这并不是什么新奇之论。几年以前，丹尼•苏利文曾经在他的博客“搜索引擎天地”（Search Engine Land）上撰写过一篇博文，虚构了谷歌拆分的情景。这篇文章发表时，适逢谷歌收购了DoubleClick公司，有些人担心这起收购会导致谷歌在搜索引擎业占据垄断地位。去年，一个名叫消费者看守者（Consumer Watchdog）的组织就曾以反垄断为由，要求谷歌进行拆分。
尽管在在线广告领域，谷歌显然面临着新的竞争，但同时谷歌也正面临着来自联邦贸易委员会（Federal Trade Commission）和美国司法部（Department of Justice）的双重反垄断调查。如公司自行拆分，则可以避免陷入反垄断诉讼。不过谷歌应该按自己的实际情况进行拆分——让各个部门专注于自身愿景；使有前景的一些小项目不被宽泛的企业文化所吞噬；给员工更多的自主权，同时提供坚实的激励。
What is it going to take for Google to get its mojo back? Google in the spring of 2011 is a far cry from Google in its startup days. Facing competition from younger companies like Facebook, Google is spending more to hire new talent, and the resulting weight on first-quarter earnings pulled its stock down 5% in aftermarket trading Thursday.
Google's (GOOG) new CEO Larry Page has a big task before him. And so far he's been busy. Page announced plans to reorganize the company -- giving engineers more control, rolling out a new bonus program and finding a place for itself in the social web. So far, those changes have drawn mixed reviews.
Most importantly, the changes overlook the hard truth that, if Google is to stay innovative in shaping the web, it needs to take much more dramatic steps than redrawing org charts or using bonuses as cudgels. Google can return a genuinely entrepreneurial culture and nurture innovative ideas, but it would mean radically restructuring the entire company.
It would mean splitting Google up.
This is hardly a new idea. A few years back, Danny Sullivan penned a fictional scenario of a Google breakup for his blog, Search Engine Land. That came in the wake of Google's purchase of DoubleClick, which left some worrying the combined company would create a search monopoly. Last year, a group named Consumer Watchdog called for a Google breakup on antitrust grounds.
Google is facing antitrust investigations from both the Federal Trade Commission and the Department of Justice. A breakup could head off any antitrust suits, even though Google is clearly facing new competition for online ads. But Google should consider a breakup on its own merits -- to keep units focused on their visions, to keep small but promising projects from being swallowed by the broader corporate culture, and to give employees more independence as well as strong incentives to deliver.
A corporate restructuring could happen in any number of ways, some more aggressive than others. For example, Google could become a holding company controlling the majority of voting rights in a number of subsidiaries. Shareholders of Google's publicly traded stock would continue to own shares in the search and display businesses, which accounts for the majority of its revenue.
The Google holding company would also own majority shares in private subsidiaries, each representing an initiative like Google Docs, YouTube, Gmail, a mobile subsidiary composed of Android software and mobile ads, and so on. Employees would receive options in their subsidiary, with an eye to IPOs when they become profitable enough. Provided the IPO markets remain friendly in coming years, this could be a strong incentive for these initiatives to deliver profits.
Another subsidiary could be focused on what is now Google Labs, an incubator of innovative projects funded by Google early on. Once employees nurture ideas into projects in Labs that grow into viable, standalone companies, they receive founders shares in that new startup.
In short, Google the holding company would oversee a keiretsu -- a network of independent companies with interlocking interests, infrastructure and shareholdings. Many of the independent companies may not be, strictly speaking, startups but they would be a lot closer to startups than what Google offers now.
Software engineering is at heart a creative profession. Success isn't measured through box-office sales or the number of books in print, but in how people respond to your work, how they incorporate it into their daily lives. Of course, such intangibles are hard to quantify, so in Silicon Valley's venture capital-steeped culture, achievement is often measured through financial metrics: revenue growth, profit margins, market value.
This isn't to say that all engineers are driven purely by avarice. Garden-variety greed may be as common in Silicon Valley as it is elsewhere, but financial success is also an indisputable yardstick for achievement. In 2010, nothing silenced Facebook's critics quite as effectively as reports of a $50 billion valuation. Google's market cap has vacillated between $130 billion and $180 billion for the past five years.
Beyond compensation, a Google breakup could also address some of the cultural issues that former executives and engineers have complained about. Promising ideas are less likely to get lost in the company's conglomerate morass, as happened with acquisitions like Dodgeball. Google could even bring outside startups into its fold without buying the entire company, but simply by investing in a stake big enough to incorporate the startup into its keiretsu.
If these aren't enough reasons for a breakup, here's one more. If Google doesn't do something dramatic, it risks an inevitable evolution toward the bloated, sedentary life of a tech conglomerate. It will become another Microsoft (MSFT), trying to shape the Internet into an outdated vision of its own imagining, rather than lithely adapting to all the surprising ways the web evolves on its own. Which is what Google is best at.
Or was best at, back when it acted like a startup.