Google's decision to build a PC operating system could be a master stroke or a colossal blunder, depending on whether the company has the resources that such an ambitious and long-term undertaking will require.
Google plays in a variety of extremely competitive markets, serving a broad scope of demanding customers and partners. Although developing an operating system could yield big rewards, it could also distract the company and make it more vulnerable to rivals.
Of chief concern is Google's continued reliance on a single type of advertising for most of its revenue, despite efforts over the years to diversify its business.
Google still makes most of its money from search pay-per-click text ads, a market that it dominates but where loyalty from consumers and marketers is thin, making the company vulnerable to the development by a rival of a significant technology breakthrough.
In short, if someone built a better search mousetrap -- as Yahoo, Microsoft, Ask.com and a host of smaller players are trying hard to do -- Google would suffer a sudden drop in search usage and consequently advertising, crippling its finances.
Google's attempts to build alternative revenue streams from display advertising remain nascent, despite the costly acquisitions of ad services provider DoubleClick and video-sharing leader YouTube, two properties Google considers key to this effort.
Bold initiatives to provide print ads to newspapers and spots to radio stations both failed. The company continues its attempts to build a TV advertising business.
Google executives are the first to admit that the company dominates the Internet search market because it toils long and hard every day to continually improve its engine technology.
Yet, not content with waging battle every day in search, Google also provides enterprise search and business collaboration software, competing against the likes of Microsoft, IBM, Cisco and Autonomy, and trying to win over business managers, IT managers and CIOs.
As if that were not enough, Google has set out to court Web application developers by opening up APIs to a wide variety of its consumer and enterprise products.
In addition, Google maintains many non-search services for consumers, such as the Orkut social network, photo management service Picasa, Reader feed manager, Knol encyclopedia, Checkout online payment system and the iGoogle personal home page.
In recent months, Google has shut down or stopped supporting several products, including Google Video, Google Notebook, microblogging service Jaiku and mobile social network Dodgeball.
"My feeling is that Google needs to stop announcing things and instead execute on completing them," said Rebecca Wettemann, an analyst at Nucleus Research who focuses on enterprise software.
For instance, Google could make many enhancements to Apps, its hosted collaboration and communication suite for businesses, which just yesterday had four key components exit the beta, or test, stage, she said.
Google's sudden announcement it is taking on the massive task of building a PC operating system makes it seem like the company has a short attention span. "They pick something up, get excited about it and work on it until they find another shiny new object they want to play with," Wettemann said.
The perception of a company with a scattered focus doesn't go down well in the enterprise software market, where IT managers and CIOs like their vendors to have a clear direction, she said.
Google said the Chrome OS project is motivated by the company's conviction that the world lacks an OS built from the ground up for people who heavily use the Web.
The Chrome OS, which will be open source, based on Linux and built to be "lightweight" and very secure, will thus be a companion to the Chrome browser Google unveiled in September 2008.
Still, developing an operating system is a major endeavor, said Burton Group analyst Guy Creese, who once worked as an OS product manager.
"I don't doubt Google's ability to create an OS if it decides to; I just question whether it wants to add that work to its plate," Creese said via e-mail.
Google should consider keeping the scope of the OS narrow, such as by limiting it to netbooks, which would reduce the testing universe for hardware compatibility and the number of device drivers to develop and manage, Creese said.
"I suspect Google is underestimating the tedious work involved in making sure the OS works with many different devices," he said.
In its announcement late Tuesday, Google said the Chrome operating system will initially be aimed at netbooks, the first of which will begin to appear for consumers in the second half of 2010. However, it also intends for the OS to run on full-sized desktop systems.
If that's the case, Google should position it as an alternative in dual-boot machines, Gartner analyst Ray Valdes said. End-users would have the option of launching, say, a full-featured OS like Windows or the lightweight Chrome OS, depending on their needs at the time, Valdes said.
Wading into the "swamp" of supporting myriad devices, drivers, graphics cards and the like at the level Windows does would be a mistake and a nightmare for Google, he said.
In fact, Google probably has about 70 percent of the Chrome OS ready, based on its extensive internal use of Linux to power its systems over the years, Valdes said.
Given that Google has succeeded so far with its Android mobile platform, creating a PC OS makes sense and is feasible, but a concern would be the company's spotty track record in customer support, said Greg Sterling from Sterling Market Intelligence.
This is a big issue, since Google will have to respond to support queries from hardware manufacturers, application developers, end-users and, if and when it's used in businesses, IT departments.
"Historically, Google hasn't been good at support," he said, referring to regular gripes from consumers and businesses that the company is often slow to respond when its services and products malfunction.
"Google can build strong products, but it has an uneven track [record] with customer service," Sterling said. "Google has the technology and money to build this operating system, but will it perform properly in the area of support?"
Google, which plans to release the Chrome OS code as open source later this year, declined to comment for this story.