Who'd have guessed that the invention of the stirrup would lead to hundreds of years of feudalism? That Gerald Ford's Whip Inflation Now program would bring us home shopping? That the invention of the spreadsheet could serve as a catalyst for the mergers and acquisitions frenzy of the 1980s?

While all three examples seem to be random, unpredictable events, they're actually not so surprising. The history of technology is full of such dramatic and unexpected consequences.

These days, technology is playing an increasingly important role in business—and is changing faster than most businesses can absorb. The result will be so many more violent surprises that it seems to be time to give them a name and declare them a new class of business problem. Let's call them "killer apps," a term borrowed from the venture capital world, where killer applications, or apps, have long transformed industries, unseated market leaders, and provided thousandfold returns on initial investments.

If businesses can focus on killer apps, they can finally understand how to expect the unexpected.

In the case of the stirrup, the unexpected consequences took hold slowly. The stirrup made lances so effective on horseback that Charles Martel's Franks defeated the previously unstoppable Saracens in the eighth century. Martel's son Charlemagne was no dummy, so he made knights a permanent fixture of his army. To support the knights financially, he gave them land and various rights that developed into feudalism. For the WIN program, an 800-number was created to let citizens suggest inflation-fighting programs. The toll-free idea caught on so rapidly that it redefined how people shop and receive customer service. The spreadsheet not only quickly transformed the computer industry by making the personal computer popular but, many say, redefined business finance.

These days, technology is changing commerce at least as fast as 800-numbers and the spreadsheet did. In fact, the consequences could be far more fundamental because of the Netscape browser, which turned a sleepy academic computer network called the Internet into a platform that may rapidly change the way most commerce is done. John Perry Barlow, who writes about cyberspace, calls the Internet "the single most important invention since fire."

How can you deal with something so elemental and so unpredictable?

Larry Downes and I have looked into dozens of killer apps, where technology has changed the dynamics of an industry. Based on those examples and on work with numerous clients, we've developed a theory for how to have your company do the killing—rather than wind up as roadkill. We came up with a dozen design principles that can help companies put important applications in place now while exploring where the true killer apps lie.

The principle that can be applied most immediately is:

OUTSOURCE TO THE CUSTOMER. Most organizations are already familiar with outsourcing, but this approach goes further. The Internet makes it possible to outsource data collection and management directly to your customers. So you simply build an interface to your information sources and give customers tools to navigate and customize them. You save time and money on data input. You reduce errors because the data are handled only once.

You also wind up with a happier customer, even though the customer is doing your work for you. That's because the customer typically feels that he or she is getting more timely and accurate service. Federal Express, for instance, has scored heavily with an Internet-based system that lets customers track a package through all the intermediate steps as it goes from sender to receiver. Holiday Inn and Hilton have won over customers with a system that helps users locate a hotel, check availability, and make a reservation.

Two other principles address businesses' biggest blind spots:

CANNIBALIZE YOUR MARKETS. Managers are justifiably frightened of launching a new good or service that gains market share at the expense of an existing company offering. However, managers must aggressively think about how competitors could use the Internet or some other technological advance to steal a big chunk of their sales, then find a way to cannibalize those sales before the competitors can gobble them up.

This isn't quite as scary as it sounds, because what appears to be cannibalism is often no such thing. If a regional newspaper like the San Jose Mercury News offers its news for a greatly reduced price on the Web, it is largely selling its highly regarded technology coverage to a new, national customer base.

But sometimes true cannibalism is necessary. For instance, securities broker Charles Schwab has begun offering a 20% lower commission to Internet customers. Schwab will have customers move to the Internet and may lose revenue. But it had no choice. Internet-based competitors like E*Trade were already offering much lower rates.

Similarly, Kodak and Fuji will have to find ways to cannibalize their highly profitable film markets. Otherwise, new competitors will steal lots of business. Digital cameras can now capture images, which can be printed on color printers. Because the cameras and printers are digital, they benefit from Moore's Law—that the cost of computing power falls 50% every 18 months. So, costs for a new competitor such as Hewlett-Packard will decline much faster than Kodak's and Fuji's.

ENSURE CONTINUITY FOR THE CUSTOMER, NOT YOURSELF. Many companies hide behind their customers when thinking of using information technology to touch them. People fear computers, companies say. The Internet is scary.

It is scary, but for whom?

Customers deal with electronic interfaces all day: the telephone, the television, ATMs, grocery scanners, car dashboards, even automated bus transfers. Customers could certainly be convinced that electronic commerce is just a more convenient, interactive version of things they already do. After all, tens of millions of people already shop over the telephone.

The disruption is much more on the merchants' side. Auto makers, for instance, have invested heavily in their dealer networks and don't want to change them too rapidly even though information technology makes them far less important.

But customers couldn't care less what auto makers or their dealers want. In fact, many would rather order from the manufacturer, especially if they could save money and get exactly the car they wanted. Auto makers need to find a way to use technology to satisfy their customers. Or someone else will.

Our dozen design principles will help you identify killer apps, so you can see them coming in time to start to react to them. But getting to the next level-where you unleash killer apps on your competitors-requires enormous corporate will.

It's hard to come up with great ideas. It's harder to believe in them. And it's hardest of all to do something about them.

Mr. Mui is co-author, with Larry Downes, of "Unleashing the Killer App: Digital Strategies for Market Dominance." He is also executive editor of Context. He can be reached at chunka@diamtech.com.