Hewlett Packard Rethinks ItselfBarbara Buell and Robert D. Hof
It's amazing that Bob Frankenberg ever got anything done at all. Until last year, the Hewlett-Packard Co. general manager dealt with no fewer than 38 in-house committees. They decided everything from what features to include in a new software program to what city would be best for staging a product launch. Just coming up with a name for the company's NewWave Computing software took nearly 100 people on nine committees seven months. "There was a lot of decision overhead," says John A. Young, HP's chief executive officer.This is Hewlett-Packard? The company whose blend of advanced technology and enlightened management defined Silicon Valley? The company that year after year topped everybody's list of America's best-managed companies? Could this be the same company whose "HP Way" encouraged innovation by abolishing rigid chains of command, eschewing fancy executive offices, and putting managers, executives, and employees on a first-name basis?
Yes, it's the same old Hewlett-Packard. But by the late 1980s, an unwieldy bureaucracy had bogged down the HP Way. A web of committees, originally designed to foster communication between HP's disparate operating groups, had pushed up costs and slowed down development. The 52-year-old company's entrenched culture, which was built around the HP Way's philosophy of egalitarianism and mutual respect, promoted consensus. But, critics say, the culture placed too much emphasis on rapport among executives and failed to penalize those who missed opportunities. "There aren't enough killers in HP's management," says analyst George F. Colony, president of Cambridge (Mass.)-based market research company Forrester Research Inc.
In the markets where HP is competing now, a few more killers might help. Today, computers and related products and services represent about 70% of HP's $13.2 billion in annual revenue, with scientific instruments, medical equipment, and electronic components making up the rest. Increasingly, HP is focusing on the fast-paced workstation and personal computer market segments, where its sluggish decision-making puts it at a disadvantage against nimble rivals such as Compaq Computer Corp. and Sun Microsystems Inc. Even giant IBM has proven more agile than HP lately.
'UNGLUED.' By last spring, Young realized things had gotten out of hand. The message came through loud and clear when a manager warned him that HP's most important project, a series of high-speed workstations being developed under the code name Snakes, was slipping a year behind schedule because of HP's burdensome bureaucracy. Workstations, which make up one of the computer industry's highest-growth areas, are so important to the company's computer strategy that in 1989, in an uncharacteristically bold move, HP paid nearly $500 million for ailing market pioneer Apollo Computer Inc. HP wanted Apollo's customers and its engineers who had expertise in the new RISC technology. But over the course of a bumpy merger, development schedules slipped, and some Apollo customers began jumping ship.
No wonder the normally even-tempered Young grew furious last spring when he learned about the Snakes delay, which was caused by endless meetings about technical decisions. "I think I may have come unglued that day," says Young. "It was a pretty clear signal to me." Young immediately attacked the problem. He removed the project's 200 engineers from the HP management structure so they could work on the Snakes workstation line free of the routine committee oversight.
That crisis convinced Young that big changes needed to be made throughout the company. Last summer, he called in Founder and Chairman David Packard, 78, to discuss a major revamping of HP's corporate structure. Packard, who remains HP's single largest shareholder, began meeting frequently with Young to discuss changes.
Young and Packard didn't have to look far for a successful computer organization. The company's own laser-printer operation was an example of how a business within a huge corporation could remain nimble and competitive. After seven years in laser printers, HP has held the lead with a 60% market share, despite competition from Apple, IBM, and a phalanx of Japanese rivals. The secret? Executive Vice-President Richard A. Hackborn's Boise (Idaho) division didn't have to deal with the usual swarm of committees.
The workstation crisis showed Young that the management structure HP had adopted for its computer businesses in the mid-1980s was failing. That unwieldy system was developed when HP was trying to move into "open systems"--computers that use nonproprietary software such as American Telephone & Telegraph Co.'s Unix. As it moved to open systems, HP needed to smooth over the differences and update its various computer lines. So it set up a series of committees to figure out what new open-systems technology to pursue, which to ignore, which of HP's products should be saved, and which should be shelved. But the committees kept multiplying, like a virus. "Everything was by committee," says Bear, Stearns & Co. analyst Clifford Friedman. "No one could make a decision."
The cure was the company's most drastic reorganization in 10 years. Young announced it in October, just as the company was finishing a surprisingly poor fiscal fourth quarter. Earnings plunged 18%, dragging down the full-year net by 11%, to $739 million on sales of $13.2 billion. Wall Street responded with a huge sell-off of HP shares, which sank to 26, about half the price they had been trading at a year earlier (chart).
So Young wiped out HP's committee structure and flattened the organization. To cut costs, HP had already launched an early retirement program. Young's October revamp divided the computer business into two main groups: One handles personal computers, printers, and other products sold through dealers, and the second oversees sales of workstations and minicomputers to big customers.
Another change: In place of the single corporate sales force established in the mid-1980s, each of the two computer units now has its own sales and marketing team. "The results are incredible," says Frankenberg, who now deals with three committees instead of 38. "We are doing more business and getting product out quicker with fewer people."
The overhaul has another benefit: It should pave the way for the ascent of a new generation of managers. Young, 58, and five of the other top 11 executives are expected to retire within the next four years. The new structure gives Young, who took HP's reins in 1978 when founders Bill Hewlett and Packard retired from day-to-day operations, a chance to evaluate the company's younger lieutenants. They now report directly to him and Chief Operating Officer Dean O. Morton. But the COO's main job now is slashing HP's overhead and overseeing its extensive quality-control program.
So far, Morton has been doing pretty well. Because HP is selling more low-margin products such as PCs and printers through dealers, gross margins have slipped from 52% in 1987 to 47%. Still, in the quarter ended on Jan. 31, earnings jumped 18%, to $205 million, thanks in part to the cost-cutting (chart). All told, the work force has been cut by 4,000, to 91,000. That has helped HP's stock bounce back to around 46.
Morton, 59, is on the retirement track too. That leaves two leading contenders for the top job: Computer Systems chief Lewis E. Platt, 49, and Computer Products head Hackborn, 53. Together, the two men control nearly three-fourths of HP's revenues. Platt is in charge of minicomputers, workstations, and the sales force that sells them directly to big businesses. Hackborn handles personal computers, printers, and other products sold through dealers. Both are engineers by training and steeped in HP culture. Cupertino (Calif.)-based Platt is an energetic manager, who has run everything from HP's analytical products division to its medical instruments group. Hackborn is based at the company's Boise peripherals plant.
The reorganization is helping to speed things up, but HP still proceeds rather cautiously by computer-industry standards. "HP is a very conservative company, and computers are a market in which people are practicing guerrilla warfare," says ASK Computer Systems Inc. CEO Sandra Kurtzig, who resells HP computers with her company's manufacturing software. "The nice guys don't necessarily win."
In the hypercompetitive workstation market, nice guys can barely keep up. While HP and Digital Equipment Corp. were still pushing their minicomputers to engineering departments, upstarts Apollo and Sun were cleaning up by outfitting engineers with powerful desktop workstations, then linking each desktop on a network.
To their credit, HP's engineers saw the workstation movement coming. They also recognized the promise of a new computer design called reduced instruction-set computing (RISC). But Sun was moving fast. Its first RISC workstations came out in 1988, while HP's RISC design, called Precision Architecture, was still being used only on its minicomputers and on one overpriced workstation.
The purchase of Apollo, which had fallen behind Sun by clinging to proprietary designs rather than open systems, was a way for HP to expand its customer list and gain expertise in RISC workstations. Moreover, combining Apollo's workstation business with its own gave HP the No. 1 slot in the market--but only for a brief time. The company stumbled in its efforts to integrate the $654 million Apollo. While HP worked to eliminate Apollo's overlapping engineers and administrators, partly through layoffs, development of new workstations fell six months behind schedule. "We'd have been better off to bludgeon things into place faster," Platt says.
HP's workstation effort also had some bum luck: Last year, it introduced a product line based on Motorola Corp.'s 68040 chip that was supposed to keep customers happy until Snakes was ready. But Motorola was six months late with its delivery. "We can identify hundreds of millions of dollars lost," says Platt. Meanwhile, IBM began shipping the RS/6000, its first competitive RISC-based workstation, and quickly grabbed nearly 10% of the market. By the end of 1990, HP had lost 20% of Apollo's coveted customers, as well as the lead in market share.
Now, the Snakes line is scheduled to be introduced on Mar. 26. Advance word in the industry is that Snakes is a hot product, with screaming speed and all the right bells and whistles. But the product's delay has cost HP dearly. Not only are Sun and IBM already well-entrenched in the RISC market but also a coalition of 45 computer makers, including PC hotshot Compaq, is expected to make an announcement on Apr. 9 that they are backing another RISC microprocessor design, made by MIPS Computer Systems Inc.
While HP has licensed its chip to Hitachi Ltd. and Korea's giant Samsung Electronics Co., for now, the Snakes line will be the only workstations to use the HP RISC design. That means that software writers may have less incentive to develop programs for it. There are already 2,800 programs for Sun's widely cloned Sparc RISC chip and 3,000 for IBM's RS/6000. And applications software makers aren't enthusiastic about taking on yet another RISC design. "It's getting perilously late for HP," says Andrew Allison, editor of the newsletter RISC Management.
'OUT OF MIND.' Late or not, with the minicomputer business waning, HP has little choice but to move further into workstations. It must also continue to try to succeed in personal computers. Although PCs have become a $750 million business at HP, the company's sales have been so weak in the U. S. that it recently shifted its personal computer headquarters to Grenoble, France.
One bright spot has been the sales of HP PCs based on Intel Corp.'s 80486 chip. By getting into that market first, HP is now on top. The company is trying to further beef up its PC business by introducing some flashier products. In April, for example, it plans to bring out a one-pound, palm-size computer that will come with a built-in Lotus 1-2-3 spreadsheet for computing on the go. The machine, which is code-named Jaguar, is expected to cost in the range of $700.
But, says John O. Dunkle, president of market researcher WorkGroup Technologies Inc., HP has a history of not following up good products with aggressive marketing--the way that IBM, Compaq, and Apple do. "Out of sight, out of mind," says Dunkle. Indeed, marketing has never been HP's long suit. Its own managers snicker that if HP sold sushi, it would probably bill it as cold, dead fish. "We are not forceful enough in positioning ourselves in the marketplace," says Computer Systems group Vice-President Franz Nawratil.
To improve matters, HP has a new advertising campaign that, for the first time, makes direct comparisons between HP computers and competing brands. But HP will need more than advertising to sell networks of its new desktop computers. Rivals such as IBM have spent years orienting their organizations to sell systems that solve customers' problems with hardware and software combinations rather than just selling them products. On that score, HP has trailed.
Indeed, analysts say that HP has a lot to learn about working with customers. Take Alpha Industries Inc., once a big HP minicomputer customer. In late 1989, John Douglas, information systems manager at Alpha, asked both HP and DEC to bid on a project for Unix workstations, networking, and data-base software. An HP salesman ignored Alpha's product requests and instead proffered HP's IBM PC clones and HP's New Wave office-management software.
RIGHT TRACK. By contrast, recalls Douglas, DEC's salesperson arranged a meeting with its Unix experts to discuss the idea and returned with several different Unix proposals. "He had Unix people here with a configuration of the network and software before HP got off the dime," says Douglas.
Lately, there's evidence that HP has been doing a better job. US West will beef up its customer service with $25 million in HP products, and the Stock Exchange of Singapore gave HP the job of automating its trading floor. The company even bested IBM in a $44 million bid to provide Wal-Mart Stores Inc. with workstations to run networks in its stores.
Still, HP is hardly over the hump. While analysts figure it can keep revenues growing at 10% annually for the next few years, HP's real test will be in building demand for those new workstations. HP insiders say they know the company is on the right track. Lately, they have been amending Oldsmobile's advertising slogan to describe the updated HP Way: "This is not your grandfather's HP."