HP, Aiming to Regain Innovative Edge, Goes Back to Drawing Board

The decision by Carly Fiorina, chief executive of Hewlett-Packard Co., to place the word “invent” within her company’s logo last year did more than prime the marketing pump. The slogan frames the central challenge for a company whose most influential achievements, such as calculators and ink-jet computer printers, seem to be fading rapidly into the past.

Fiorina realized that HP faces a crucial juncture: A new cadre of managers must adopt a revitalized approach to research and development to drive the spirit of innovation that built one of technology’s great companies.

“They need to be relevant again. [HP] fell so far behind in terms of making a difference in technology that they became a commodity provider” of low-cost consumer products such as printers and PCs, said Rob Enderle, an analyst with Giga Information Group.

Fiorina’s credibility took a beating this month when an unforeseen shortfall in quarterly earnings sent HP’s stock down sharply to a 52-week low.

But her plan to reinvigorate Hewlett-Packard’s R&D; efforts continues apace. Last month Fiorina brought in Richard DeMillo as HP’s chief technology officer and Stephen Squires as chief science officer. Squires had been special assistant for information technology for the director of the U.S. Defense Advanced Research Projects Agency. DeMillo was a research executive at Telcordia Technologies (the phone industry think tank formed after the break-up of AT&T;). They join Richard Lampman, director of HP Labs, as leaders of the company’s R&D; efforts.

DeMillo said HP aspires to reach the same pinnacle of strategic importance in the future of technology as Apple Computer and the Xerox Palo Alto Research Center achieved when they developed the graphical user interface, which opened the world of computing to the masses.

To achieve such lofty goals, HP must embody the essence of free inquiry and maintain a long view, Squires said. “People have to feel empowered to try something without permission.” The entire organization should “think about the future in a relentless way.”

HP already spends $2.6 billion, or 5.4% of sales, a year on R&D.; HP Labs conducts research in e-commerce, storage systems, encryption, biometrics, Internet appliances and quantum mechanics, among other areas.

Although HP’s overall R&D; spending is huge, IBM invests at close to three times HP’s rate. The payoff for Big Blue in recent years has included fundamental breakthroughs in microprocessor and storage technologies leading to tens of billions of dollars in component sales and licensing royalties.

But a fixation on budgets ignores how innovation works in the real world, Squires said. “It’s not about spending money,” he said. “It’s about ideas. . . . If we had five or 10 times the budget, it wouldn’t help much if we didn’t have the corresponding ideas.”

And industry partnerships can be as critical as new ideas. HP has always been mindful of such relationships, such as a key alliance with chip titan Intel to design a new family of processors that will power the next generation of server computers that manage networks, as well as sophisticated engineering workstations.

Those processors--dubbed “Itanium” by Intel and due out next year--should quickly become the top offering in a highly lucrative niche, analysts say. As co-designer, HP is positioned to offer customers faster, more reliable systems with tighter integration between hardware and software.

The relationship with Intel could pay other dividends down the road.

“We know finally that Moore’s Law [the industry maxim that microprocessor computing power doubles every 18 to 24 months] is at its limit,” Squires said. “As soon as you know you are at the end of one curve, you have to be at the beginning of a new one.”

The new curve, he hopes, will be based on molecular-scale processors thousands of times smaller--and millions of times more powerful--than today’s silicon chips.

If such processors can be perfected, they could eventually force the industry and society as a whole to rethink their assumptions about the role of computing in everyday life. Molecular-scale processors could lead to supercomputers small enough to be woven into fabrics, using body heat or ambient light as a power supply.

HP scientists are at the forefront of such research, and Squires predicts the company will show working prototypes of devices operating on molecular-scale processors within four years.

Because of HP’s relationship with Intel, “they could dovetail [molecular-scale processors] into something that’s mainstream much sooner,” said Giga’s Enderle.

That rapid progression from the lab to the market has been an HP trademark, leading to a number of watershed products:

* In 1968, the first programmable desktop calculator.

* In 1972, the first pocket scientific calculator, which made the slide rule obsolete.

* In 1984, the first thermal ink-jet printer for PCs.

To its credit, “HP has always done a great job of moving its research out of the lab into business,” said Paul Saffo, a director of the Institute for the Future in Menlo Park, Calif.

In contrast, Xerox virtually invented modern computing with its development of networking, the graphical interface display on a PC screen and a stream of other innovations at PARC--yet lost out to competitors that exploited those technologies in the marketplace.

Yet in recent years, HP’s research work has been incremental, experts say--one reason why the company hungers for a giant leap forward, as molecular-scale computing could be.

HP’s emphasis on ideas and apparent commitment to intellectual freedom aside, a willingness to spend freely has usually been a prerequisite for big accomplishments.

Financial strains that have hammered HP’s stock recently can have a similarly chilling effect on costly research projects that may take years to complete.

And last year when HP spun off Agilent--its test and measurement business--it lost the division’s billion-dollar-plus R&D; budget.

HP also faces significant cultural challenges, said Donald Norman, president of UNext Learning Systems and author of “The Invisible Computer,” a book about information appliances. Norman was formerly research chief for Apple and worked briefly at HP Labs, leaving in frustration in 1998.

“The problem with HP has long been that it is a set of exclusive fiefdoms,” Norman said. The company’s organization into numerous geographically dispersed business units discourages rolling the dice on big ideas, he said. For one of the nation’s largest technology firms, an occasional $100-million gamble makes sense, but no small division can afford it.

“The result is that they ended up being a very conservative company--basically afraid to take risks, afraid to move in new directions,” Norman said. Fiorina’s efforts may change that, but not quickly, given ingrained patterns, he said.

Joel Birnbaum, former director of HP Labs and now a consultant to Fiorina, said that HP should consider new types of entrepreneurial partners to keep up with the pace of change.

“Sometimes . . . it’s better to work with venture capital and small start-ups than trying to build everything inside,” Birnbaum said. “If it’s not faster-moving, it’s not going to survive.”

In the long run, DeMillo said, HP will place many research bets and some will never pan out. But it will keep returning to key areas, including e-commerce, printing, imaging and, in particular, services.

“Information technology, wherever it’s deployed, has to be judged on how it delivers value to end users,” DeMillo said. But how can HP provide the same service quality to all its customers, from the Fortune 500 to “a rural schoolhouse in Cambodia? You can’t.”

So DeMillo proposed a goal: Invent a layer of technology that effectively substitutes for that service for anyone who touches HP equipment anywhere.

He added: “I think it’s a huge challenge.”

(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)

The R&D; Race

Technology companies that consistently push the industry in new directions normally have large research and development budgets. IBM, Microsoft, Xerox and Hewlett-Packard historically have spent a higher percentage of their revenue on R&D; than have other tech companies. Dollar amounts in billions:

*--*

R&D; As % of Company spending* revenue IBM $12.7 14.6% Microsoft 3.9 16.8 Hewlett-Packard 2.6 5.4 Agilent** 1.2 12.3 Xerox 1.0 5.4 Dell 0.5 1.6 Apple 0.4 4.8

*--*

* Data correspond to last four fiscal quarters.

** Most recent quarter is estimated.

Source: Companies