50 Years In Tech. Part 2: Why HP Fell

By Jean-Louis Gassée

by Jean-Louis Gassée

Hewlett-Packard’s decline starts in 1999 when then-CEO Lew Platt, having just spun off the measuring instruments business into Agilent, bucks tradition and recruits an outsider as his replacement. Indeed, Carly Fiorina’s six years at the helm are marred by unfulfilled promises and the catastrophic $25B acquisition of Compaq, a company that’s wounded by low margins and its own disastrous acquisition (DEC, $9.6B)

Mark Hurd, Fiorina’s successor, provides hope for a brighter future…and promptly trips over (never-confirmed) allegations of “fraternization” and the sophomoric expense account fiddling meant to cover up the supposed liaison.

In comes Leo Apotheker…and out he goes, ousted in less than a year due to the ill-considered Autonomy acquisition and by rumors of his intention to spin off the PC business.

Meg Whitman replaces Apotheker in 2011 and tears the company apart, literally. She “liberates” the PC-and-printers business as HP Inc., and stays as head of HP Enterprise. After a series of quarters with uninspiring results — Bloomberg gives her the top spot in their list of underachieving CEOs for 2013 — she resigns in November 2017. (To be fair to Meg Whitman, and to keep a bit of distance from such lists, Tim Cook found himself ranked 13th “underachiever” on that same list.)

Add repeated incidents of board strife and an internecine spying scandal and you get the full picture, from Fiorina to Whitman and on into second-class citizenship…

…or so goes Silicon Vally conventional wisdom. I hold a different view: HP’s troubles started many years before Carly Fiorina walked in.

— — — — — — — — — — — — —

In November, 1970, I take my first trip to the US, to the HP offices in Loveland, Colorado where “my” desktop computer products are designed and made. I arrive in the dark of night and, when jet lag wakes me up, I open the curtains to a mile-high Technicolor scene, the impossibly exaggerated red earth white clouds blues skies of John Wayne movies. But this is real; the majestic Colorado landscape. As an introduction to the US, it can’t be beat.

I hop in the “not too big” rental car (a mere Ford Galaxie) and drive to the office for a training session. To this day, I remember how relaxed and hospitable people were, holding hands to say grace before a dinner of 28oz porterhouse steaks and weak coffee at The Rear Of The Steer (I’m not joking…that was an actual restaurant, darkly lit as the custom of the day dictated). On a frigid Saturday afternoon I see my first college football game, complete with pompon girls and cannon shots to celebrate touchdowns. On a trip to a convention in Las Vegas, we fly inside the Grand Canyon, something forbidden today. I know now this was a partial view of my future country, but I still treasure the memories.

The HP I saw was the powerful, benevolent, prosperous technology company that had started Silicon Valley before World War II. But the company was already torn between two cultures: On one side, we had science lab engineers designing and selling high-precision voltmeters, oscilloscopes, and Fournier analyzers; on the other was the sexier computer business propelled by the unrelenting progress of silicon-based logic. The tension bubbled up in seemingly petty ways, such as arguments over office flooring. The instruments team believed in traditional plastic tile; the computer engineers wanted fancy, expensive carpet (again, I’m not joking).

Despite these differences, everyone lived by HP’s unwritten credo: We design products for the engineer at the next bench. This worked well for the instruments business, and it was also plainly visible in HP’s new line of desktop and mobile computers. These were machines that only an engineer could love — or even use. (The one exception was the outstanding financial HP 80, descendants of which you can still see in offices today.)

True to the credo, HP’s instruments and computers were more bought than sold, paid for by one’s employer (invariably some sort of lab). HP had no empathy for individual users; it saw no profit in marketing to “mere mortals”. According to one persistent industry joke, HP was so inept at marketing that if they sold sushi, they would position it as “a small ball of cold rice surmounted by a piece of dead fish”.

An example: In 1977, HP created the visionary HP-01 calculator watch (picture from the refreshingly old-school Museum of HP Calculators):

A technical tour de force, the HP-01 contained six IC chips for an equivalent of 38,000 transistors. (For comparison, the 6502 microprocessor inside the contemporaneous Apple computer had fewer than 4,000 transistors.) But to HP, a watch isn’t a computer; it was only sold in…high-end jewelry stores.

Another example: The brilliant HP-35 was only sold by direct mail. Why? Not enough margin was the official explanation, although the real reason was that the company was simply indifferent to, and ignorant of, the retail business.

At HP France, I found a way around the edict: I got a Paris slide-rule retailer to collect customers’ mail-order forms against a measly commission authorized by my bosses. Orders flooded in and, at a trade show where actual selling was forbidden, we rented a small store just outside the building and delivered machines to eager customers.

In the mid-70’s things began to change. Apple started in 1976, Microsoft in 1975. They were part of a movement fueled by the advent of affordable 8-bit microprocessors from MOS Technology, Motorola, and, most notably, Intel. Ohio Scientific, Victor, Commodore Tandy… dozens of microcomputer companies sprouted in the late-70’s. As noted in the Hewlett-Packard Wikipedia page, Steve Wozniak designed the Apple I while at HP and gave the company first right of refusal. Wozniak said HP turned him down five times…

Indeed, why would HP want a puny, 8-bit, home-brew machine? It already had a “personal computer”, its ultra-sophisticated 9800 line culminated with at 16-bit 9830 with optional modules for terminal emulation and matrix computations and compatibility with hard disks from the company’s minicomputer line. It was clearly superior…but at $10K or more it failed the true PC test: a machine you could lift with you arms, your mind, and your credit card.

We’re left with a mystery: Up through the mid-70’s, HP was unequalled in its creation of a world-class line of desktop and mobile computers, including a Because We Can smartwatch. Then, nothing. Other computer companies caught on quickly: In 1976, Data General saw the tide was turning and rushed their 16-bit microNOVA design down the street to their Sunnyvale factory. Despite not lacking access to money, people or industry connections, HP didn’t see fit to design its own 16-bit microprocessor.


Jumping outside of tech, consider the world-class Volkswagen Golf introduced in 1974 to replace the beloved but outdated Beetle, in and of itself a perilous exercise. But the most remarkable story unfolds as, year after year, VW keeps improving the Golf without corrupting its identity, gestalt if you prefer: a sexy, modern small car — but not a boring appliance. VW never lost the plot. But HP did.

Blinded by its loyalty to engineers and its infatuation with the ozone of the lab, HP couldn’t see that it was about to be pacmanned by “inferior” products. To paraphrase Saint-Exupery, HP’s heart couldn’t see “it”, the future of desktop and mobile computers. Then they lost themselves in the pursuit of bigger minicomputers that were exterminated by the likes of Sun Microsystems — which would later fall to servers built on PC derivatives.

HP had the means, but one or more leaders must have lost the desire.

We now have the modest, stable $7.6B Agilent, the hard-to-characterize $29B HP Enterprise “solutions” business, and HP Inc, the $52B PC and printers business. Yes, the baton was dropped by HP’s recent Board and CEOs, but the blindness goes back decades.

— — — — — — — — — — — — —

Future "50 Years In Tech" Notes will appear at variable intervals, interleaved with more contemporaneous topics.

— JLG@mondaynote.com