Launching an e-commerce initiative has been likened to jumping out of an airplane without a parachute, in the hope that you can sew one together before you hit the ground. But building a “dot-com” business is actually more like flying a plane. In flight, speed transforms thin air into lift for an airplane’s wings. In Internet-based commerce, companies that implement their ideas rapidly enough can accomplish the cyber equivalent of Bernoulli’s Principle and soar off into the heavens.

Of course, those that don’t move fast enough will stall and, in Air Force lingo, “burn a billion-dollar hole in the ground.” A recent cover story in Fortune magazine listed more than a dozen high-profile chief executives who it said were fired even though they were smart, successful, and had solid strategies. The problem: They couldn’t get their organizations to implement those strategies fast enough.

In today’s chaotic environment, your ability to implement quickly is your strategy.

This need for speed changes everything about how you manage your business. You have to push ideas into the marketplace far earlier, when they’re really still in the prototype stage, or risk having a competitor steal your thunder. You have to learn—on the fly, if you will—what works and what doesn’t. You have to kill the losers quickly and turn the winners into big businesses. Then, you have to do it again, and again. The new mantra is: Think big. Start small. Test quickly. Scale rapidly.

It’s easy to avoid obstacles when you’re moving slowly, but frightfully hard when you go supersonic. So, lest you start crashing into mountains, you need to impose a new set of disciplines on how you manage programs. To help, I’ve put together a three-point checklist based on the hard lessons that my partners and I have seen companies learn over the years. Following these suggestions will let you avoid the major, common pitfalls and help you implement ideas better and faster.

ORGANIZATION: Just as the first U.S. rocket plane, the X-15, was dropped from the belly of a conventional jet, a digital venture has to be autonomous to fly high and fast. Otherwise, because it represents such a threat to the existing order, it will be sabotaged by groups in your business. A dot-com can’t be just a project. It has to be at least a program and, more likely, a separate business unit reporting directly to the chief executive officer. Toys “R” Us, for one, recently took a shortcut into the on-line world by setting up a highly autonomous venture, in a partnership with venture-capital firm KPL, to fight back against upstart eToys.

The venture’s independence will help you give it the necessary flexibility, creativity, and gung-ho spirit. But you’ll still need partners to give you some “Web DNA” and to help you get to market quickly. These partners include contractors, developers, and consultants, who can provide specific expertise. Partners should also include businesses that augment yours—for instance, if you sell cars, you need to be able to offer car insurance and financing.

PEOPLE: An A team will win even with a B business plan, because it will adapt to market conditions and competitors. The big question is how to get eager young e-jockeys to work for your boring old company when they can go to a hip start-up and, they assume, make millions.

The first answer is stock options (known as lottery tickets in Silicon Valley). Equity—and dreams of riches—is what brings in the talent and then creates the siege mentality that gets that last, needed 10% out of the team. The options should be in the dot-com venture, not the parent, which may mean creating a “tracking stock” that reflects the results of the digital enterprise. Hanover Direct, a catalog retailer, says it has had great results from using lots of equity as incentives in its digital ventures—but Rakesh Kaul, Hanover Direct’s chief executive, says he may, in fact, not have gone far enough yet.

Even if you don’t create options, it’s best to pay your staff in a way that’s perceived as equitable with what any contractors and consultants get. That may mean cash bonuses as your venture reaches milestones, but the expense will be worth it because you’ll have everyone working toward the same objectives.

Fortunately, money isn’t everything. W.W. Grainger, a distributor of industrial parts—in other words, the least sexy business ever created—found it could attract good people by convincing them they’d learn a lot and be involved in something cool. This opportunity to be a creator is very compelling to technologists. Members of the development team for Apple Computer’s first Macintosh had their signatures imprinted inside its case. They later said this was more motivational than stock options.

GETTING THINGS DONE: Once many project managers boot up Microsoft Project, they can’t help themselves. They start planning, instead of prototyping, and are slow to get to market. The traditional “waterfall,” or sequential work plan, simply doesn’t work anymore. Cargill, for one, knows better. When the agricultural company needed to stitch together 25 different software tools from almost as many vendors, conventional wisdom called for working on the individual tools and then integrating them at the end of the process. But Cargill did an early test—a sort of prototype—and learned crucial lessons, gaining valuable time. Cargill had focused on what really matters: getting something out there, then learning and adapting through constant iteration.

Traditional planning metrics don’t work, either. For instance, percentages of completion. Software developers say that the first 90% of the code takes 80% of the time—and the final 10% takes 80% of the time. That’s pretty much how things work on e-commerce projects, too. The only way to deal with estimating time is to be binary. Either something is finished or it isn’t.

Traditional planning cycles have to go, as well. Big companies generally operate with annual budgets, but today’s market can’t be divided into neat, yearlong chunks. An e-commerce venture that looks promising needs to be financed immediately. A loser has to be killed off just as fast. In effect, you have to create the instant feedback loop of an entrepreneur, rather than the central-planning techniques of corporate bureaucracy.

Remember: Your aim is, in part, to be the first mover. After all, does anyone recall the names Kohl, von Hunefeld, and Fitzmaurice? They were the first to fly from Europe to America, only a few months after Charles Lindbergh’s epic 1927 flight the opposite way. But who cares?

Mainly, though, your aim is to be the first finisher. Think of Nungesser and Coli, famed French aviators in the 1920s but now long-forgotten. They took off 12 days before Lindbergh but simply vanished into the Atlantic.

Gutstein is chief operating officer of Diamond Technology Partners. He can be reached at gutstein@diamtech.com.