Start-ups have it easy. That may be a politically incorrect statement. The popular image of start-ups has some poor guy borrowing to the limit on all his credit cards so he can wage a desperate battle against Big Ogre Inc. and maybe, just maybe, eke out a living.

Not!

When it comes to innovation, big companies are the ones that really face complications. Start-ups just have to locate a monster market no one else has uncovered, produce a brilliant product, and manage the company flawlessly. Established companies not only have to do all those things, but they also have to fight cultures that encourage nearly everyone to mitigate risk—if not avoid it altogether. (Who dares do anything that might disappoint Wall Street for even a quarter?) In addition, corporations setting up "intrapreneurial" ventures have to overcome not only inertia, but also the objections of existing business units that may be threatened by a new operation's innovations.

In other words, start-up ventures fail only if they mess up. "Intraventures" can fail a lot of ways. For example, they can fall apart if the corporate parent has cultural problems. Even if both parent and intraventure are in great shape individually, the venture can still flop because the relationships between the two are too complex and hard to manage.

What to do? The answer is obviously complex, but we can point to three key areas where large companies typically fall down and explain how to start to address the problems.

We do so based on more than a decade of use of the Bell-Mason Diagnostic. This tool, which involves asking key executives hundreds of questions, initially was used to gather best-practices data on start-ups—how successful start-ups look, along 12 dimensions, at different stages of their development. The tool was then used to evaluate several hundred groups that were seeking venture funding, to predict which groups would succeed. Over the past eight years, the Bell-Mason Diagnostic has been adapted to evaluate the more complex problems that intraventures face and has been used to assess more than 200 ventures.

The tool has helped enough venture capitalists find Davids that could slay Goliaths. It's time to help the Goliaths fight back. Here are the three immutable rules of intraventures:

FIND THE RIGHT PEOPLE. That's not as obvious as it sounds. Lots of organizations find good people for intraventures; they just aren't the right good people.

For instance, when companies pick teams for important projects, they understandably look for people with strong track records. Often, those people are managers. But intraventures don't need many managers; they need doers.

Also, companies often decide that certain people can contribute some percentage of their time to an intraventure. But "fractional" people, no matter how talented, create problems of coordination. Intraventures require people who can devote their hearts and souls to the project.

If companies have the foresight to set up an advisory board for the intraventure, they tend to pick program managers. Someone from finance, for instance, checks to make sure that revenue forecasts are justifiable. What intraventures need, though, are very senior advisers with independent perspective and considerable vision. Start-ups have advisers like this—Siebel Systems, a young software company, began life with Charles Schwab and the dean of the Stanford business school on its board—so how can intraventures compete without having similar sounding boards?

To address the people issues, companies forming intraventures should focus on three functions:

First is the venture board, which should be as close as possible to the sort of board that the Siebels of the world have. Heterogeneity is a must. A liberal sprinkling of outsiders is a good idea.

Second is the venture executive officer. This is the corporate insider who leads the venture board, acts as the chief fund-raiser, and is the high-level emissary inside and outside the corporation. This individual must have enough stature within the corporation to be an effective champion. Often, this person needs to be the parent's chief executive. Otherwise, a company can wind up in the fix that has caught many of the big brokerage firms. They keep saying they're going to move aggressively into on-line stock trading, but the executive sponsors of the intraventures aren't senior enough to overcome opposition by those representing the brokers, who stand to lose commissions. While the firms sort out the political infighting, Charles Schwab & Co. and some start-ups are cleaning up.

Third is the venture general manager. This is a rare bird. He needs to have a bit of a rebel in him, but, unlike the chief executive of a start-up, he also has to be a team player who is savvy about dealing with the parent organization. While the corporate environment and pay structure make it hard to attract true entrepreneurs to these venture GM positions, people experienced in brand management for consumer products seem to make a good substitute because they often have experience bringing new products to market.

HAVE CLEAR GOVERNANCE RULES. Many companies can't decide how to treat a new idea such as electronic commerce. Typically, companies establish lots of programs as experiments. But intraventures need to be established with the idea that they'll become entirely new divisions, if not separate businesses. In fact, the more disruptive an intraventure might turn out to be to the existing business, the more autonomy it needs. Otherwise, the intraventures will be smothered, much as the giant brokerage firms have done with their attempts at on-line trading.

But, with independence, comes the need for clear agreement, upfront, on: how to resolve conflicts with other operations; what corporate assets the intraventure may draw on; what variations from corporate policy will be allowed; what the triggers will be for decisions on subjects such as additional financing, and a host of other things.

This last point is especially important because many companies mistakenly force intraventures to operate on standard, annual budget cycles. As with a start-up, an intraventure's product is ready when it's ready. The venture needs financing when it needs financing. There is simply too much uncertainty associated with true innovation to allow for normal budget processes.

DEVELOP A PORTFOLIO OF OPTIONS. Venture capitalists don't make just one bet on electronic commerce or biotechnology. They build a portfolio, on the understanding that, while most won't pay off, the winners will more than cover the minimal losses on the failures. Corporations need to take the same approach—establishing a whole series of intraventures that explore risky opportunities. As the intraventures progress, companies need to "weed and feed." That means nurturing those that look like they'll succeed, building businesses once success is demonstrated—and killing off, quickly and ruthlessly, those that are going nowhere.

In World War II, even the world's sleepiest organization, the sprawling bureaucracy in Washington, managed astounding feats of creativity. Unleashing "can-do" entrepreneurs like the manic Andrew Jackson Higgins, inventor of the D-Day landing craft, the U.S. rapidly produced everything from Liberty ships to P-51 fighters to instant deep-water ports in the waters off Normandy. In the process, Washington overcame the enormous advantages that the Germans and Japanese held at the beginning of the war.

Big organizations can innovate. It's just hard.

Mason is director of the Bell-Mason Group. She can be reached at hbmason@aol.com. Rohner is a partner with Diamond Technology Partners. He can be reached at rohnert@diamtech.com.