Businessweek Archives

The Technology Paradox

Cover Story


When Computer Associates International Inc. introduced an accounting program, Simply Money, in 1993, it picked a daring price point for its first million copies: zero.

Toshiba Corp. happily admits that its forthcoming digital-movie player will never earn back the investment poured into it.

Order service from Teleport Communications Group Inc., and a Teleport crew will install a dozen optical fibers with a million times the capacity you wanted--at no extra charge.

Are these companies crazy? No, they're obeying a new set of rules--seeming paradoxes that only make sense in light of the ongoing revolution in the price and capability of digital technology. To wit: Computer Associates gave away Simply Money software on the theory that favorable word-of-mouth would outweigh the trivial expense of making the diskettes--and persuade customers to buy upgrades and related programs. Toshiba plans to recoup its development cost for digital videodisks with spin-offs in other products, from high-capacity audio players to storage devices for laptop computers. As for Teleport: Optical fibers are so cheap that it makes sense to install enough capacity to last a lifetime.

That's the technology paradox: Businesses can thrive at the very moment when their prices are falling the fastest. "The only thing that matters is if the exponential growth of your market is faster than the exponential decline of your prices," says George M.C. Fisher, chairman and CEO of Eastman Kodak Co. The challenge is enormous, he says. "Companies have to project out: `How will I be competitive in a world (in which) technology will be virtually free?"'

VANISHING POINT. High tech has had its own inverted economic logic ever since the invention of the transistor at AT&T Bell Laboratories in 1948: Power goes up and price comes down in lockstep. In the past few years, though, cheap technology has crossed an invisible threshold to assume a central role in economies around the world. In 1994, Americans spent more on home PCs than on TV sets. Cyberspace became a middle-class suburb. People magazine named Internet guru Vinton G. Cerf one of the 25 most fascinating people of the year. Five thousand Vietnamese owned digital cellular phones. Wounded by the microprocessor revolution, supercomputer giant Cray Research Inc. saw its stock fall to barely 10% of its 1987 value.

The business world, in short, is being stood on its head. This creative destruction has fractured rock-solid principles of commerce of decades past. Among the casualties: 10-year plans and deliberate pacing of product cycles. Clear distinctions between custom and mass markets. Giant soup-to-nuts systems from single companies--such as IBM's long-dominant mainframe computer networks.

The new rules require more than ingenuity, agility, and speed. They call for redefining value in an economy where the cost of raw technology is plummeting toward zero. Sooner or later, this plunge will obliterate the worth of almost any specific piece of hardware or software. Then, value will be in establishing a long-term relationship with a customer--even if it means giving the first generation of a product away. This won't happen with cars and other products built from bulk materials such as steel, but it's already happening with the electronics systems that are increasingly becoming an integral part of these products.

No single set of rules works for every player. A strategy of domination and control befits the likes of Intel Corp. and Microsoft Corp., which own vital standards and charge handsomely for them. There's also money to be made at the other extreme, in pure commodities, from basic disk drives to plain-vanilla consumer electronics. Dynamic random-access memory (DRAM) chips, for example, perhaps the purest example of a commodity, have once again become the biggest source of profit for Japan's hard-pressed electronics giants as huge demand has propped up their prices.

For most companies, however, the technology revolution dictates a middle course--latching on to a standard rather than creating one, and then putting in a little something extra to elevate one's product above commodityhood. Compaq Computer, Dell, and Packard Bell Electronics survive the PC wars by tweaking the Intel-Microsoft standard to boost the speed or enhance ease of use. Sharp Corp. made a best-seller out of an ordinary video camera by building a flat-panel TV screen into it.

To companies struggling to adapt, this brave new world isn't so wonderful. Their high-tech crown jewels are suddenly all but worthless. AT&T has taken billions of dollars in charges since 1984 to get rid of obsolete jobs and equipment. IBM, DEC, and Wang Laboratories are still trying to recover from the blow dealt by the microprocessor.

Japanese companies have been hit especially hard. The country's consumer-electronics industry has been mired in single-digit growth since the mid-1980s. Who needs a $500 camera when disposable models come with panoramic, telephoto, and underwater lenses? Dime-store digital watches keep time as well as costly chronometers. And the sound on cheap portable CD players is so good that savvy Tokyoites plug them into speakers in lieu of buying fancy audio systems. "How do you assign prices or value in a world where quality is perfect and nothing breaks?" asks Yotaro Suzuki, senior vice-president of the Japan Institute of Office Automation in Tokyo.

FROM NIBBLE TO CHOMP. The wave of downsizing that swept the computer industry is about to come crashing down on telecommunications gear (page 84). Companies such as Cisco Systems, Synoptics Communications, and Cabletron Systems stole a march on the likes of AT&T, Northern Telecom, and Siemens with equipment that links PCs in networks. Says Daniel Lynch, chairman of Interop Inc., which runs trade shows: "The new guys start out nibbling. The big guys never notice. And then, boom!"

Successful strategists soak out costs, cut prices, and then wait for business to roll in. Step 1 is a price decline--say, in DRAM chips. At first it causes chaos, as in the early 1980s, when American producers fled the DRAM market amid cries of Japanese dumping. In Step 2, the market finds a new use for the cheap resource. Case in point: Windows software, which is ubiquitous and gobbles megabytes of DRAMs. Unit prices of the chips may fall, but gross revenue soars--a point that is lost on those who once again predict doom for commodity chipmakers. The next wave of chip demand will come from the likes of computers that obey spoken commands and communicate in 3-D images. After that? Believable virtual reality and intelligent artificial intelligence.

The new rules of "techonomics" practically guarantee that high-tech product gluts will be temporary. That's because the economy and society will reshape themselves to take advantage of the cheap resource, whether it's computers that talk or satellites that track stolen cars. Demand for digital resources--unlike demand for, say, food and clothing--is almost infinitely elastic. Witness the hockey-stick shape of graphs plotting usage of online services. Marvels Steve Case, president and CEO of America Online Inc.: "Things take a while to coalesce, then they explode."

Even the ubiquitous microprocessor was once seen, by its own inventors, as a niche product. Intel co-founder Gordon E. Moore rejected a 1970s proposal for a home computer built around an early microprocessor. "I personally didn't see anything useful in it, so we never gave it another thought," he later recalled. In a list of possible uses for its 286 chip written before the success of the IBM PC, Intel omitted the personal computer, thinking instead of industrial automation, transaction processing, and telecommunications. But then, history is replete with overlooked opportunities, from microwave ovens to fiber optics (table, page 80).

The signs that the old way of doing business was failing first became apparent in the mid-1980s. Electronics seemed like a dead-end street. Intense global competition had ripped profits out of the consumer-electronics industry, and the chip and computer industries seemed destined to follow. Says Mark Rosenker, vice-president of the Electronic Industries Assn.: "We were walking around with our heads in our hands saying, `My God, what's happening to us?"' Many theorists argued that the only way to win was to emulate huge, vertically integrated Japanese conglomerates, which could afford to make all their own components, lose money launching new products, and weather debilitating price wars.

The pessimists were proven wrong. By the early 1990s, a new strategy took shape in the chip, PC, and communications wars. To survive, smart middle-tier companies, which were neither owners of a standard nor pure commodity players, stopped focusing on the product alone. They started concentrating on whole "architectures," or grand schemes for integrating many products into smoothly functioning systems. The approach required an intimate knowledge of customer needs, the ability to read technology road maps three to five years off, and a willingness to look beyond easy opportunities in the current mass market.

Small chipmakers such as LSI Logic, Cirrus Logic, and VLSI Technology couldn't sell a new PC microprocessor: Intel had that business sewn up. So they took aim at so-called chip sets that added sound and visual effects for PCs and on emerging-device areas, such as digital settop boxes that link up to the Information Superhighway.

FITTING IN. Europe and Japan offered fertile ground for these American hotshots. Open up a Canon single-lens reflex camera, a Ricoh fax machine, or a Philips Magnavox television, and you'll see specialized U.S. chips. Sony Corp.'s much admired game machine, the Play Station, runs on a chip designed by LSI Logic.

Companies don't have to create a new architecture--only fit into one. That's how C-Cube Microsystems gained leadership in the fast-growing field of digital signal compression. Managers didn't even think of protecting their 2 1/2 years of work on a powerful video-compression chip behind a wall of patents and copyrights. Says C-Cube founder Alexandre Balkanski: "The whole trick is to get in quickly and run as fast as you can. There are no birthrights."

Digital satellite TV provides a glimpse of the strategies that will work in the future. A band of companies led by General Motors Corp.'s Hughes Electronics began planning it four years ago after calculating that the necessary components--then far too costly--were about to become affordable. They were right: Partner RCA, a unit of France's Thomson, was able to offer a home dish and decoder starting at $699. Since last June, more than 400,000 American homes have signed up.

The confidence to take the gamble on satellite TV came from the designers' familiarity with the unique economics of silicon. Every 18 months or so, improvements in chipmaking technology make it possible to double the performance of silicon at no increase in price. More compact circuitry makes chips faster because electrons have less distance to travel. And as chips get smaller, more of them can be stamped out of the same slice of silicon. Consequently, single chips have taken over functions that used to be performed by refrigerators full of diodes, triodes, and capacitors. And chipmakers have assumed roles once performed by a wide range of companies, from makers of discrete electronic components to suppliers of software. "The food chain is collapsing," says Gilbert F. Amelio, CEO of National Semiconductor Corp.

The ascent of chips means that rules of play that originated in Silicon Valley are governing an ever growing segment of the economy. Around 2000, high-volume microprocessors will crack the "bips barrier"--execute more than a billion instructions per second. That will provide a playground for designers to come up with an almost limitless range of products, from holographic videoconferencing to Oracle Corp. CEO Lawrence J. Ellison's pet craving--a personal digital assistant that alerts your cardiologist if your company's stock falls. But manufacturers beware: These miraculous devices will be subject to the same punishing cost curves as PCs and cellular phones.

Advances in optic fiber will mirror the miracles in silicon. Engineers continually upgrade the capacity of hair-thin glass fibers by jacking up the pulse rate and splitting the light beams that carry information into multiple wavelengths. Already, the cost of carrying one more phone call is practically zero, making the call a tricky product to price. Some long-distance carriers have even adopted "postal pricing," in which the cost of a three-minute call is the same anywhere in the continental U.S.

The same is true of software. The incremental cost of manufacturing one more diskette or CD-ROM containing software is close to zero. Consumers know that, which is why many feel justified in making illegal copies. Some idealistic programmers, like those affiliated with the Free Software Foundation in Cambridge, Mass., even argue that copying should be free. To help counter that movement, software publishers know they must give customers something more than a platter of bits. Makers of mainframe software are stressing customization or 24-hour maintenance agreements. Entertainment-software companies are acting like magazine publishers, building brand names by churning out titles frequently at easy-to-swallow prices.

"SPLENDID" PIRATES. There's an old joke about a dim manager who brags to his boss that their company is losing money on every sale--but making it up on volume. In the era of "free" technology, that manager doesn't seem so stupid anymore. With the production cost of hardware and software so low in comparison to the development cost, it actually can make sense to give stuff away in order to establish a market toehold and start a profitable long-term relationship.

Michael Goldhaber, president of the San Francisco-based Center for Technology & Democracy, sees these grabs for mind share as the early manifestations of a coming high-tech "attention economy" in which attention from others is the scarcest and hence the most valuable commodity. Makes sense: Bits and bytes are virtually unlimited, but there are still only 24 hours in a day.

Media baron Rupert Murdoch understands that. In India, pirates steal the signal from News Corp.'s Star TV satellite and profit by reselling the programs to people over cable. "Some cynics have said this will be fatal for Star. We disagree," Murdoch said in a speech last year in Melbourne. His logic? The pirates--or "splendid entrepreneurs," in his words--are simply broadening the potential market for Star TV and allowing Murdoch to raise advertising rates.

If attention is the most precious resource in a free-tech economy, then it makes sense to throw battalions of cheap bits into capturing a share of it by making products exciting, easy to use, or preferably both. Work on the interface between human and machine already consumes three-quarters of the development work on electronic products, says Gary A. Curtis, a Boston Consulting Group Inc. vice-president and leader of its worldwide information-technology practice. Nonetheless, technology keeps getting more costly in terms of the time required to master it. Consultant Gartner Group Inc. calculates that the lifetime cost of owning an ordinary corporate PC--troubleshooting, training, and all--comes to more than $42,000.

In Goldhaber's attention economy, attention is reciprocal: Producers get it from consumers by showering it on them. Success will hinge on "letting customers define what they want to buy," then building it to order, says Roger N. Nagel, deputy director of Lehigh University's Iacocca Institute. The cost of the parts will be almost incidental to the price of the final package, which includes service and follow-up modifications. Those who offer attractive packages can virtually give away their hardware, says Nagel: "Tomorrow's factories will sell customer gratification, not things."

Ultimately, companies will "mass-customize" their products, which means serving the mass market with products that are nonetheless tailored to individuals. Dell Computer Corp. already bundles each PC with software and peripherals that the customer demands--a level of customization previously reserved for cars and houses. The goal is to live in a commodity marketplace without serving up a commodity. Says Kodak's Fisher: "Every unit coming down the line should be capable of being different from the one that preceded it."

This, after all, is where the Japanese went astray. Much of their electronics universe is analog and electromechanical, with high-priced paraphernalia of spinning platters, motors, shutters, and lenses that take teams of experts to modify or enhance. So while American companies rushed forward with customized, multifunction chips that formed the heart of high-value, new product categories, Matsushita, Sony, and Sanyo--the original commoditizers--were endlessly mired in low-margin "box businesses." Their VCRs and microwave ovens could not exploit the economies of the silicon curve.

Nintendo Co. was one of the first to find a path out. "There's no way to charge a premium on hardware," says Nintendo President Hiroshi Yamauchi. So the world's leading video game company charted a business model in which the game consoles would be "given away" to consumers at cost, or even below, to boost sales of software. Virtually all of Nintendo's considerable profits flow from sales, license fees, and manufacturing charges on the game software.

Free technology demands that engineers learn a whole new discipline: wastefulness. Makers of custom logic chips regularly stitch together quick and loose designs on silicon, squandering what used to be precious "real estate" on the chip, to get products to market faster. It's hard on designers who take pride in writing tight code and building efficient systems. But why waste time and effort making efficient use of something that isn't scarce? The logic also applies to software. Microsoft's Windows 95 will be huge and slow on today's PCs. But Microsoft knows that customers will simply buy new computers or add memory to their old ones.

To be sure, there will always be some sectors, such as video transmission today, where shortages of capacity make efficiency a winning strategy. Efficient software for data and video compression, for instance, is a boom industry. But compression only creates more room--on a disk or over a phone line--for some other programmer to be creatively wasteful.

While free technology makes room for creativity, it also intensifies global competition. Cheap communications mean chipmakers in Silicon Valley can farm out work to Taiwan or Tel Aviv. "We have no monopoly in this country in 1994 on technology smarts and capabilities," says Warren McFarlan, senior associate dean of Harvard B-school. A visit last summer to four software companies in New Delhi opened his eyes: "They are dominated by Harvard and MIT PhDs and are every bit as good as any firm I've looked at in the U.S."

Yet it's misguided to worry that in a free-tech economy, all manufacturing, service, and even engineering jobs might migrate to low-wage nations. Where product cycles are shortest, such as in PCs, factories have remained inside the biggest markets--such as the U.S. and Japan--to save shipping time. Engineers also prefer to be close to the largest, most sophisticated markets. Taiwan's Acer Inc., one of the most aggressive PC price-cutters, is moving most of its assembly for the U.S. market to San Jose.

Information technologies will never be literally free--just when they start to seem so, someone finds a way to consume so much that they start to seem scarce all over again. Then the cry goes out for still more capacity and speed. That virtuous cycle promises to keep rolling as long as there are people with the imagination to envision something better.

The New Rules Of The Game

The advent of "free" technology is forcing companies to master some slippery concepts. Warning: Not every rule is appropriate for every business.

Products are most valuable when they're cheapest. The niche for high-priced products gets ever smaller. Low prices and high volumes are the way to go. Just ask Compaq Computer.

Make money by giving things away. High tech loves shaving economics--giving away the razor to sell more blades. Mosaic software for the Internet started out free. Now, its creators sell it as Netscape.

Teamwork conquers all. The complexity of the latest electronics gadgets--such as digital satellite TV--requires the kind of collaborative systems design that used to be the province gf general contractors on aircraft, ships, and moon shots.

Mass-customize. To avoid the me-too commodity trap, use agile manufacturing techniques to make each product off the line unique. That's how Dell sells PCs and Matsushita sells mountain bikes.

Hurry up and waste. Engineering efficiency in product design may be nice, but with computing and communication resources so cheap and speed to market so essential, quick-and-dirty is often the best route.

Don't fear gluts. Demand for individual products rises and falls, but hunger for such things as computing power is, in the long term, insatiable. Memory chips are fantastically profitable despite being lowly "commodities."By Neil Gross and Peter Coy with Otis Port in New York, with bureau reports

American Apparel's Future

(enter your email)
(enter up to 5 email addresses, separated by commas)

Max 250 characters

blog comments powered by Disqus