In short, Google is aiming to render desktop software irrelevant. To thwart them, Microsoft needs Windows to do things that a browser can't--or do the same things significantly better.
Interestingly, if Microsoft wants some tips on how to do this, it might want to look toward Apple. Essentially, this has been Apple's challenge all along: make the Mac experience enough better than a generic PC that it is worth the added cost.
The Mac's resurgence came when it had a strong OS--Mac OS X--combined with iLife applications that really nailed the experience for the tasks that people wanted to do on their computer at the time.
This is an area where Windows has been languishing in recent years. Although most people wouldn't want to give up their favorite desktop applications (Windows or Mac), the Web has been gaining ground. Even areas that were once squarely in the desktop's domain--such as photo editing, productivity software, and personal finance--are making their way onto the Web. What Windows really needs is a new generation of killer apps.
Microsoft also has to do something that Apple doesn't--aim for the masses. Part of Apple's success story has been about choosing its battles and accepting that it can't win everywhere. The Windows model depends on ubiquity, so it needs answers with nearly universal appeal.
One area where Microsoft has been investing is around the area of doing the same things better. Its focus on touch screens in Windows 7 is an example of this. Although multitouch is likely to remain a niche in the short term, it shows the power that a desktop interface can have.
Microsoft also needs to minimize the downsides associated with Windows. On that score, Microsoft has made significant strides with Windows 7. The operating system boots quicker and behaves better than its predecessor.
On the Office side, Microsoft needs to create software that is enough better than Google's that companies want to pay for it.
Next week, Microsoft is expected to talk more about Office 2010, the next version of Office, which is due out next year. Microsoft is taking a two-pronged approach.
CNET News Poll
First, it is taking Google Apps head-on with lightweight browser-based versions of Word, Excel, PowerPoint, and OneNote that can run on Safari, Firefox, and Internet Explorer.
It will offer them to consumers via its Windows Live service--a service that today is free--and businesses will also be able to give the browser-based apps to their workers.
But Microsoft is also doing more on the desktop, adding in the kinds of features it hopes will make the Office suite worth paying for.
The path for Microsoft is clear. The big question, though, is whether Google will be able to be "good enough."
Microsoft has some time, but not a ton. Google's operating system won't even arrive on PCs until the second half of next year. Plus, for now, Windows has the advantage of legacy application support--i.e., businesses and consumers want to run their existing programs. But to stay in front for years to come, it will have to do better than that. It needs to figure out--and quick--the next set of tasks users want to do with their computer and how to make those tasks demonstrably better on a PC.
The company also has another option as well. It can work on Windows' successor. It could be that it needs a lightweight browser-based OS of its own.
Indeed, the thinking beyond its Gazelle research project is that the browser needs to be more like an operating system. In that case, the browser doesn't actually take on the operating system's complete role, but rather relies on Windows. However, Microsoft has other operating system work under way as well, including its top-secret Midori project.
My guess is Microsoft will take both approaches, but hold off on the latter unless and until it needs to. That's pretty much what Microsoft has done with Office vis-a-vis Google Apps. It was only after large business customers started threatening to go to Google Apps that Microsoft conceded that it needed to offer full-on browser apps.