doubled down: since its controversial $25 billion acquisition of Compaq in 2002, now seen as Fiorina’s most important legacy, the company has invested tens of billions more either building its own factories to make PC components or forming long-term partnerships with suppliers.
David Tennenhouse, a former Intel research director who’s now a partner at San Mateo, CA-based New Venture Partners, calls this strategy “reverticalization”—in essence, a return to the vertical integration that characterized the computer industry until the 1980s. Before that time, the dominant hardware makers (think IBM, Burroughs, Unisys, NCR, Control Data, Honeywell, DEC, Prime, Wang, Data General, and HP) also sold the proprietary systems software, applications, and services needed to make their machines useful. But then came the rise of giants like Intel and Microsoft, who supplied basic platforms like CPUs and operating systems and supported a broad, horizontal system of specialization in areas like storage, peripherals, networking, and of course, software.
“There is a tendency to assume that horizontalization is the way the world should be, but it’s probably just one swing of the pendulum,” says Tennenhouse. “I think HP has done a masterful job of staying ahead of the swing, and realizing that the swing worked for HP and so jumping on it and swinging harder.”
In the PC business, the printer business, and the enterprise computing business, HP’s vertical integration “doesn’t go all the way down to the silicon, which is still horizontal, but they have really reverticalized everything going up,” says Tennenhouse. In fact, HP’s successes cutting costs through consolidation and then gaining market share may have galvanized other infotech powerhouses to follow suit. “I’m not sure what Oracle’s thinking was around the Sun acquisition, but a certain amount of it must have been just looking at what those other guys are doing,” says Tennenhouse. “I wouldn’t give Mark Hurd credit for driving the whole landscape, but you can certainly say that a lot has changed on his watch, and he was a leading player in it.”
Parts of HP’s reverticalization effort are half-formed and incomplete, of course. HP hasn’t made its own semiconductors since spinning off Agilent in 1999, though it has a close partnership with AMD, and there’s even been speculation that the perpetual second-place chipmaker might be one of HP’s next big takeover targets. (Apple and IBM are also mentioned as potential acquirers.)
And it’s not clear what role Hurd envisioned for Palm inside HP, although the obvious speculation is that HP hopes to emulate Apple’s success pursuing a vertical model in the smartphone and tablet market. (Palm was probably more valuable for WebOS, its smartphone operating system, than for its actual devices, which aren’t best-sellers.)
But in smartphones and tablets, HP will have to compete not just with Apple itself, but with an expressly horizontal model in the Google-led Android initiative, which is gaining steam quickly. So the pendulum keeps swinging—but for now, the departing CEO seems to have HP riding it in the right direction.
