What Google is really up to

The key to understanding what Google is up to will be found by looking at Google's own business. Its horizons are much broader than what happens on the desktop.
Written by Phil Wainewright, Contributor on

I think the stock market's reaction to yesterday's Google-Sun announcement gets it about right: Google down $7, Sun up a penny. As many bloggers have already opined, (and as I had feared it would be), it was all a bit of a disappointment. Maybe Sun will benefit from this link-up, but it's hard to see what added value it's going to bring to Google. Sun's track record in on-demand applications and services is well documented, and it doesn't look good.

The worry is that CEO Eric Schmidt is about to draw Google into the kind of anti-Microsoft obsession that in the past has dragged down both his previous employers, Novell and Sun. But perhaps that's exactly what Google wants us to think, says Microsoft watcher Joe Wilcox. As noted by Dan Farber yesterday, he believes the tie-up with Sun is simply a tactical PR move to turn the tables on Microsoft (down 52c yesterday):

"[Google] can ... cause Microsoft to waste resources, be distracted chasing Google, while the rival executes brilliantly where the Redmond folks aren't looking."

This certainly makes a lot more sense to me, and it squares with the emerging story that the alliance was first suggested by Sun's engineering team and then taken up by top management.

But nothing has justified the frenzy of expectation these past few weeks that Google was about to deliver the knock-out blow either to Microsoft's dominance of the desktop PC market or to the entire telecoms industry's dominance of, well, telecoms.

The key to understanding what Google is up to will be found by looking at Google's business, not at its competitors'. There is no doubt in my mind that Google (or some other Web 2.0 company if Google messes up) will one day be bigger than Microsoft. But it won't be by seizing Microsoft's own markets, just as Microsoft grew bigger than IBM without getting into the mainframe business. All the ballyhoo about bolstering OpenOffice against Microsoft Office is just a distraction. Google's horizons are much broader and the real battle won't be fought over desktop applications as we know them. Some people will still use Office in twenty years' time, just as some people today still ride on railroads or canals. Superceded technology never disappears, it just declines into obscurity.

As everybody knows, the core of Google's business is search — or more accurately, not the act of searching itself but the search results that come back: what used to be called information retrieval. Google's business model revolves around helping people find better answers faster.

To do that it needs to embed its search and retrieval technology ever more tightly into each user's individual information processing. Yes, of course that means being present on the desktop, but even more so on the palmtop, the cellphone and wherever else web access is taking place.

Teaming up with Sun wins a lot of PR, gets Microsoft to waste cycles on competitive analysis, and puts Google on a few more desktops via Java runtime downloads. That's enough of a win to make the alliance worthwhile, but the real action is happening elsewhere:

  • exploring the viability of funding city-center Wi-Fi services with location-based advertising
  • buying up fiber to cut vital milliseconds off the time it takes to analyse the user's environment and deliver appropriate contextual information
  • researching client technologies that can help cache code and information locally to reduce download waits and continue to service users during those moments when they're disconnected
  • getting ready, as Adam Bosworth mentioned in his presentation at Dreamforce the other week, for the moment when "mobile VoIP phones suddenly become pervasive."

All the time, Google will, to borrow Bosworth's phrase again, 'run like mad' , not to deliver a rehashed version of yesterday's applications or devices, but to react intelligently to newly emerging user needs.

Editorial standards