HP Kept Hiring CEOs to Fix It. The Thing They Were Fixing Was Its Soul.

On January 1, 1939, two engineers formed a partnership with $538 of capital in a rented Palo Alto garage, and flipped a coin to decide whose name came first.¹ Sixty years later, the company that grew out of that garage hired its first CEO who was not named Hewlett or Packard and had never worked there. Over the next sixteen years it would hire four more chiefs to run the founders' company - and force two of them out, watch a third last barely eleven months, and finally hand the last one a saw and tell her to cut the place in half. The headstone everyone reads says 'leadership instability.' The real cause of death is more interesting, and far more expensive.

The official story is that HP had a talent problem - a string of bad CEO picks, a board that kept hiring the wrong outsider, simple bad luck. The truth underneath is that HP had a structural problem, and every new CEO was hired to solve it the one way that could never work: by imposing a single transformational thesis on a company whose entire identity was the founders themselves.

Every successor was hired to overwrite the founders

Look at the sequence and a pattern jumps out that no single resignation explains. Carly Fiorina arrived in July 1999 as the first outsider, charged with dragging a slow consensus culture into the internet age. Mark Hurd came in April 2005 to run it like an operator and cut. Léo Apotheker arrived in September 2010 to pivot the whole company toward software. Meg Whitman took over in September 2011 to clean up the wreckage.² Four CEOs, four wholesale reinventions, each one premised on the idea that what HP needed was a bold new direction. None of them was hired to be a custodian of what HP already was. And that was the tell, because what HP already was - the engineering-led, consensus-driven 'HP Way' - was not a strategy a CEO could swap out. It was the founders, still in the walls.

The Compaq vote was a referendum on the founders, not the deal

Fiorina's defining bet - merging with Compaq in 2002 - is usually told as a near-run proxy fight she barely won. The vote was indeed close — reported at roughly 837.9 million shares for versus 792.6 million against, a margin of less than three percent. But that headline number hides the real story. Nearly half of the 'against' votes were tied to the Hewlett and Packard family foundations. Both founding families' institutional arms — not just Walter Hewlett's side, as the legend has it — went on record opposing the deal, and Walter Hewlett filed a formal proxy statement urging shareholders to reject it.⁵ Strip the families out and the independent institutions backed Fiorina by roughly two to one. So the market liked the strategy. The people whose name was on the building did not. That gap - capital on one side, lineage on the other - is the whole story of HP's succession in a single tally.

The severance agreement ending Fiorina's employment is dated February 8, 2005 — the day the board acted, before the news became public the following morning.³ Popular memory inflated the parting into a golden parachute many times larger than the documented figure and into a dramatic public firing. Both are embellishments. The SEC filing shows a cash severance of exactly $14 million — 2.5 times her base plus target bonus — and her roughly six million stock options vesting on the way out.³ There was no firing announcement; the board asked her to step down, and Fiorina later said on CBS's 60 Minutes that she had been blindsided by the decision.⁹ The amount and the manner were both smaller and quieter than the legend. What was large was the precedent: HP had now established that its CEO could be removed when the thesis and the institution stopped fitting.

The pattern got faster, then it got terminal

Mark Hurd lasted five years by doing the one thing the institution could tolerate - operating, not reinventing - until he was forced to resign in August 2010 over inaccurate expense reports tied to a personal relationship with a contractor; HP's own investigation found the sexual-harassment allegation against him was not substantiated, but the expense irregularities were deemed a violation of its standards of business conduct. Then the board reached for the most radical thesis yet. Léo Apotheker, a software man hired to turn a hardware company into a software company, was asked to step down after eleven months.⁶ Eleven months. By that point the gap between hiring a savior and firing one had collapsed almost to nothing - which is what happens when an organization keeps prescribing surgery for a condition that is actually its skeleton.

The split that finally admitted what no CEO could fix

In 2015 Meg Whitman did the thing her four predecessors had been hired, implicitly, to avoid. She stopped trying to make one leader carry the whole legacy and divided it. The old Hewlett-Packard Company renamed itself HP Inc. and kept the PCs and printers; Hewlett Packard Enterprise spun out as a separate public company with the servers, storage, and services. Whitman ran HPE; Dion Weisler took HP Inc.⁷ This is celebrated as strategic clarity, and operationally it was. But read it as a succession outcome and it says something blunter: the founders had built a company that no single successor could run, because the 'HP Way' had to mean one thing and the two businesses needed it to mean two different things. The split was not the strategy finally arriving. It was the board conceding, after sixteen years and a graveyard of theses, that the original problem had no single-CEO answer.

The fair objection is that this is too tidy - that HP's troubles were ordinary corporate misfortune, not a curse written into its DNA. Fiorina's stock decline overlapped almost exactly with the dot-com bust that flattened every large-cap tech company, so blaming her for all of it ignores the macro collapse. Hurd left over an expense scandal, not a strategy failure. Apotheker may simply have been a poor pick. Each chapter has a mundane explanation, and each is partly true. But the steelman has a hole: mundane explanations don't repeat in a pattern this clean. One forced exit is bad luck; an eleven-month CEO and a company sawn in two is a system telling you something. The individual stories vary. The shape - outsider hired to overwrite the founders, outsider rejected by the founders' company - does not.

Two engineers flipped a coin in a garage to decide whose name came first, and for forty years it didn't matter, because they were both there. The moment they weren't, the question of whose company it really was became unanswerable - and HP spent sixteen years and four CEOs discovering that you cannot inherit a culture the way you inherit a balance sheet. In the end it didn't find the right successor. It found that the job, as built, could not be done by one - and split the name back in two.