Sometimes I think I am the only one that does not see a logical connection between Microsoft's business and "search". Living part time and working in Silicon Valley has given me some perspective into the workings of tech giants though. The initial reaction of a Scott McNealy, Larry Elison, or Steve Balmer to a new kid on the block is to beat them into the dust before they can gain any momentum. Remember poor old Netscape? So Balmer has become obsessed with Google, even though there is not an iota of evidence that Google is challenging Microsoft's business. Sure, Google is using its giant cash hoard ($12B) and its army of smart engineers to launch all sorts of revolutionary web based apps and even threatening to get into the cell phone business. But, has Microsoft lost a sale to Google yet? Is there a big organization that has dumped Word and Excel for Google Docs? No. Yes, Google has $18 billion in ad generated revenue. But that is revenue taken away from newspapers, television and radio. Before Google, did Microsoft see buying newspapers as something important to its business? No. The problem with Microsoft is that they have stopped innovating. Microsoft invented their own market for Windows and successfully dominated the PC industry. They bought into and dominated the desktop productivity tools space. They are still the largest gaming platform (not XBOX, the PC). Are all those things going away? Should development come to a halt? Is ad serving really that big a deal? I think it makes more sense for the investment community to lobby for a break up of Microsoft than to encourage the acquisition of Yahoo! Here are the reasons: 1. While not the most important reason, breaking up Microsoft would settle the ongoing anti-trust action against the company. That would re-open some valuable markets, especially in the EU. (And Thomas Penfield Jackson would be a very happy man!)2. A blind reverence for Windows is hurting innovation. There is no reason to use the core Windows OS on servers, entertainment systems, gaming consoles, or cell phones. Free up each of those groups to innovate and use the best possible solutions for their platforms. 3. There are two ways to fight the move to Software As A Service (SAAS). One is to abandon the old applications and try to convert them to hosted solutions. The other is to make those desktop apps so feature rich, sleek, and easy to use that there is no way SAAS services can catch up. Spinning off the desktop application group is the only way to make that happen. 4. Online services are indeed important. But according to Clayton Christensen, author of the Innovator's Dilemma, rarely do established players invent the next paradigm. For Microsoft to compete in the next evolution of computing they must spin out a group that is not caught up with thoughts of protecting its installed base. 5. Security. This is the point I have made before. If there were separate code bases for Microsoft platforms you would no longer have universal vulnerabilities like RPC DCOM and rendering issues. The other reasons benefit Microsoft shareholders. Increased security in the computing eco-system is a benefit for all. Breaking up is hard to do. It has costs associated with it. When Steve Rosenbush suggested it on May 8th Microsoft had already lost $24 billion in market cap since the Yahoo! debacle began. At $235 billion today it is off another $37 billion in just two months. I would guess that breaking up will increase total shareholder value, not decrease it by $61 billion as the Yahoo! foray has. Failing to break up can be even more costly. I was a contractor at General Motors during the formative Roger Smith days. Roger was a numbers guy and he decided to demolish the text book structure that Alfred Sloan had engineered for GM. Sloan's mantra of "centralized decentralization" had created a mega-company with shared administrative and financial reporting structure but separate marketing and manufacturing organizations for each car division. The central staff would decide the market segment that Pontiac, Buick, Chevrolet, Cadillac, Oldsmobile, and the truck divisions would serve. Each division would design, build and sell its own products. They had to compete with each other. Faced with competition from Honda and Toyota, Roger felt that change was needed. He created two divisions (Big Cars and Really Big Cars). Now multiple brands were manufactured on the same assembly lines. He invested $20 billion in new stamping presses that could have their tools switched out in hours instead of weeks. All of these efficiencies through shared engineering, sales, and tooling were supposed to protect General Motor's market share. Twenty years later we know the story. GM continued to lose market share. Today you could buy all of GM for $5.4 billion. Its stock price is at 1970's levels. The reason? While it is true that Toyota was more efficient than GM, the real survival measure was innovation and GM lost that ability when it centralized. Lesson learned? Microsoft should split up before it becomes the General Motors of the Internet age.