Before a couple gets engaged, they often consult with their family and friends as a sanity check, to see if the marriage makes sense. In the case of Hewlett-Packard and Compaq, which announced their engagement at a news conference last September, the advice was loud and clear: stay single!

The merger announcement was made at the Copley Marriott Hotel in Boston by Hewlett-Packard CEO Carly Fiorina and Compaq head Michael D. Capellas. If the merger were to go through (and at this writing, that's uncertain), it would be the largest in computer industry history. At the time of the announcement, the combined company would have had assets of around $56 billion, annual revenues of about $87 billion, and annual operating earnings of $3.9 billion, with operations in more than 160 countries and more than 145,000 employees. With falling stock prices and industry layoffs, the real numbers will be lower. Even so, a merged company would approach IBM in size.

While called a merger, it would not be a marriage of equals. Hewlett-Packard's shareholders would own approximately 64 percent of the company, and the merged company would be based in the Palo Alto, HP's hometown. HP's Fiorina would be chairman and chief executive, with Capellas serving under her as president. The new name has not been revealed, but "Hewlett-Compaq," "HewPaq," and "ComPackard" are unlikely. Fiorina has called the proposed company the "new HP."

At the news conference, the two company heads painted a rosy picture, but as soon as they left the stage, many proclaimed the marriage, if not a mistake, at least an act of desperation. "They are two pathetic companies just trying to survive," said Sanjay Jhaveri, a technology analyst at Zurich-based Vontobel Asset Management, to Business Week magazine. "Their backs are against the wall." "This is not a case of 1+1=2," said Todd Kort, principal analyst for Gartner Research, to Time magazine. "More like 1+1=1.5."

Fortune magazine said that what ails HP and Compaq aren't bloated or redundant costs. "The real problem facing both these companies, who've become almost clones of one another, is that they are not positioned to take advantage of the new areas of growth where value---and profits---are greatest" in software, services, and storage.

Encouragement did come from some influential sources, including Intel CEO Craig R. Barrett and Oracle CEO Larry Ellison. "This will be the biggest hardware company in the world, and that gives them tremendous market power," Ellison told Business Week. "Keep in mind that the [stock] market is the one that said Ariba Inc. was more valuable than Daimler Benz. That's the market for you." The research firm IDC calling the deal "as significant as any in the evolution of the IT industry."

But those pro-merger thoughts were lost on Walter B. Hewlett and David W. Packard, sons of the late HP founders, who announced they would oppose the deal. Hewlett spoke on behalf of his family, which controls 5 percent of HP's stock. Packard controls another 1.3 percent. A former professor of Greek at UCLA, Packard said the merger would go against his father's philosophy of hiring people for the long-term. "My father never said everyone has tenure," he said to the Wall Street Journal. "But the company took very seriously the importance of employees and having a long-term outlook." The David and Lucile Packard Foundation, which owns 10 percent of HP, said it was still looking into the matter, and some consider its vote decisive. At this writing, analysts are giving the merger only a 50-50 chance of succeeding, and are predicting that if it doesn't go through, it will be a vote of no confidence for Fiorina.

Anti-merger mania

With so many computer industry mergers, why so much resistance to this one? Part of the reason is the inability of so many mergers to live up to their hype. All are announced to great fanfare, with promises of synergy and operating efficiency, which usually translates into layoffs. But once out of the spotlight, many mergers grow troubled.

Some large acquisitions have also been problematic. Such was the case of Compaq's acquisition of Digital Equipment Corp, a $9.6 billion deal that was, in its time, the largest in computer industry history. That deal, said the Economist, "proved to be a total disaster, distracting management at a time when the merged firm's core businesses badly needed more attention." On the other hand, Compaq swallowed up Tandem Computer with little indigestion. Hewlett-Packard's smaller purchases have raised few concerns. These include a proposed acquisition of Comdisco, which provides IT services, as well a buyouts of Bluestone Software, maker of J2EE and XML application server technology, and Dazel Corp., which made electronic information delivery software.

But this deal is different. It combines two companies that have been seen more or less as equals in the computer business. Both have lines of PCs, low-end servers, and hand-held PCs running Microsoft's Pocket PC operating system. HP is distinguished by its 50 percent share of the printer business.

The problem with mergers of this scope is two-fold: cultural and technical. Companies have cultures---a set of working assumptions, legends, and personalities, which often reflect the leader at the top. Apple's Steve Jobs operates differently than IBM's Lou Gerstner. Oracle's Larry Ellison is a different animal than Bill Gates, and so are their respective companies. Cultures not only determine how companies operate, the chances they take, the research they invest in, the products they produce, but the employees they attract. When two companies merge, inevitably, no matter how much the CEOs protest it won't happen, one culture predominates and talent flees. That was the case when AOL bought Netscape. And could well be the case here---where Compaq's Texas-based culture would be absorbed by the Silicon Valley.

The technical issues brought on by mergers are more quantifiable. A combined Hewlett-Packard/Compaq would be the world's largest PC maker and major force in the low-end server market. The combined companies are weaker when it comes to high-end servers, compared with Sun and IBM, and while they have a service infrastructure, they are at the less profitable low end of the business. For uniprocessor servers, HP has had success with HP-UX, its flavor of UNIX, in the 64-bit Intel IA-64. Compaq has made less of a splash with its own UNIX version, Tru64, and the betting is that HP's OS will prevail. Or perhaps a merged company would put even more emphasis on Linux.

PC business in the doldrums

The most troubling aspect of the HP/Compaq deal is what it says about the state of the PC industry. Born in the late 1970s, burgeoning during the dot.com boom of the 1990s, the PC business has stalled out. Or maybe we've just become spoiled. Perhaps all that's happened is that the PC industry now resembles other parts of the consumer good market. How often do you have to replace your television, your compact disk player, and for that matter, your refrigerator? PCs may be becoming more like that. But for an industry accustomed to burgeoning growth, a slower growth model will be painful.

"While there's plenty of potential, this deal is beset by mind-numbing problems," wrote Business Week. "If the deal goes through, the merged company will have a 19% share, making [Hewlett-Packard CEO Carly] Fiorina the global PC capital queen---just as the PC is dealing with the worst-ever downturn that has resulted in $1.2 billion in losses and 31,000 layoffs so far this year." The research firm Dataquest said that in 2001, U.S. PC sales dropped 18.7 percent, with worldwide shipment dropping 11.6 percent. International Data Corporation put those numbers at 21 percent and 13.7 percent, respectively.

The problem facing Hewlett-Packard and Compaq is in distinguishing their brands. Besides the logos, what's the difference? Wall Street Journal columnist Walter S. Mossberg argues that "the desktop PC has become a boring commodity device, shorn of most of its variety and innovation." He notes that 72 PC makers, including some real innovators, competed in 1991. Now that number is fewer than ten, and it continues to shrink. "Building personal computers, once an exciting and lucrative business, has become a low margin stultifying affair. Most PC makers are now little more than resellers for Microsoft software and Intel chipsets, wrapping metal and plastic around those products. Nearly all the profits are sucked up by Microsoft and Intel, and those two companies perform nearly all the research in the PC business."

HP and Compaq don't agree. Citing a larger R&D budget, Cabellas says that innovation is what will separate HP/Compaq from Dell, enabling the new HP to lead in new product categories. But one could argue that Dell's method of doing business is itself an industry milestone. "What Dell has been able to do in PC marketing---building computers to spec and delivering them quickly---has been a remarkable innovation," says David Allison, chairman of the Division of Information Technology and Society at the Smithsonian Museum of American History.

John C. Toole, executive director and CEO of the Computer History Museum, notes that the biggest innovations in the PC industry happened early on, with IBM's initial architecture, and "luggables"---the laptop's heavy great uncles---from Osborne and Compaq. Memory has gotten cheaper, disk drives larger, while buses and I/O connections have gotten faster---but these innovations have by and large come from component manufacturers, not PC makers. "The fundamental things about a PC have been commoditized," Toole says. "The innovation is happening elsewhere---on PDAs and other portable electronics, for example."

Ironically, Compaq was one company to benefit from recognizing that the key to PC hardware was not innovation, but cloning. The Compaq Deskpro, introduced in 1984, was the first PC clone---able to run the same applications as the IBM PC. While this "me too" strategy solidified the PC architecture, it's principal claim to fame was its ability to run Flight Simulator---at that time, the benchmark for PC compatibility.

Other than on price, PC brands scarcely compete. It matters less than ever whose name is on the box. Inside, you know you're getting an Intel chip or a clone, and whatever version of Windows Microsoft is currently pushing. The casing is similar. Telephone support is universally dicey. Reliability is more or less the same, with similar guarantees in case of failure. More than televisions, automobiles or toothbrushes, choosing between PC brands has become one of image and brand identity, not features and innovation.

From the consumer's point of view, there has never been a better time to buy a PC---because PC makers have never been so hungry to sell you one. In the U.S., personal computers are now sold primarily by two channels, both set-up for mass merchandising. You can purchase a system configured to your specifications from a mail order company, primarily Dell and Gateway, which also operates 300 stores across the U.S. Or you can get one off-the-shelf, quite literally, from discounters like Costco, Staples, and Best Buy. At Costco, you can choose your computer, shop for discounted applications like Microsoft Office from a large table, and pick up some packaged salmon and a pound of chocolate for desert. The cashier gives you a voucher slip for the computer and they hand you the box, which fits in the over-sized cart. You roll it back to your car and you're set: ready for dinner and computing by night. The shopping experience is no different from buying a VCR or television: even the setup may be no more difficult.

The off-the-shelf channel is so compelling that Dell itself has gone after it with its SmartSteps PC, a pre-configured machine with prices starting at $600. It is the lowest price home computer the company has ever sold, and competes directly with pre-configured machines from Compaq and Hewlett-Packard.

The XP panacea?

The PC industry has also stalled because, finally, it seems to have run out of killer applications. Moore's Law may still be in effect, but for most people it no longer matters. If your PC is fast enough to run Windows 95, that's enough to run office programs, to download and send email, to read HTML pages off the web. The fastest PC games may require more power, but most gamers will buy a Sony Playstation, a Sony Game Cube, or perhaps, a Microsoft Xbox. And if you want to want to play chess, that application will run just fine on an x286.

It's no wonder then that the PC hardware industry has gazed with lust at Windows XP, Microsoft's newest operating system, the first to put NT technology onto PCs bound for the home consumer market. Estimates vary over the additional computation that XP requires. Microsoft, itself, suggests a CPU with at least a 300 MHz clock speed, and 128MB of RAM. Others have put those numbers higher. If XP were a "must have" purchase, one might expect a stampede of users exchanging their slower Pentium machines for more souped up models.

But Windows XP does not merit that kind of enthusiasm. Its single most useful attribute is stability, which Microsoft has been promising for years and is already available in Windows 2000. Beyond that, XP provides instant messaging and makes it easier to create, edit and share digital movies and photos, , burn CDs and download music---not with MP3 files but with Microsoft's proprietary format. There are some enhancements for games, provisions for multiple users within a household, and a firewall for networking.

On the minus side, the features of XP have come in for much criticism. Windows Product Activation (WPA) requires that users lock their version of the operating system to their hardware. As an anti-piracy move, WPA is understandable and Microsoft has every right to charge what it specifies on its license: a fee for every PC. And consumers have every right not to buy the product. The idea that you must call Microsoft if you make too many changes to your PC also does not sit well. Like most computer companies, Microsoft support is tough to reach, especially if you want a living, breathing human being.

Microsoft's Passport feature, which stores personal information to make web transactions easier, has also garnered criticism. Concrete benefits are hard to find and the fees involved seem to benefit Microsoft more than the consumer. Microsoft has not hidden its intent to earn more of its revenues from subscriptions, thereby getting a steadier flow of revenue. But customers may not be sympathetic---or comfortable handing over personal information to Bill Gates and company.

The problem of diminishing returns has also affected Intel, which during the latest World Series ran awkward advertisements for the Pentium 4. The ads featured computer-generated extra-terrestrials, which makes some sense given that graphics are the sole remaining class of desktop application to benefit from faster processing power. One suggested application, Photoshop-like manipulation of digital images, is a clear winner for the faster Pentium. But if you use your PC to do the household budget and send email, the extra power is overkill.

With the PC industry in the doldrums, what is one to make of a combined HP and Compaq? If such a merger were to matter, it would make its impact not in market share, but research and development. What does it matter if a company rivals IBM in size or is the largest PC maker or provides competition for Dell? What has driven the computer industry since the beginning has been experimentation and brainpower. If a company want to compare itself with IBM, that's the place to start. IBM, with its formidable research facilities, has been the leader in U.S. patents for the past seven years. That gauntlet is what HP, Compaq, or HP/Compaq must pick up if it expects to drive the PC industry forward.