BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

The PC Industry Is Digging Its Own Grave

This article is more than 10 years old.

This week, the PC community turned in its worst performance since IDC and Gartner started counting.  Worldwide shipments were off by 14% year on year, according to IDC.  Gartner says the decline was 11%.  The two differ slightly in methodology.

Those of us who have been following the industry since its inception have reason to be dismayed as to the causes.  Yes, the market has shifted to high mobility devices (i.e., smartphones and tablets), but much of the pain among PC suppliers is self-inflicted.

Of course, a fish can’t become a bird just because flying is in fashion.  There are some limitations to a company’s ability to change what it does.  But Apple proves the point that a computer company can become a high-mobility company.  And, in fact, particular vendors’ degree of pain maps almost perfectly to the degree to which they have been able to make this transition.

For example, Samsung has been making notebooks for years, but is now right in the middle of the Google -fed Android revolution in smartphones and tablets.  Lenovo, an amalgam of IBM ’s old PC company and the first major Chinese PC maker, has been moving into phones in China, India, Philippines, and Vietnam.  Lenovo also offers a variety of tablets in developed markets.

Dell and Hewlett-Packard (HP), on the other hand, have nearly non-existent positions in high mobility.  Dell, with its emphasis on computing for businesses, has reacted to organizations’ slowness to adopt new technologies by being slow to introduce them.  HP bought Palm, intending to use the smaller company’s webOS for both phones and tablets, only to abandon the project and sell off the assets after a relatively short period.  Both hardware makers are working to expand their high-mobility portfolios, but Dell is more focused on enterprise solutions, and HP has indicated that these products will not be rushed to market.

However, the real responsibility for the current situation lies with the main suppliers: Microsoft and Intel.

Although Intel, under CEO Andy Grove, was able to abandon the memory business and focus on microprocessors at a time when Japanese firms, backed by generous government subsidies, were driving their American rival out of the memory market, the company has not been able to pull a similar trick when it comes to the architectural transition from x86 to ARM.  Not that the company can’t make the transition technically.  It is philosophically against such a move.  Orthodoxy has calcified the place since Mr. Grove’s tenure.

The problem is mostly cultural.  And here, I depart from the usual disclosure practice to note not who Endpoint does have consulting contracts with but who it doesn’t.  Both Microsoft and Intel terminated their relationships with my company in the past year or so.  And the reason?  Because they don’t like independent analysts, who have to tell it like it is even when the picture isn't pretty.  They want “message force multipliers,” tame hacks who will help their public relations efforts.  Tell them what they don’t want to hear, and you’re shown the door.

Microsoft didn't like a study we ran that showed that its InTune offering was badly priced.  The company wanted marketing support, not real information.  Intel was angry that Endpoint wasn't supporting Ultrabooks enough.  Neither program has been particularly successful in reinvigorating the industry.

This bull-headed attitude is causing these suppliers to take the whole market down with them.  They refuse to see the pixels on the screen.  A contributing dynamic is the old Zen adage: “your strength is your weakness, and your weakness is your strength.”  The very profitability of Windows and x86 has kept Microsoft and Intel from looking beyond these technologies.  At Microsoft, anything that might hurt the Windows franchise is instantly killed.

But here’s the thing: the concept of the product cycle — in PCs, the three-year cycle (now four or five) — is entirely a fabrication.  The industry is only 35 years old.  The founders just made this stuff up to sell more product.

About five years ago, Intel had a program for explaining to companies how PC failure rates rocketed after three years and how productivity could be greatly improved by changing out the whole fleet on a regular basis.  But in today’s world, a PC’s local content can be mirrored to the cloud instantaneously, and if the machine itself craps out, it can be thrown away and replaced with no loss of productivity.

Hard drives, which themselves are being replaced by even more robust solid state drives, have a design life of five years.  This is the specification that drive suppliers like Seagate and Western Digital meet to sell them to PC hardware OEMs like Dell and HP.  And the drive’s moving spindle represents the most friable part of the machine.  A PC can last for a decade.

When Microsoft first shipped DOS, it did so under license.  Essentially, you bought the right to use it, but didn’t actually own the bits.  However, a version would ship, you paid for it, and it more or less worked.  Point releases (the numbers that increment after the “.” — as in, version 3.1) came out, but buyers could opt not to install them.  The process was pretty mechanical.  But every few years, Microsoft came up with a new product.  The transition from DOS to Windows did great things for the company.  And early Windows versions, with numbers like 1, 2, and 3, gave way to Windows 95, arguably the company’s most successful product ever.  Somewhere along with way, the software became stickier.  Microsoft educated buyers that it had the right to enter their machines over the network and change the bits with “updates.”  So, now, users not only didn’t own what they paid for, but didn’t really control it, either.  The rationale was bug removal, virus protection, compatibility fixes, and so on.  But the effect was increased dependence on Microsoft and its sometimes arbitrary whims.

With transitions being so profitable, can we really blame Microsoft for wanting to do it over and over?  Well, yes.  Resistance from the installed base has been detectable for years, with CIOs grousing about having to qualify an entirely new image just because Microsoft wants to turn its base over again.  Evidence of this phenomenon is how support for Windows XP, which Microsoft really wanted to cut, has been extended again and again.  Neither Microsoft nor Intel has ever accepted that some computing is “good enough,” to use a common industry phrase.

With Windows 8, Microsoft entirely screwed the pooch.  A badly conceived OS, designed to compete with Apple’s iOS and yet remain a traditional PC, did neither.  It only confused and repelled users.  Windows 7, a decent replacement for Windows XP, finally, could have sustained the industry easily for a decade.

The dilemma facing the industry is familiar to anyone who knows the children’s “cookie jar” story: if you drop half the cookies, you can take your hand out of the jar with some cookies and then go back for another modest handful.  Instead, we have the stupid industry, standing there on the stool with its hand in the jar, fistful of cookies unable to pass the rim, red-faced, crying, and looking like a fool.

Disclosure: Endpoint has consulting relationships with the PC hardware OEMs mentioned in this story.

Twitter: @RogerKay