Google (NASDAQ:GOOG), already burdened with a reputation for wanting to be all things to all people -- a search engine, a cell phone operating system supplier, a renewable energy investor, etc. -- now wants to be a venture capitalist, The Wall Street Journal reported last week.

David Drummond, Google's top attorney and senior vice president for corporate development, is to lead the new unit. Bill Maris, a tech entrepreneur who's a former associate of Sergey Brin's wife, Anne Wojcicki, will reportedly help to establish the fund.

The Big G a VC?
The idea seems laughable at first. Focus is a central theme of American capitalism. If outsourcing is our religion, Google -- which eschews specialization in favor of broad experimentation -- is a corporate sinner. So culturally distinct is this way of thinking that skeptics assume that upstart search specialists such as Powerset and Cuil, which debuted with fanfare last week in spite of nagging technical problems, can take share from Google when Microsoft (NASDAQ:MSFT), Yahoo! (NASDAQ:YHOO), and IAC's Ask.com have all tried and failed.

To me, Google as VC is as lucid an idea as Intel (NASDAQ:INTC) as VC. All managers are investors; how they invest, and in what, is what separates them. Some firms are too small to invest in anything but urgent capital needs. Many Silicon Valley start-ups fit that description. But others, like Google, have too much cash. They're like Berkshire Hathaway in that they have more capital than ideas; $12.7 billion more, to be precise. Should CEO Eric Schmidt simply let these billions sit idle?

Perhaps. Certainly he shouldn't act quickly. But corporate venture investing, when done well, can be a catalyst. Witness Intel Capital. SEC reports show that the value of its equity holdings -- which included would-be rebels Clearwire (NASDAQ:CLWR) and NYSE Euronext (NYSE:NYX) -- grew from $1.02 billion at the end of the 2007 first quarter to $1.23 billion as of March, and with 21 fewer holdings. That's a 20.5% year-over year paper gain in a market that's caused nothing but trouble. Intel blew away second-quarter earnings estimates last month.

But Intel appears to be the exception. Corporate VCs funded just 7% of all venture capital deals in the first half of 2008, down from 8.4%, according to PriceWaterhouseCoopers and the National Venture Capital Association. So even if Google thinks it's ready for the VC business -- and investors like me approve of the idea -- it's likely going to be tough to win deals.

How can Google build a fund that start-ups will want to do business with? Here are three ideas:

1. Invest for returns, not research. Some firms invest in start-ups to control them. Google should avoid that temptation and be an advisor whose only goal is to grow the business and create a successful liquidating event for the founders, employees, and its investors. If the result is still a buyout, so be it, but that should be nothing more than a happy coincidence. Disney's (NYSE:DIS) Steamboat Ventures can be a model for this approach.

2. Partner relentlessly. Some of the best deals in the history of the venture capital industry have been collaborative. Early funding for Google, for example, came from two of Silicon Valley's finest: Sequoia Capital and Kleiner Perkins Caulfield & Byers.

3. Move on quickly. Top investors cut losers when the underlying business fails to deliver on its central promises. Google can't allow its VC money to act as an R&D fund that tolerates losses in the quest for the Next Big Thing. Rarely is the Next Big Thing actually the Next Big Thing, and thus, R&D should remain its own operation.

Will Google's billions be up for grabs? I won't object if so. The Web is changing fast, enabling bold ideas and even bolder business models, and Google is helping lead the shift. But it won't always be that way. During those times, Google the investor, if not Google the e-gadget geek, will still have a chance to profit.

Get your clicks with related Foolishness: