that appears to be difficult for established companies to maintain. Which is why there’s so much attention on Google and Apple (two companies with radically different but equally effective methods of maintaining innovative company culture).

  2. It might be that PART of the reason Steve Jobs is so good is because of his “years in the wilderness” (working with NeXT, PIXAR) after being ousted from Apple in 1985

  3. After reading your great post, it seems to me, that for such a smart man, Uncle Bill certainly can act pretty dumb at times!

  4. what did Harel Kodesh develop since he left MS? or was he just the lone voice in the crowd, which 10 years after the fact, has resonance?

  5. Wow. That is the smoking gun.

  6. I find it increasingly difficult to care what happens to Microsoft. It doesn’t make me angry. I couldn’t care less. It’s been hawking vaporware and general mediocrity for too long.

    Let them go, folks. After all this time, they still can’t make decent software. They probably never will.

  7. Pingback: The day the Windows died « Asymco

  8. Wow. I had no idea about that little twist in the Microsoft saga, which is kind of saying something.

    I worked in Microsoft’s Windows CE group from 1996 to 2008. I lived through any number of changes in direction, strategy, and so forth for the group. I was not the recipient of one of Kodesh’s 100 copes of Innovator’s Dilemma, though; I’m sure those went to the group’s program managers and developers. I was a documentation guy.

    I remember Harel Kodesh well. He was a couple of levels up, org-wise, from me but he was a nice guy. Very smart.

    In hindsight, of course, it’s easy to see that he was probably right in his memo. Of course, Gates is the one who’s supposed to be the visionary, seeing years further down the road than anybody else, at how this stuff will play out.

    It kind of makes me sad to learn that Bill dropped the ball on that one. I mean, I can see why: what Kodesh was proposing would have been hugely expensive, with a poor risk/reward ratio (at least on the face of it). With Ballmer no doubt involved in the decision, I can see how Gates ended up saying no. But it’s still sad.

  9. Microsoft has a history of producing mediocre software. It achieved success despite this. Its business model made PCs a generic commodity. This appealed to corporate customers who desired uniformity for ease of administration, as well as to developers who wanted to focus their efforts on a single platform. Businesses were initially the largest customers of PCs at first.

    As prices came down, people began purchasing home computers that could run the same programs as used at work. They did this to insure document compatibility, and avoid having to learn anything new.

    Once PCs had infiltrated the home, a growing number of games served to further cement the PC’s dominance.

    In the world of handheld devices, Microsoft was never able to establish themselves in the same way. Handheld devices were too small to effectively run office productivity applications. People on the go carried laptops for that.
    But the killer-app for smart phones turned out to be email with Blackberry emerging as the corporate darling.

    Now, with the iPhone leading the way, we are moving into the next phase in handhelds where the internet is the killer app. It is too late for Microsoft to “own” this market with proprietary lock-in. They will have to compete based on bold innovation and quality… something they are not use to doing.

    Kodesh was right. Had they done that starting with a blank sheet of paper ignoring the Windows legacy and preconceived notions of user interface, creating an OS specifically optimized for handheld devices, maybe it would have been Microsoft not Apple, leading the pack.

  10. I’ve said it before and I’ll say it again. Gates was never the visionary that Jobs was (and is). MS has always been (and might always be) a bandwagon jumper using its financial and market-position clout to beat down, drown out, or take over smaller, faster, more innovative players. That strategy, though, looks like it’s ultimately going to spell the beginning of the end for MS’s dominance.

  11. Thanks for the shout out!

  12. While I agree wholeheartedly that Microsoft dropped the ball by trying to make every device into the same computer, regardless of the user, I feel I should point out that Apple uses the same underlying OS and a similar-when-not-identical set of APIs between the iPhone, iPad, AppleTV, and Macintosh computers.

    The problem isn’t that Win32 was everywhere, the problem was that Win32 wasn’t any good (it’s overly complicated, but most things were at the time), and they tried to shoehorn Windows 95 conventions into Windows CE (a ‘Start Menu’ for example).

    On the other hand, Apple, thanks to NeXT, has the Cocoa (née NeXTStep) framework which is simple to work with and generally uniform across platforms, and they understood that a different computing environment needs a different interface and philosophy.

  13. It seems as though HP is trying to stuff a full version of Windows 7 on their tablet and these geek users think this is such a great idea. The average consumer probably wants nothing like that since there is no call for it to do a few light tasks. A company should build devices with just enough of an OS to get the job done it was designed to do. Forget about trying to shoehorn Microsoft Office into handheld devices. Yeah, the geek users are happy at the thought of stuffing desktop apps into tablets and smartphones, but it’s impractical and a poor idea for a low-tech consumer user to deal with.

    Microsoft seems unable to move away from legacy equipment and support. Why a smartphone would need to support all that old stuff, I’ll never quite understand. There is a place for Microsoft and Windows in the enterprise, but it will be a burden when all the other companies can quickly move into the future without having to support their long, gone past.

    It’s going to take a long, hard climb for Microsoft’s mobile division to catch up with the rest of the smartphone industry no matter how much money they throw at it. I think they’ll just continue to follow Apple as closely as possible.

    I don’t see the iPad replacing the Windows desktop. I’m unable to see the demise of the standard Windows PC since it’s been around for such a long time.

  14. Thanks for the great book. I read it ages ago and have carried that excerpt in my PDA for years. I happened upon it this morning and thought it’d interest everyone. I hope it leads people to buy your book, which is a must-read.

  15. Pingback: Who Fatally Wounded MSFT? | DASH7 Wireless Sensor Networking Blog

  16. “We don’t have time to start from scratch.”

    My bet is he would say the same thing today.

  17. Great post. I must get Breaking Windows. Wish it were available in e.

  18. I’m still amazed Bill didn’t insist that the Xbox run Windows Vista.

    It’s been a long time since I read Breaking Windows, but it’s very well done. I couldn’t put it down.

  19. The situation described is so similar to IBM’s dillema when they had to come up with a PC. The Skunkwerks works. :-)

    Google Books has a lot of the book online.

  20. Hindsight is 20/20

  21. Microsoft wasn’t operating in a vacuum, so that’s a poor excuse.

    But an even bigger failure was Palm, which was inept in ways business schools will build entire curricula on!

  22. Part of the reason for this strategy is because it’s what worked for MS in the past. Back when they work working on OS/2 with IBM, they had shipped Windows 3.0 and it took off. IBM told them to ditch Windows and Bill said no. He realized that it was the application developers that made the platform, and their intellectual investment in an API (Win16, in this case) established momentum behind the platform. This is how he beat IBM.

    Shortly thereafter, they extended the API to 32-bits on NT and Windows 95 and the strategy worked–they had a huge number of developers writing 32-bit apps when Windws 95 launched.

    It’s working today with Xbox, which is just a special build of Windows XP. Developers can build on Windows, using familiar interfaces, like DirectX, then port to Xbox with minimal pain (relatively speaking).

    So their strategy of Windows everywhere has worked for the most part.

    Steve Jobs is employing a similar strategy. The iPhone runs the same kernel as OS X. It uses the same programming tools. That’s why you can only write apps on a Mac.

    The reason the iPhone is winning is not the API or the “tax” Kodesh was talking about. It’s user demand. Windows phones aren’t cool (at least until v7). iPhones are.

    The issue of handset adoption of WinCE, etc. has more to do with the competitors, including Symbian and Linux, the capabilities of each, and the cost of licensing vs. WinCE.

    So there are multiple forces at work. Of the three I mention (user demand, handset manufacturer/carrier OS choice, and OS/API as it relates to developer momentum and tools), I would argue that the API consideration is the least significant. Developers go where the users are and where they can make money. In the old days, you made money writing Windows apps. Now you make money writing Web apps and iPhone apps.

  23. Very interesting and telling read. I’m commenting because I see many comparing Bill and Steve and I don’t think you can really do that.

    Steve Job is a visionary and an evangelist period. He envisions how things are going to work and get someone else to build it to his vision.

    Bill Gates is a business man with practical business sensibilities. He built things that could be commoditized because he saw computers and technology as tools and not as a lifestyle.

    Apple right now is influencing the cellphone the same way it did with the Apple IIs and personal computers in the 80s. But they’re repeating history with their ownership and integration of hardware and software and their efforts in controlling both.

    Ultimately, Apple lost to IBM PC compatibles that flooded the market and were cheaper and could run any OS; it wasn’t just Microsoft back then, there were a lot of different OSes available. I see the same happening in the future. Google will beat Apple with Android because Android works on open hardware.

    Apple leads by innovation and always has. Microsoft leads by being practical and always has. When Steve has gone, which I consider a tragedy for technology, Apple will flounder because there’s no one to replace him. Microsoft has already handed over the reins once and it’s still running. Who’s going to be at the finish line? Not Apple.

  24. “Why has Microsoft found itself in third place, behind Apple and now Google, in the mobile segment of computing?” It never fails to amuse me how frequently Nokia gets totally ignored in this type of analysis. http://wp.me/pFuDI-U

  25. Microsoft execs basically copped to this a few weeks ago, at least according to the NYT.

    “In a counterintuitive moment, Microsoft’s executives called the software a departure from traditional PC software, saying past fixations on Windows traditions had stifled the company.”


  26. It’s because Nokia Does Not Matter. Nokia’s giant market share is in low-end dumb and feature phones. Nokia is basically a commodity phone pimp. They have no massive presence in smartphones despite what Europeans might be able to buy on an unlocked basis. Nokia also cannot do anything approaching good software. I had the Nokia 770, with Maemo. It was utter and total unbelievable shit — and has remained so. Nokia is in a death spiral and the only two people who seem to understand this are Nokia and me.

  27. Apple’s huge misstep back then was pricing the 1984 Mac at a whopping $2,500. That hurt them massively. And the LISA was a jaw-dropping ten grand. The Mac, touted as being “For The Rest Of Us,” was really priced only for “The Rich Of Us.” Jobs did not have total control of Apple as he does now. Had he, the LISA would have never come out.

  28. >>>Windows phones aren’t cool

    If you’re betting it’s only “cool” that accounts for the ginormous success of iPhone, you’re mistaken. Cool has a limited shelf-life. Ease of use is the issue.

  29. Pingback: Top Posts — WordPress.com

  30. “They have no massive presence in smartphones”
    Actually, Nokia increased its marketshare in the Smartphone segment to 40% in 4Q 2009.
    I’d say that’s a pretty significant presence. It’s not about Europe or the US, it’s about India and China.

  31. >>>it’s about India and China.

    And those are countries where you expect expensive smartphones to sell?

  32. Are you kidding me? YES!

  33. We’ll see.

  34. The iPad ” is set to revolutionize all of computing”?

    Give me a break. I’m calling out this ludicrous claim for what it is. You can’t make a claim like that without significant substantiation.

    It’s a double sized iPhone. And that’s going to revolutionize “all of computing”?

  35. Hey, I can’t help stupid fucks like you. Maybe reality will bitchslap you around and you’ll catch on. If not, Darwin will take you out. No loss.

  36. Pingback: 7 Examples of Ego Leading to Business Blunders | Money and Risk

  37. A simple NPV technology decision on what Kodesh recommended would have proved Kodesh right, despite starting over