In the wake of last week’s Microsoft-for-Yahoo rumors (and desperate hopes from the likes of me that the rumored deal would fall through), readers questioned my theory that a Microsoft acquisition would kill both MSN and Yahoo. I argued that Microsoft was already spread too thin, competing with Oracle/IBM on one side and Sony/Apple/Google/Time Warner on the other. Pshaw, some readers said: Just look at GE. GE sells jet engines, sitcoms, locomotives, and credit cards…so surely Microsoft can handle a few different flavors of software. Today, John Battelle has picked up the same theme.
Here’s the difference between Microsoft and GE, at least as it pertains to the Internet. One reason Microsoft has struggled for twelve years with its Internet business is that it has always managed the business with an eye to protecting and/or augmenting its core desktop monopoly. Microsoft’s competitors, meanwhile, have approached the Internet business with an eye to doing whatever is best for the Internet business, desktop monopoly be damned. And now that it is clear that the Internet business is going to compete with the desktop business, Microsoft finds itself in an even stickier situation: How can it do what it needs to do to win in the Internet without hastening the demise (or at least slowing the growth of) the desktop cash cow?
GE can compete in dozens of different businesses because it is a true conglomerate: Each division is free to do whatever it needs to do to win (and the divisions also aren’t that competitive with one another). At Microsoft, meanwhile, all the sideline divisions exist to protect or augment the major businesses, and Microsoft’s Internet team has to work within this system. For Microsoft’s Internet business to have a real chance to win, it has to be able to launch wholesale attacks on Microsoft’s core business. And it’s hard to see that happening within the Redmond corporate structure as currently defined.
In short, the reason Microsoft can’t “be GE” is that GE is a conglomerate and Microsoft isn’t. If Microsoft wants to become a conglomerate, fine, but this will require a wholesale DNA transplant. And by the time it got one, the Internet opportunity (and dozens of others) would have long since passed it by.