(FORTUNE Magazine) – Quick: When was the fax machine invented? 1955? 1968? Give up? Alexander Bain, a Scottish clockmaker, patented the first fax in 1843. Two decades later an Italian priest named Giovanni Caselli created the first commercial system. Napoleon III, Emperor of the French and an early adopter long before the word was coined, became a big customer, setting up a dedicated fax link between Paris and Lyons in the 1860s. This system used synchronized pendulums and a telegraph wire to transmit pictures and stock market results, according to Ira Flatow's history, They All Laughed...Nearly a century and a half later, fax machines became high-volume consumer items, with 1994 sales of $2 billion, while generating substantial revenues from paper, toner, and telephone usage time.

During the intervening years, the elements needed to transform this brilliant technological idea into a marketplace winner gradually fell into place: Telephone companies built a global infrastructure to carry the signals; Federal Express and its competitors taught people to expect fast delivery of documents; and the microprocessor made possible robustly powerful machines priced for the mass market.

To hear the experts tell it, laboratories all over the world are glutted with good ideas. The hard part is figuring out what to do with them. "It's easy to get the technology to work," says L. John Doerr, a leading Silicon Valley venture capitalist who is a partner in Kleiner Perkins Caufield & Byers. "What I see over and over is that the technology works but misses the market." In applied technology, there is often more mystery in selling than in inventing.

To understand what transforms technology into dollars, compare some of the biggest hits of the past 40 years, shown on the time line below, with underachievers in the same period. Remember the hype about gallium arsenide semiconductors and high-temperature superconductivity? Driven an electric car lately? Dusted off your Sony Betamax or MCA laser videodisk? Know anyone who actually uses a pen computer or a personal digital assistant? When will we be served by ISDN, the fabled all-digital telephone network?

IBM deserves special mention for its extraordinary record of failure to exploit superb technology during the years before CEO Lou Gerstner arrived. Besides being involved in the laser disk project, IBM also invented the fast RISC microprocessors now used in computer workstations and PowerPCs, but ceded the early profits from them to Sun Microsystems and others. Although IBM won wide acceptance for its personal-computer design in the 1980s, the company squandered its early lead and more recently has lost money in PCs. So what makes for a winning idea? Rule No. 1, says Joseph Morone, dean of the management school at Rensselaer Polytechnic Institute: The technology must be essential to the strategy of the corporation developing it. In the course of bringing something radically new to market, he explains, corporations almost inevitably run into serious trouble. "The companies that persist through all the uncertainty are those that have no choice but to see it through," he says. "If that sense of strategic imperative isn't there, the likelihood of proceeding through years of pain and investment is exceedingly remote."

History--and not just at IBM--suggests he's right. In the 1920s, 50 years after Napoleon III was deposed, the New York Times Co. began dabbling with fax technology. Focused on conventional publishing, the Times didn't get around to starting a faxed delivery service until 1990, by which time digital transmission was beginning to make more sense; the fax business remains tiny.

Commitment pays off. Cushioned by healthy revenues from drugs, Pfizer dropped out of the race to develop computerized tomography, or CT, scanners. General Electric's medical systems unit, which made X-ray machines, had to persist to protect its core diagnostic-imaging business. Thanks in part to hefty CT sales, the GE unit's 1994 revenues reached an estimated $3.5 billion, big enough for inclusion in the Fortune 500 if the unit were independent. Similarly, AT&T, with enormous cash flow from long-distance service, didn't pursue fiber optics during the 1970s. With much more to lose, Corning lavished 15 years and $100 million to develop the technology, now a major source of the company's profits.

Rule No. 2: Technologies need champions. Says Arno Penzias, a Nobel laureate in physics who is vice president of research at Bell Labs: "The definition of a lousy product is one that has no enemies inside the company." Sales of almost any new product will eat into someone's existing business-and the more strategically important the new product is to a company, the more likely the products cannibalized will be its own. If not, argues Penzias, "you know none of your customers will be interested."

One way to address that problem is to double your risk. In 1984, Hewlett-Packard introduced two competing lines of print- ers for PCs: laser printers, in which H-P established a strong market position, and ink-jet printers. Lured in part by the prospect of selling high-profit-margin disposables, such as ink cartridges, the company resisted the temptation to narrow its focus to laser machines. Now H-P is the world leader in both laser and ink-jet printers. No matter what they say, few companies are as willing as H-P to cannibalize existing businesses on the way to competitive advantage.

"People who bring products to market have scars all over them," says Larry Moore, the Lotus vice president who helped shepherd Notes groupware to market. "You face skepticism in the company and in the market. You're always battling for resources and mind share. You've got to have the courage of your conviction--people respect utter, unshakable commitment. But at the end of the day, you're getting people to sign on to a hunch."

That's where marketing comes in. Rule No. 3: The technology must have high and clearly defined value to customers. Says Tony Engberg, who manages five research and development labs at H-P: "You've got to be able to describe it so people can intuitively grasp its importance to them and their business. You want an 'Oh, yeah-wow!' reaction." When Microsoft finally introduced a well-designed version of Windows software in 1990, millions of PC users long bewildered by the enigmatic c> on their screens cried, "Oh, yeah--wow!" almost in unison. To date, Microsoft has sold more than 80 million copies of Windows.

The question of timing is delicate. Any number of great products that later caught on suffered from a premature introduction: among them, the typewriter (1874), the Xerox-type dry copier (1938), the microwave oven (1953), and the computer mouse (1964). Maybe electric cars belong on that list-time will tell. Lotus's Moore says the need a new product meets must be strong but not yet articulated; if an idea comes up in a focus group, forget it, he says. Its time has already passed.

In a new book called Harmony: Business, Technology, and Life After Paperwork, Penzias defines three great needs that demand satisfaction from products of almost any category. Tomorrow's products, he writes, must be in harmony with users, other products, and the environment. The very opposite, in other words, of many successful products today, such as PCs, video entertainment systems, and cellular phones--all baffling to users, unlikely to work with each other, and fated to quickly become unrecyclable junk.

Rule No. 4: Make it easy. Ease of use is important, as Penzias notes, but not sufficient. A new product is far likelier to succeed if designed to permit an easy transition from whatever it replaces. Usually the pickings are best when a new product can displace an old one in a market that already exists. "But when people want a cart with ten more horses," warns Engberg of H-P, "getting them to use a car is difficult."

A third kind of ease is equally important: ease of deployment. To make his light bulbs useful to customers, Thomas Edison had to create an electrical infrastructure from scratch; he ended up owning power plants and wire factories. Far easier to piggyback on an existing business framework, as when aircraft-engine makers talked airlines into switching from propellers to jets.

Oh--one more thing. George Heilmeier is a respected scientist who made important contributions to liquid-crystal technology and now runs Bellcore, the software research and consulting company jointly owned by the seven regional Bell telephone companies. Says he: "We can sound very logical about what it takes to bring a technology to success in the marketplace. But the bottom line is that there's always luck involved."

Microsoft was a humble vendor of computer-language software in 1980 when Lady Luck came calling. IBM, needing an operating system for its forthcoming personal computer, first approached Digital Research, a company that was already selling a similar program. According to industry legend, the head of Digital Research failed to respond to Big Blue's advances, giving Microsoft the opportunity to provide the basic software, known as DOS, that PCs need to run. The success of IBM's PC made DOS a best-seller and gave Microsoft its opportunity to change the world. So that's Rule No. 5: Get lucky.