Microsoft's Windows dilemma

Here are all the talking points you'll hear about Kevin Johnson's departure as the chief of Microsoft's sprawling Platform and Services Division — and what to say about them. The failed Yahoo bid killed his prospects of becoming Microsoft's CEO. Perhaps, but Steve Ballmer, who is more to blame for the Yahoo debacle, wasn't going anywhere, and Johnson may not have been prepared to wait. Johnson was charged with competing with Google in search and advertising, and he failed. And you would have done any better? Facebook took Microsoft for everything it's worth in striking its deal for Microsoft to invest and sell ads on the social network — and that's Johnson's fault. True enough, but Microsoft's $240 million investment is pocket change for the software giant. Enough with the cocktail-party chatter. Here's why I think Johnson really left.

It all comes down to Windows — and Windows Live, Microsoft's attempt to extend its monopoly operating system online. The basic strategy of combining the Windows operating system with Microsoft's online-media division, which created Johnson's job, was flawed. The skills involved in coding an operating system and creating websites and selling ads are fundamentally different; mastering them all would be beyond any executive's reach. Johnson was a talented executive, most agree, and his departure is a loss to Microsoft. But succeeding at his job would require focus, and a smaller ambit to his assignments. He needed to have a bigger job; and Microsoft needed someone with a smaller job.

The reorganization following his departure leaves things in a muddle. Windows will still be tied with some online services, under the "Windows Live" rubric; other online operations will be in a separate division. But what goes where? Hotmail, for example, was recently rebranded as a Windows Live product; but it's hard to see Microsoft's online operation giving up such a reliable traffic generator to the Windows group.

In Johnson's wake, expect more infighting, more reorganizations, more mass rebranding campaigns — everything, that is, except the kind of purposeful progress Microsoft desperately needs to compete with Google.